ProgramadeGestionHCR 2013-2019

Confeccionado por: Grenedys López Tápanes.
Curso 2012-2013
Febrero 2013
Índice
Introducción…………………………………………………………………………..7
1. Almeida ASd, Picon PD, Wender OCB.

Resultados de pacientes submetidos à cirurgia de substituição valvar
aórtica usando próteses mecânicas ou biológicas.
Rev Bras Cir Cardiovasc [Internet]. 2011;26:326-37; Disponible en:
http://www.scielo.br/scielo.php?script=sci_arttext&pid=S0102-
76382011000300006&nrm=iso. ………………………………………….……….10
2. Anderson AJPG, Barros Neto FXdR, Costa MdA, Dantas LD, Hueb AC,
Prata MF.
Preditores de mortalidade em pacientes acima de 70 anos na
revascularização miocárdica ou troca valvar com circulação
extracorpórea.
76382011000100014&nrm=iso. …………………………………………………..22
3. Badreldin AMA, Doerr F, Ismail MM, Heldwein MB, Lehmann T, Bayer O,

et al.
Comparison between Sequential Organ Failure Assessment Score (SOFA)
and Cardiac Surgery Score (CASUS) for Mortality Prediction after Cardiac
Surgery.
Thorac Cardiovasc Surg [Internet]. 2012;60(01):35-42; Disponible en:
http://dx.doi.org/10.1055/s-0030-1270943. ………………………………….....29
4. Barros de Oliveira Sá MP, Barros de Oliveira Sá MV, Lopes

deAlbuquerque AC, Barreto Gomes da Silva B, Muniz de Siqueira JW,
Santos de Brito PR, et al.
GuaragnaSCORE satisfactorily predicts outcomes in heart valve surgery
in a Brazilian hospital.
Rev Bras Cir Cardiovasc [Internet]. 2012;27(1):1-6; Disponible en:
76382012000100003&nrm=iso. ………………………………………………….37
5. Bhukal I, Solanki SL, Ramaswamy S, Yaddanapudi LN, Jain A, Kumar P.

Perioperative predictors of morbidity and mortality following cardiac
surgery under cardiopulmonary bypass.
Saudi J Anaesth [Internet]. 2012;6(3):242-7; Disponible en:
http://dx.doi.org/10.4103/1658-354X.101215. ……………………………….....43
6. Cantero MA, Almeida RMS, Galhardo R.

Análise dos resultados imediatos da cirurgia de revascularização do
miocárdio com e sem circulação extracorpórea.
76382012000100007&nrm=iso. …………………………………………….……55
7. Cordeiro da Rocha AS, Monassa Pittella FJ, Rocha de Lorenzo A,
Barzan V, Colafranceschi AS, Reis Brito JO, et al.
A idade influencia os desfechos em pacientes com idade igual ou
superior a 70 anos submetidos à cirurgia de revascularização miocárdica
isolada.
76382012000100008&nrm=iso ………………………………………………….62
8. Di Leoni Ferrari A, Pelzer Süssenbach C, Vieira da Costa Guaragna JC,

Escobar Piccoli JdC, Ferreira Gazzoni G, Klein Ferreira D, et al. Bloqueio
atrioventricular no pós-operatório de cirurgia cardíaca valvar: incidência,
fatores de risco e evolução hospitalar.
76382011000300010&nrm=iso. …………………………………………………..69
9. Dorneles Cde C, Bodanese LC, Guaragna JC, Macagnan FE, Coelho JC,
Borges AP, et al.
The impact of blood transfusion on morbidity and mortality after cardiac
surgery.
http://www.scielo.br/pdf/rbccv/v26n2/en_v26n2a12.pdf. ………………......78
10. Espinoza G R.
Tendencia en volumen hospitalario y mortalidad operatoria para cirugías
de alto riesgo.
Rev Chil Cir [Internet]. 2012;64:319-; Disponible en:
http://www.scielo.cl/scielo.php?script=sci_arttext&pid=S0718-
40262012000300018&lng=es. …………………………………………………….86
11. Garcia-Hernandez JA, Benitez-Gomez IL, Martinez-Lopez AI, Praena-

Fernandez JM, Cano-Franco J, Loscertales-Abril M.
Marcadores pronósticos de mortalidad en el postoperatorio de las
cardiopatías congénitas.
An Pediatr (Barc) [Internet]. 2012;77(6):366-73; Disponible en:
http://dx.doi.org/10.1016/j.anpedi.2012.03.021 ……………………………….87
12. Holm J, Hakanson E, Vanky F, Svedjeholm R.

Mixed venous oxygen saturation predicts short- and long-term outcome
after coronary artery bypass grafting surgery: a retrospective cohort
analysis.
Br J Anaesth [Internet]. 2011;107(3):344-50; Disponible en:
http://dx.doi.org/10.1093/bja/aer166 …………………………………………....95
13. Javierre C, Ricart A, Manez R, Farrero E, Carrio ML, Rodriguez-Castro
D, et al.
Age and sex differences in perioperative myocardial infarction after
cardiac surgery.
Interact Cardiovasc Thorac Surg [Internet]. 2012;15(1):28-32; Disponible
en: http://dx.doi.org/10.1093/icvts/ivr13 …………………………………..…102
14. Kristensen KL, Rauer LJ, Mortensen PE, Kjeldsen BJ.

Reoperation for bleeding in cardiac surgery.
en: http://dx.doi.org/10.1093/icvts/ivs050. ……………………………….…..107
15. Lee DH, Buth KJ, Martin BJ, Yip AM, Hirsch GM.
Frail patients are at increased risk for mortality and prolonged
institutional care after cardiac surgery.
Circulation [Internet]. 2010;121(8):973-8 Disponible en:
http://dx.doi.org/CIRCULATIONAHA.108.84143710.1161. ……………..….112
16. Lopez-Delgado JC, Esteve F, Javierre C, Perez X, Torrado H, Carrio

ML, et al.
Short-term independent mortality risk factors in patients with cirrhosis
undergoing cardiac surgery.
Interact Cardiovasc Thorac Surg [Internet]. 2012;Publicado en línea con
antelación:[aprox. 14 pp.]; Disponible en:
http://dx.doi.org/0.1093/icvts/ivs501 ……………………………………..…..118
17. Mejía OAV, Lisboa LAF, Tiveron MG, Santiago JAD, Tineli RA, Dallan
LAO, et al.
Cirurgia de revascularização miocárdica na fase aguda do infarto: análise
dos fatores preditores de mortalidade intra-hospitalar.
76382012000100011&nrm=iso. ………………………………………………..125
18. Morlans Hernández K, Santos Gracia J, Cáceres Lóriga FM, Pérez

López H, Mirza Saadat A.
Factores de riesgo y evaluación del riesgo de muerte hospitalaria en la
sustitución valvular mitral con prótesis mecánica.
Rev Cubana Cir [Internet]. 2003;42(2):[aprox. 14 pp.]; Disponible en:
http://scielo.sld.cu/scielo.php?script=sci_arttext&pid=S0034-
74932003000200006&nrm=iso. …………………………………………….…..134
19. Pang PY, Sin YK, Lim CH, Su JW, Chua YL.
Outcome and survival analysis of intestinal ischaemia following cardiac
surgery.
en: http://dx.doi.org/10.1093/icvts/ivs181 ………………………………...….141
20. Pearse RM, Moreno RP, Bauer P, Pelosi P, Metnitz P, Spies C, et al.
Mortality after surgery in Europe: a 7 day cohort study.
Lancet [Internet]. 2012;380(9847):1059-65; Disponible en:
http://dx.doi.org/10.16/S0140-6736(12)61148-9. ………………………….....145
21. Radaelli G, Bodanese LC, Guaragna JCVdC, Borges AP, Goldani MA,
Petracco JB, et al.
O uso de inibidores da enzima conversora de angiotensina e sua relação
com eventos no pós-operatório de cirurgia de revascularização
miocárdica.
76382011000300011&nrm=iso. …………………………………………………164
22. Rahmanian PB, Adams DH, Castillo JG, Carpentier A, Filsoufi F.

Predicting Hospital Mortality and Analysis of Long-Term Survival After
Major Noncardiac Complications in Cardiac Surgery Patients.
Ann Thorac Surg [Internet]. 2010;90(4):1221-9; Disponible en:
http://www.sciencedirect.com/science/article/pii/S0003497510011458.
………………………………………………………………………………………..171
23. Ried M, Haneya A, Kolat P, Potzger T, Puehler T, Schmid C, et al.

Acute Renal Dysfunction Does Not Develop More Frequently Among
Octogenarians Compared to Septuagenarians after Cardiac Surgery.
Thorac Cardiovasc Surg [Internet]. 2012;60(1):51-6; Disponible en:
http://dx.doi.org/10.1055/s-0031-1295567. …………………………………..180
24. Schloss B, Gulati P, Yu L, Abdel-Rasoul M, O'Brien W, Von Visger J, et

al.
Impact of aprotinin and renal function on mortality: a retrospective single
center analysis.
J Cardiothorac Surg [Internet]. 2011;6:103; Disponible en:
http://dx.doi.org/10.1186/749-8090-6-103 ……………………………………186
25. Schulman S.
Pharmacologic tools to reduce bleeding in surgery.
Hematology Am Soc Hematol Educ Program [Internet]. 2012;2012:517-21;
Disponible en: http://dx.doi.org/10.1182/asheducation-2012.1.517 ……..191
26. Shahin J, DeVarennes B, Tse CW, Amarica DA, Dial S.

The relationship between inotrope exposure, six-hour postoperative
physiological variables, hospital mortality and renal dysfunction in
patients undergoing cardiac surgery.
Crit Care [Internet]. 2011;15(4):R162; Disponible en:
http://dx.doi.org/10.1186/cc10302. ………………………………………….....196
27. ter Horst R, Markou AL, Noyez L.
Prognostic value of preoperative quality of life on mortality after isolated
elective myocardial revascularization.
en: http://dx.doi.org/10.1093/icvts/ivs184 ……………………………...…….206
28. Vlaar AP, Cornet AD, Hofstra JJ, Porcelijn L, Beishuizen A, Kulik W, et
al.
The effect of blood transfusion on pulmonary permeability in cardiac
surgery patients: a prospective multicenter cohort study.
Transfusion [Internet]. 2012;52(1):82-90; Disponible en:
http://dx.doi.org/10.1111/j.537-2995.011.03231.x. ……………………….....210
29. Zangrillo A, Biondi-Zoccai G, Ponschab M, Greco M, Corno L, Covello

RD, et al.
Milrinone and Mortality in Adult Cardiac Surgery: A Meta-analysis.
J Cardiothorac Vasc Anesth [Internet]. 2012;26(1):70-7; Disponible en:
http://dx.doi.org/http://www.sciencedirect.com/science/article/pii/S105307
701100468X. ………………………………………………………………………219
La Mortalidad
Intrahospitalaria en la
Cirugía Cardiovascular
Dr. Osvaldo González Alfonso
y Lic. Grenedys López
Tápanes.
Introducción: La mortalidad
intrahospitalaria derivada de la
cirugía cardiovascular es motivo
de renovado interés, con el
desarrollo de nuevas técnicas quirúrgica menos invasivas y de protección
miocárdica más integral. La mortalidad intrahospitalaria es sin duda uno de los
marcadores mes evidentes del desempeño de cualquier equipo quirúrgico,
pero también es esencial para valorar la introducción de estas nuevas técnicas.
La predicción del riesgo de morir es un tema recurrente entre los artículos
presentados aquí, de acuerdo a la patología, de base, la edad, el sexo y hasta
algunas comorbilidades como la diabetes u otras complicaciones como el
infarto perioperatorio, la insuficiencia renal, el uso de transfusiones, entre otras.
Esperamos que este paquete informativo le sea de utilidad.
Tema de la Búsqueda:
La Mortalidad Intrahospitalaria en la Cirugía Cardiovascular.
Estrategia de Búsqueda:
La estrategia de búsqueda se basó en la localización de artículos de revisión a
texto completo referenciados en las tablas de contenidos obtenidas por
suscripción a cartas electrónicas de alerta por materia y por publicaciones en
revistas de alto impacto y referencias cruzadas en la bibliografía de los mismos.
Se realizó una búsqueda en PubMed con los MESH apropiados al caso con
un gestor de referencias (EndNote) con limitantes en humanos y entre los años
2013- 2008, mientras que se realizó una búsqueda para obtener los textos
completos con los descriptores en SciELO Regional y EBSCO por último se
hizo una búsqueda avanzada en Hinari en las revistas de impacto mediante el
DOI obtenido con las citas del gestor de referencias y se colocó todo los
resultados en una biblioteca dentro de un gestor de referencias (EndNote) cuyo
resumen es el siguiente:
Nombre de la biblioteca virtual: mortalidad intrahospitalaria cir. cardiovasc.enl

Localización: H:\Bibliografia en EndNote\mortalidad intrahospitalaria cir.
cardiovasc.enl
Última actualización: martes, 26 de febrero de 2013 02:36:32
Registros: 196
Subgrupos: 2
Tipos de referencias Usados: 1
Tipos de referencias más usados: Artículo de revista
Textos completos: 29
Figuras: 0
Lista de Términos por Autores: 1361
Lista de Términos por Revistas: 78
Lista de Términos por Palabras Claves: 313
Las palabras claves por las que hicimos la búsqueda: octogenarians,

cardiac surgery, acute kidney injury, renal dysfunction, mortality, Age Factors
Aged, Analysis of Variance,Biological Markers/blood, Cardiac Surgical
Procedures/ adverse effects/mortality, Cardiopulmonary Bypass/adverse effects
Female, Hospital Mortality, Humans, Incidence, Intensive Care Units, Length of
Stay, Male, Myocardial Infarction/blood/ etiology/mortality, Perioperative Period
Risk Assessment, Risk Factors, Sex Factors
Fuente Núcleo: En este paquete informativo sobre la Mortalidad

Intrahospitalaria en la Cirugía Cardiovascular, la fuente núcleo principal
consiste en artículos a texto completo del tema en revistas como: Rev Bras Cir
Cardiovasc (9 artículos 2012 -2011), Interact Cardiovasc Thorac Surg (5
artículos 2012), entre otras y se brindan las direcciones de los autores más
importantes y productivos para poderlos contactar de ser necesario como son:
Dr. Giovanni Landoni,

Department of Cardiothoracic Anesthesia and Intensive Care, Istituto
Scientifico San Raffaele, Via Olgettina 60, Milan 20132, Italy.
landoni.giovanni@hsr.it
Dr. Alexander P J. Vlaar,

Laboratory of Experimental Intensive Care and Anesthesiology
(L.E.I.C.A.),Academic Medical Center, Room M0-228, Meibergdreef 9,
Amsterdam, 1105 AZ, the Netherlands;
a.p.vlaar@amc.uva.nl.
Dr. Luc Noyez

Department of Cardio-Thoracic Surgery—677, Heart Center, Radboud
University Nijmegen Medical Center, PO Box 9101, 6500 HB Nijmegen,
Netherlands.
l.noyez@ctc.umcn.nl
Dr. Juan Carlos López Delgado,

Department of Intensive Care, Hospital Universitari de Bellvitge, IDIBELL
(Institut d’Investigació Biomèdica Bellvitge; Biomedical Investigation
Institute of Bellvitge), C/Feixa Llarga s/n. 08907 L’Hospitalet de Llobregat,
Barcelona, Spain.
juancarloslopezde@hotmail.com
Dr. Katrine Lawaetz Kristensen,

Department of Cardio-Thoracic and Vascular Surgery, Odense University
Hospital, Odense, Denmark.
katrinelawaetz@gmail.com
Dr. Javierre, Casimiro
Department of Physiological Sciences II, School of Medicine, University of
Barcelona, IDIBELL, L'Hospitalet, Barcelona, Spain.
cjavierre@ub.edu
Dr. J.A. García-Hernández

Unidad de Gestion Clinica de Cuidados Criticos y Urgencias, Hospital
Infantil Universitario Virgen del Rocio, Sevilla, Espana.
garcier@gmail.com
Dr. Antônio Sérgio Cordeiro da Rocha

Coordenação de Pesquisa Clínica Rua das Laranjeiras, 374/5º andar – Rio
de Janeiro, RJ Brasil – Zip Code: 22040-006.
ascrbr@centroin.com.br
Marcos Antonio Cantero. Rua Delmar de Oliveira, 1725

Vila Progresso –Dourados, MS
Brasil – Zip Code: 79825-115
E-mail: marcoscantero@sbccv.org.br
Dr. Michel Pompeu Barros de Oliveira Sá

Av. Eng. Domingos Ferreira, 4172/405 – Recife, PE, Brazil – ZIP code:
51021-040.
michel_pompeu@yahoo.com.br
ORIGINAL ARTICLE Rev Bras Cir Cardiovasc 2011;26(3):326-37
Outcomes of patients subjected to aortic valve

replacement surgery using mechanical or
biological prostheses
Resultados de pacientes submetidos à cirurgia de substituição valvar aórtica usando próteses mecânicas
ou biológicas
Adriana Silveira de Almeida1, Paulo Dornelles Picon2, Orlando Carlos Belmonte Wender3
DOI: 10.5935/1678-9741.20110006 RBCCV 44205-1287
Abstract (P=0.057). Factors associated with reoperation were: renal

Objective: This paper evaluates outcomes in patients failure, prosthesis endocarditis and age. Probability free of
subjected to surgery for replacement of the aortic valve using bleeding events at 5, 10 and 15 years after surgery using
biological or mechanical substitutes, where selection of the mechanical substitute was 94.5%, 91.7% and 91.7% and, for
type of prosthesis is relevant. bioprostheses, was 98.6%, 97.8% and 97.8%, respectively
Methods: Three hundred and one patients, randomly (P=0.047). Factors associated with bleeding events were:
selected, who had been subjected to aortic valve replacement renal failure and mechanical prostheses.
surgery between 1990 and 2005, with a maximum follow-up Conclusions: The authors have concluded that: 1)
period of 20 years. mortality was statistically similar in the groups; 2) patient
Results: Survival at 5, 10 and 15 years after surgery using characteristics at baseline were a major determinant of late
mechanical substitute was 83.9%, 75.4% and 60.2% and, for mortality after surgery; 3) there was a tendency toward
biological substitute, was 89.3%, 70.4% and 58.4%, reoperation in the bioprostheses group; 4) patients using
respectively (P=0.939). Factors associated with death were: mechanical prosthesis had more bleeding events as time
age, obesity, pulmonary disease, arrhythmia, bleeding and passed; 5) data presented in this paper is in accordance with
aortic valve failure. Probability free of reoperation for these current literature.
patients at 5, 10 and 15 years after surgery using mechanical
substitute was 97.9%, 95.8% and 95.8% and, for those using Descriptors: Bioprosthesis. Heart Valve Prosthesis
bioprostheses, was 94.6%, 91.0% and 83.3%, respectively Implantation. Aortic Valve. Heart Valve Prosthesis.
1. Specialist in Cardiovascular Surgery at MEC/Master’s Degree in This study was carried out at Clinics Hospital of Porto Alegre, Porto
Cardiology and Cardiovascular Sciences at Federal University Alegre, RS, Brazil.
of Rio Grande do Sul (UFRGS); Cardiovascular Surgeon Physician
of the Cardiovascular Surgery Service at Hospital Nossa Senhora Correspondence address: Adriana S. de Almeida
da Conceição, Grupo Hospitalar Conceição, Porto Alegre, RS, Av. Princesa Isabel, 729 – sala 403 – Porto Alegre, RS
Brazil. Brazil – Zip Code: 90620-001.
2. PhD in Cardiology at UFRGS; Adjunct Professor of the Internal E-mail: adriana@analisys.com.br
Medicine Department at UFRGS Medical School, Porto Alegre,
RS, Brazil. Financial Support: Fundo de Incentivo à Pesquisa e Eventos do Hospital
3. PhD in Medicine at Munich University, Germany; Associate de Clínicas de Porto Alegre (FIPE/HCPA).
Professor of the Surgery Department of UFRGS Medical School,
Head of the Cardiovascular Surgery Service at Clinics Hospital of Article received on March 25th, 2011
Porto Alegre, Porto Alegre, RS, Brazil. Article accepted on August 30th, 2011
326
Almeida AS, et al. - Outcomes of patients subjected to aortic valve Rev Bras Cir Cardiovasc 2011;26(3):326-37
replacement surgery using mechanical or biological prostheses
Resumo reoperação foram: insuficiência renal, endocardite de prótese

Objetivo: Esse estudo avalia resultados em pacientes e idade. Probabilidade livre de eventos hemorrágicos em 5,
submetidos à cirurgia para troca valvar aórtica utilizando 10 e 15 anos após cirurgia utilizando substituto mecânico
substituto biológico ou mecânico, com poder de relevância foi de 94,5%, 91,7% e 91,7% e, para bioprótese, foi de 98,6%,
na seleção do tipo da prótese. 97,8% e 97,8%, respectivamente (P=0,047). Fatores
Métodos: Foram selecionados, randomicamente, 301 associados com eventos hemorrágicos foram: insuficiência
pacientes submetidos à cirurgia para troca valvar aórtica renal e prótese mecânica.
entre 1990 e 2005, com seguimento máximo de 20 anos. Conclusões: Os autores concluíram que: 1) mortalidade
Resultados: Sobrevivência em 5, 10 e 15 anos após cirurgia foi estatisticamente semelhante entre os grupos; 2)
utilizando substituto mecânico foi de 83,9%, 75,4% e 60,2% características basais dos pacientes foram os maiores
e, para substituto biológico, foi de 89,3%, 70,4% e 58,4%, determinantes de mortalidade tardia após a cirurgia; 3)
respectivamente (P=0,939). Fatores associados com óbito houve tendência à reoperação para o grupo com bioprótese;
foram: idade, obesidade, doença pulmonar, arritmias, eventos 4) pacientes com prótese mecânica tiveram mais eventos
hemorrágicos e insuficiência valvar aórtica. Probabilidade hemorrágicos ao longo do tempo; 5) dados encontrados no
livre de reoperação desses pacientes em 5, 10 e 15 anos após presente estudo são concordantes com a literatura atual.
cirurgia utilizando substituto mecânico foi de 97,9%, 95,8%
e 95,8% e, para bioprótese, foi de 94,6%, 91,0% e 83,3%, Descritores: Bioprótese. Implante de Prótese de Valva
respectivamente (P=0,057). Fatores associados com Cardíaca. Valva Aórtica. Próteses Valvulares Cardíacas.
INTRODUCTION influence of comorbidities on the outcome in a period of 20

years.
Aortic valve replacement is the standard surgical Within this context, the aim of this study was to assess
procedure for patients with symptomatic valvular heart mortality, reoperation, and bleeding events in patients
disease [1], accounting for 13% of all cardiac surgery in undergoing surgery for aortic valve replacement by
adults [2], being the most common procedure for all valve mechanical or biological, in a tertiary referral hospital for
surgeries in the United States [3 ] and the second most heart surgery in southern Brazil.
common heart surgery in the UK [4]. In Brazil, according to
a survey of the Department of the Unified Health System METHODS
(DATASUS), implantation of prosthetic valve corresponds
to 17.4% of highly complex heart surgery performed in The study design was observational, retrospective and
January 2008 to August 2010, being the second most cohort.
frequent [5].
More than 30 years after the introduction of modern Sample and sampling
prostheses, the choice between the biological and Nine hundred and thirteen patients, aged 18 years,
mechanical aortic valve remains controversial [6,7]. This is underwent surgery for isolated aortic valve replacement at
because there is no ideal substitute to provide a long life, the Clinics Hospital of Porto Alegre, Rio Grande do Sul in
without the use of oral anticoagulants, with no increased the period from 1 January 1990 to December 31, 2005. The
risk of thromboembolism and operating mechanism similar study excluded cases of aneurysm, and dissection of heart
to the native valve [8-14]. Still, the clinical decision becomes surgery. Of the remaining cases, 301 patients to the cohort
increasingly challenging with the increase in life expectancy through the PEPI software (Programs for Epidemiologists)
and the presence of comorbidities such as advanced age, version 4.0 were randomly selected. With the same software,
congestive heart failure, coronary artery disease, pulmonary it was calculated the sample of sufficient size to detect a
disease and renal failure [6]. magnitude of effect (difference between groups) compared
The choice between the types of prostheses in adults to 15% mortality among the types of prosthesis, maintaining
is determined primarily by assessing the risk of bleeding a statistical power of 80% and a significance of 5%. The
related to anticoagulation, with a mechanical prosthesis magnitude of effect was estimated taking into account the
versus the risk of structural valve deterioration, with a study of Hammermeister et al. [17].
bioprosthesis [15,16]. All surgeries were performed under cardiopulmonary
There are few Brazilian studies comparing biological and bypass with moderate hypothermia (32°) and cardiac arrest,
mechanical prostheses, as well as studies describing the using standard techniques of the Department of
327
Cardiovascular Surgery, Clinics Hospital of Porto Alegre, It is understood by mortality related to the prosthesis
including anesthetic procedures. All prostheses used were the death caused by structural deterioration, nonstructural
a double-leaflet and all biological prostheses implanted were dysfunction, thrombosis, embolism, hemorrhage,
provided by the National Unified Health System After endocarditis, or death related to reoperation of a previously
surgery, all patients were transferred to the ICU for operated valve. Deaths caused by heart failure in patients
postoperative mechanical ventilation in cardiac surgery. with advanced myocardial disease and valvular function
The maximum follow-up was 20 years, averaging 9.2 ± 4.8 without changes are not included.
years and median of 8.9 years. Cardiac death are all the deaths resulting from cardiac
The main objective was to compare mortality among causes, including deaths related to the valves or not. They
individuals with mechanical and biological valve are included in this category deaths from congestive heart
substitutes. The secondary objectives were: 1) to compare failure, acute myocardial infarction and arrhythmia
the probability of reoperation-free time and bleeding events documented, among others.
between groups, 2) to assess predictors of death, and Sudden death is considered the unexplained and
reoperation for bleeding events. unexpected deaths of unknown cause. Its relationship with
The clinical and surgical aspects of the treatment during the valve operated is also unknown. Item reported as a
the study period were completed from information in the separate category of valve-related mortality if the cause
written records of these patients. The data were evaluated can not be determined by clinical or autopsy.
by at least two authors independently. For quality control Hemorrhagic event is defined as any episode of internal
of the team’s performance, 10% of the protocols were or external bleeding that causes higher mortality,
randomly selected to be reviewed by the main investigator. hospitalization, permanent injury such as stroke or loss of
The methodology of this study was based on the vision, or even the need for blood transfusion.
STROBE guidelines (Strengthening the Reporting of
Observational Studies in Epidemiology) [18]. Statistical analysis
The death records were searched in the State Health Quantitative variables were described by mean and
Secretariat of Rio Grande do Sul, Center for Health standard deviation in cases of symmetrical distribution, or
Information - NIS, in Porto Alegre. median and interquartile range in case of skewed
Complications related to the prosthesis were recorded distribution, and qualitatives through absolute and relative
in accordance with the Guidelines for Reporting Mortality frequencies. The comparison between groups was
and Morbidity After Cardiac Valve Interventions [19.20]. performed by Student’s t test for independent samples
(symmetrical distribution) or Mann-Whitney (asymmetric
Ethical aspects distribution) in the case of quantitative variables and chi-
With respect to privacy and confidentiality, the squared or Fisher’s exact test for qualitative variables (rates
anonymity of patients and the use of data obtained in the and proportions).
survey only for the purpose of the project were guaranteed. To assess the survival time, the probability of
The research project received approval from the reoperation for bleeding events we used the Kaplan-Meier
Commission on Ethics in Health Research at the Clinics curve. We applied the chi-square log-rank test to compare
Hospital of Porto Alegre, registered under n° 08-147, to curves between groups.
obtain permission to perform the study in that hospital, To control confounding factors, we used the
with financial assistance from the Incentive Fund to proportional hazards model of Cox As a measure of effect,
Research and Events (FIPE/HCPA). we calculated the ratio of incidences (HR) with their
respective ranges, with 95% confidence. The criterion for
Definitions entering the variable in the model was to produce a P value
less than 0.20 in the bivariate analysis, except for the type
The definitions listed below were all obtained in the of prosthesis that was considered in all models because it
Guidelines for Reporting Mortality and Morbidity After was the main factor under study.
Cardiac Valve Interventions [19.20]. The level of significance was 5% and data were
Total deaths represents all deaths resulting from any analyzed with SPSS (Statistical Package for the Social
cause in patients undergoing aortic valve surgery. Sciences) version 17.0.
Perioperative mortality is defined as any death within
30 days after surgery, regardless of geographic location of RESULTS
the patient.
Hospital mortality is death during the hospital stay after Of the selected patients, 158 (52.5%) underwent
surgery. implantation of mechanical prostheses [St Jude (n=117),
328
Carbomedics (n=25) and Sorin (n=16)] and 143 (47.5%) underwent implantation of a bioprosthesis were
implantation of porcine prostheses [Biocor (n=70), Flumen hospitalized longer (P <0.001), presented longer stay in
(n=55), Bioval (n=14) and Braile Biomédica (n=4)], p = 0.387. the intensive care unit (P = 0.001), total mechanical
The characteristics of patients enrolled in the study are ventilation time significantly higher (P <0.001 ) and a larger
listed in Table 1. number of cases of pneumonia (P = 0.045), as shown in
Considering the hospital outcomes, patients who Table 2.
Table 1. Characteristics of the sample

Variables Sample Mechanical prosthesis Biologal prosthesis P
(n=301) (n=158) (n=143)
Mean age±DP 61.4±12.9 58.0±12.9 65.1±11.9 <0.001
Age range - n (%)
<50 years 60 (19.9) 43 (27.2)(3) 17 (11.9)
51 - 60 years 66 (21.9) 38 (24.1) 28 (19.6) <0.001
61 - 70 years 97 (32.2) 50 (31.6) 47 (32.9)
>71 years 78 (25.9) 27 (19.1) 51 (35.7)(3)
Gender - n (%)
Male 183 (60.8) 88 (55.7) 95 (66.4) 0.074
Female 118 (39.2) 70 (44.3) 48 (33.6)
BMI (kg/m²) - Mean±SD 25.8±4.4 26.5±4.5 25.0±4.1 0.005
(1)
Obesity - n (%) 48 (15.9) 32 (20.3) 16 (11.2) 0.047
Morbid obesity (2) - n (%) 6 (2.0) 4 (2.5) 2 (1.4) 0.687
Functional Class (NYHA) - n (%)
I-II 155 (51.5) 88 (55.7) 67 (46.9) 0.156
III-IV 146 (48.5) 70 (44.3) 76 (53.1)
Chronic atrial fibrillation - n (%) 23 (7.6) 12 (7.6) 11 (7.7) 1.000
Diabetes mellitus - n (%) 35 (11.6) 20 (12.7) 15 (10.5) 0.685
COPD - n (%) 108 (35.9) 56 (35.4) 52 (36.4) 0.963
Stroke- n (%) 11 (3.7) 5 (3.2) 6 (4.2) 0.866
SAH - n (%) 244 (81.1) 122 (77.2) 122 (85.3) 0.100
MI - n (%) 14 (4.7) 8 (5.1) 6 (4.2) 0.934
Creatinine>2 mg/dL - n (%) 7 (2.3) 2 (1.3) 5 (3.5) 0.263
COPD - n (%) 1 (0.3) 0 (0.0) 1 (0.7) 0.475
Emergency surgery - n (%) 4 (1.3) 2 (1.3) 2 (1.4) 1.000
Endocarditis - n (%) 12 (4.0) 6 (3.8) 6 (4.2) 1.000
Rheumatic fever - n (%) 104 (34.6) 55 (34.8) 49 (34.3) 1.000
Pathology
Failure 62 (20.6) 32 (20.3) 30 (21.0)
Stenosis 164 (54.5) 85 (53.8) 79 (55.2) 0.605
DL - predominant stenosis 62 (20.6) 36 (22.8) 26 (18.2)
DL – predominant failure 13 (4.3) 5 (3.2) 8 (5.6)
(1)
BMI>30 Kg/m² e (2) BMI>40 kg/m², according I Brazilian Guideline for Diagnosis and Treatment of Metabolic Syndrome [21].
(3)
Statistically significant association by adjusted residual test (P>0,05).
SD = standard deviation, BMI = body mass index, NYHA = New York Heart Association Class, COPD = chronic obstructive pulmonary
disease, SAH = hypertension, AMI = acute myocardial infarction; DL = double lesion
329
Table 2. Hospital outcomes

(n=301) (n=158) (n=143)
CPB time (min) - mean±SD 72.2±24.2 70.6±23.2 74.0±25.1 0.218
Ischemia time (min) - mean±SD 54.8±18.1 53.4±18.5 56.3±17.7 0.166
Length of hospital stay (days) - median (P25-P75) 13 (10-21) 12(10-18.3) 15(11-23) <0.001
Time in ICU (days) - median (P25-P75) 3.1 (2.8-4.2) 3.0 (2.7-3.9) 3.5 (2.9-5.0) 0.001
PO hospitalization time (days) - median (P25-P75) 9 (8-12) 9 (8-12) 10 (8-14) 0.064
Mechanical ventilation time (h) - median (P25-P75) 14.6 (10.2-19.3) 13.3 (8.9-16.5) 15.9 (13.5-20.9) <0.001
Mechanical ventilation >5 days - n (%) 7 (2.3) 3 (1.9) 4 (2.8) 0.712
AMI - n (%) 1 (0.3) 1 (0.6) 0 (0.0) 1.000
Stroke - n (%) 11 (3.7) 6 (3.8) 5 (3.5) 1.000
Pneumonia - n (%) 43 (14.3) 16 (10.1) 27 (18.9) 0.045
Arrhythmias requiring cardioversion/defibrillation - n (%) 10 (3.3) 7 (4.4) 3 (2.1) 0.342
Dialysis - n (%) 4 (1.3) 3 (1.9) 1 (0.7) 0.624
Reoperation for bleeding - n (%) 12 (4.0) 4 (2.5) 8 (5.6) 0.289
Tamponade - n (%) 2 (0.7) 0 (0.0) 2 (1.4) 0.225
Permanent TAV- n (%) 4 (1.3) 1 (0.6) 3 (2.1) 0.349
CPB= cardiopulmonary bypass; SD = standard deviation; AMI = acute myocardial infarction; TAV=total atrioventricular block
(CI 95% = 62.20% -78.60%) and 58.40% (CI 95% = 48.40% -

Survival data 68.40%), respectively. There was no statistically significant
Figure 1 shows the long-term survival of patients in the difference in survival of patients in both groups (P = .939)
study. Survival at 5, 10 and 15 years after surgical valve throughout follow-up.
replacement by a mechanic substitute was 83.90% (CI 95% Using the multivariate Cox regression, the type of
= 78.00% -89.80%) 75.40% (95% CI = 68, 04% -82.80%) and prosthesis remained with no association with death (P =
60.20% (CI 95% = 45.90% -74.50%), and by biological 0.556), as shown in Table 3. The factors that remained
substitute, was 89.30% (CI 95% = 84.20% - 94.40%) 70.40% statistically associated with death were: age over 70 years,
Table 3. Independent predictors of death by Cox regression

analysis
Variables HR (IC 95%) P
Age> 70 years 2.48 (1.51-4.08) <0.001

Aortic valve insufficiency 2.68 (1.61-4.46) <0.001
COPD 1.97 (1.26-3.08) 0.003
Bleeding events 3.67 (1.57-8.57) 0.003
Arrhythmias with cardioversion
and/or defibrillation in the ICU 3.06 (1.13-8.28) 0.027
Obesity 1.95 (1.02-3.73) 0.043
Chronic atrial fibrillation 1.79 (0.84-3.80) 0.129
Embolic events 2.41 (0.72-8.12) 0.156
Diabetes mellitus 1.47 (0.79-2.74) 0.220
SAH 1.42 (0.77-2.65) 0.264
Creatinine>2 mg/dL 1.71 (0.61-4.80) 0.310
Fig. 1 - Kaplan-Meier to assess probability of survival Type of prosthesis (biological) 0.87 (0.54-1.40) 0.556
Postoperative stroke 0.76 (0.20-2.88) 0.685
Preoperative stroke 0.89 (0.28-2.80) 0.837
CHF class III and IV 1.05 (0.66-1.67) 0.843
CHF = congestive heart failure, COPD = chronic obstructive
pulmonary disease, SAH = hypertension; ICU = intensive care
unit
330
Table 4. Incidence of deaths by period

(n=301) (n=158) (n=143)
n (%) n (%) n (%)
Total deaths 88 (29.2) 40 (25.3) 48 (33,6) 0,149
Perioperative 34 (11.3) 25 (15.8) 9 (6,3) 0,015
Hospitalar(1) 21 (7.0) 15 (9.5) 6 (4,2) 0,115
ICU(1) 8 (2.7) 5 (3.2) 3 (2,1) 0,726
Other 54 (17.9) 15 (9.5) 39 (27,3) <0,001
ICU = Intensive Care Unit.
(1)
None exceeded the perioperative period
obesity, chronic obstructive pulmonary disease, Regarding the causes of deaths, patients who
arrhythmias requiring cardioversion and/or defibrillation underwent implantation of mechanical substitutes were
in the intensive care unit, bleeding events and aortic valve more likely to prosthesis-related death (P = 0.07), which
insufficiency. can be seen in Table 5.
The incidence of deaths by period is shown in Table 4. As shown in Table 6, during the follow-up period, the
Considering all the perioperative period, the group with incidence of reoperation was higher in patients with
mechanical replacement had higher mortality than the group biological valve replacement (P=0.021). Major hemorrhagic
with implanted bioprostheses (P = 0.015). In the remaining of events tended to be more frequent in patients with
the follow-up period, mortality for patients with bioprostheses mechanical replacement (P = 0.084).
was higher than those with mechanical prostheses (P <0.001). Figure 2 shows the cumulative probability of remaining
free of reoperation in these patients during the follow-up.
This probability in 5, 10 and 15 years after surgical valve
replacement by a mechanic substitute was 97.90% (CI 95%
= 95.50% -100.00%) 95.80% (95% CI = 92, 10% -99.50%) and
95.80% (CI 95% = 92.01% -99.50%), and by biological
substitute, was 94.60% (CI 95% = 90.70% - 98.50%) 91.00%
(CI 95% = 85.90% -96.10%) and 83.30% (CI 95% = 74.70% -
91.90%), respectively. Patients with biological substitute
tended to be more likely to have another surgery, especially
after the first 10 years of follow-up (P=0.057).
Using the multivariate Cox regression, patients with
bioprosthetic remained with a greater tendency for
reoperation (P=0.093), as shown in Table 7. The factors
that remained statistically associated with reoperation were:
serum creatinine levels above 2 mg/dL, prosthetic
endocarditis and patients older than 70 years.
Figure 3 shows the cumulative probability of remaining
free of bleeding events. This probability in 5, 10 and 15
Fig. 2 - Kaplan-Meier to assess time freedom from reoperation years after surgical valve replacement by a mechanical
substitute was 94.50% (CI 95% = 90.80% -98.20%) 91.70%
Table 5. Causes of death

Causes of death Sample Mechanical prosthesis Biologal prosthesis P
(n=88(1)) (n=40) (n=48)
n (%) n (%) n (%)
Cardiac 57 (64.8) 27 (67.5) 30 (62.5) 0.791
Related to the prosthesis 20 (22.7) 13 (32.5) 7 (14.6) 0.093
Sudden or unexplained 5 (5.7) 3 (7.5) 2 (4.2) 0.834
Noncardiac 31 (35.2) 13 (32.5) 18 (37.5) 0.791
(1)
Deaths equivalent to 29.2% of the total sample
331
Table 6. Outcomes in the cohort during the follow-up period

(n=301) n (%) (n=158) n (%) (n=143) n (%)
Reoperation for valve replacement 20 (6.6) 5 (3.2) 15 (10.5) 0.021
Bleeding events (1) 14 (4.7) 11 (7.0) 3 (2.1) 0.084
Embolic events (1) 17 (5.6) 9 (5.7) 8 (5.6) 1.000
Endocarditis 8 (2.7) 3 (1.9) 5 (3.5) 0.484
Stroke 14 (4.7) 7 (4.4) 7 (4.9) 1.000
Hemorrhagic 3 (1.0) 2 (1.3) 1 (0.7) 0.803
Embolic 11 (3.7) 5 (3.1) 6 (4.2) 0.547
(1)
Including Stroke
(95% CI = 86, 80% -96.60%) and 91.70% (CI 95% = 86.80% -

96.60%), and by biological substitute was 98.60% (CI 95%
= 96.60% - 100.00%) 97.80% (CI 95% = 95.30% -100.00%)
and 97.80% (95% CI = 95.3% -100.00%), respectively. There
was a greater likelihood of patients who underwent
implantation of biological substitutes to remain free of
bleeding events (P=0.047).
Using the multivariate Cox regression, according to
Table 8, the factors that remained statistically associated
with hemorrhagic events were: serum creatinine levels
above 2 mg/dL and mechanical prostheses.
DISCUSSION
Fig. 3 - Kaplan-Meier time to assess event-free bleeding Mortality

In this study, there was no difference in survival in
both groups, considering a follow-up period of up to 20
years (P=0.939). This fact is due possibly to increased risk
Table 7. Independent predictors of reoperation by Cox regression of bleeding in patients who received a mechanical
analysis
prosthesis be offset in part by the increased risk of
Postoperative endocarditis 199.20 (30.70-1291.00) <0.001 reoperation in those with biological prostheses.
Age>70 years 0.05 (0.01-0.58) 0.016 The mortality observed in this cohort was 29.2% and
Creatinine>2 mg/dL 9.11 (1.06-78.40) 0.044 not statistically significant when comparing the differences
Type of prosthesis (biological) 2.59 (0.85-7.88) 0.093 between the groups receiving biological and mechanical
Preoperative stroke 0.25 (0.02-2.27) 0.249 prostheses (P=0.149), results similar to those found in a
Aortic failure 1.68 (0.57-5.00) 0.348 cohort of 816 patients (24.9% in 25 years) [22]. This is
HR=Hazard Ratio possibly due to the fact that over half of the patients were
older than 60 years and the presence of comorbidities such
as obesity and chronic obstructive pulmonary disease,
Table 8. Independent predictors of bleeding events by Cox which were predictive of death in this sample. Chronic
regression analysis
obstructive pulmonary disease was an independent
Creatinine>2 mg/dL 33.30 (5.50-199.00) <0.001 predictor of death (P <0.05) also in the historical cohort
Type of prosthesis (mechanical) 5.52 (1.40-21.80) 0.015 studied by Bose et al. [23], with 68 patients older than 80
Preoperative MI 4.42 (0.92-21.20) 0.063 years who underwent aortic valve replacement between
Pneumonia in ICU 2.66 (0.74-9.56) 0.134 April 2001 and April 2004, with a mean of 712 days.
Morbid obesity 4.68 (0.50-44.30) 0.178 Additionally, one should not forget that only 17% of the
Diabetes mellitus 2.16 (0.56-8.40) 0.265 deaths were related to the prosthesis in this study, 11.3%
Mechanical ventilation> 5 days 2.25 (0.21-24.00) 0.503 related to mechanical prosthesis and 5.7%, biological
HR=Hazard Ratio; Stroke = stroke, MI = myocardial infarction; prosthesis (P = 0.070). Similar data were found in the
ICU = Intensive Care Unit. mortality study by Stassano et al. [24], observed in 27.74%.
332
Of these, 6.7% and 8.1% are related to mechanical and expectancy of less than 10 years. In this study, as well as
biological prostheses, respectively (P=0.80). pulmonary disease and age over 70 years, also obesity,
Hammermeister et al. [17] found an even higher aortic insufficiency, bleeding events and arrhythmias were
percentage of deaths, as follows, 66 ± 3% and 79 ± 3% for statistically significant predictors of death.
patients with mechanical and biological prostheses, Edwards et al. [28] used data from The Society of
respectively (P = 0.02). This is a prospective, randomized Thoracic Surgeons National Database to identify, in a study
clinical trial comparing mechanical and porcine prostheses of prevalence, risk factors associated with surgical valve
in 394 patients, with 18 years follow-up in 13 medical centers replacement, including 32,968 patients operated between
in the United States, operated between 1977 and 1982. In January 1994 and December 1997, with prevalence of
this study, 37% of the deaths were related to mechanical mortality of 4%. Age was the only risk factor significantly
prosthesis and 41% to the bioprosthesis. This may be related to the type of prosthesis. The factors most strongly
because many deaths related to bioprostheses occurred in associated with mortality were the procedures performed
10 to 15 years after surgery, and can be attributed to primary on an emergency, the need for reoperation and renal failure,
graft dysfunction, with or without reoperation. It is likely not confirmed in this study as predictors of death.
that the high mortality rate recorded in the study is a result Butchart et al. [29] followed a cohort study where they
of the implants performed in the late 1970s and early 1980s, collected 82,297 blood samples for obtaining the
when the technology of valve prostheses and surgical international normalized ratio (INR) of 1,476 patients who
techniques and myocardial protection were still poorly underwent surgery for valve replacement by mechanical
evolved [25]. prostheses between 1979 and 1994 and were followed up
Another important clinical trial was performed in the until 1998, noting that the high anticoagulation variability
United Kingdom in Edinburgh [26], comparing the is the most important independent predictor of survival.
evolution of 211 randomized patients undergoing aortic The variability of anticoagulation was expressed for each
valve replacement between 1975 and 1979 to receive patient, the percentage of INR values †outside the limits
mechanical or porcine prostheses. The results showed a between 2.0 and 4.0. The incidence of deaths related to the
advantage regarding survival in 12 years of follow-up for prosthesis was significantly higher in patients with high
the group with mechanical prosthesis, but this advantage variability of anticoagulation control (changes greater than
disappeared with 20 years of follow-up (P=0.39). Survival or equal to 30%) compared to those who had low variability
at 10 and 20 years after valvar replacement surgery by of the intermediate control (variations between 0 and 29%,
mechanical substitute was 64.0% and 28.4%, and by 9%), showing a linear rate of 1.4% versus 0.5% deaths per
biological substitute was 65.7% and 31.3%, respectively, year (P <0.001). In this study, there was tendency for a
showing no statistical significance (P=0.57). These data higher number of deaths related to the prosthesis for
corroborate the results of this cohort, although they are patients who underwent implantation of mechanical
proportionally lower. prostheses compared to patients undergoing implantation
Also, Kulik et al. [27] found a survival curve similar to of bioprostheses (P=0.070). It is possible that this is due to
the present study when evaluating a cohort of 423 patients, high variability of anticoagulation control, which would
aged between 50 and 70 years who underwent aortic valve require further studies for confirmation.
replacement between January 1977 and July 2002, with a In relation to hospital outcomes, comparing patients
mean of 4.9 ± 3.9 years and a maximum of 15.8 years. The who underwent surgery for implantation of mechanical and
survival at 5, 10 and 15 years after surgical valve replacement biological prostheses, patients in the second group spent
by a mechanical substitute was 89.0 ± 2.1%, 73.2% ± 4.2 longer time on mechanical ventilation (P<0.001), probably
and 65.3 ± 6.0% and by biological substitute was 87.6 ± because older age (P <0.001). The combination of these
5.7%, 75.1 ± 12.6% and 37.5 ± 27.3%, respectively, with no two factors may have resulted in higher incidence of
statistically significant difference between groups (P=0.55). pneumonia in these patients (P = 0.045). Thus, they had
Peterseim et al. [6] performed a retrospective analysis longer hospital stays in the intensive care unit (P=0.001)
of a cohort of 841 patients operated from 1976 to 1996, and, consequently, longer hospital stay (P <0.001). This,
comparing outcomes in patients undergoing aortic valve therefore, did not increase mortality during the perioperative
replacement with porcine and mechanical prostheses. In 10 period in this group. Also, in study published by Florath et
years after surgery, survival free of health problems related al. [30], assessing determinants of mortality at 30 days
to the prosthesis was higher in patients with mechanical postoperatively in a cohort of 2198 patients operated on
substitute and age below 65 years and in patients with between 1996 and 2003, infection was not found as a
biological substitute and older than 65 years. Patients with predictor of increased mortality.
lung disease, kidney disease, ejection fraction less than Tjang et al. [3] performed a systematic review of 28
40%, coronary disease and age over 65 years had a life original articles published between 1985 and 2005 that
333
contained follow-up of patients undergoing surgery for especially by low social, economic and cultural levels, or
aortic valve replacement to identify predictors of mortality. those of difficult clinical management and also due to
There was strong evidence that the risk of early mortality uncertainties about the true intensity of anticoagulation.
was increased in cases of emergency surgery, while the These uncertainties are due to the fact that measures of the
risk of late mortality was increased in older patients with system depend on the INR calibration of thromboplastin
preoperative atrial fibrillation. It was noted also moderate reagents, tissue factors whose contents vary from one
evidence that the risk of early mortality was increased by commercial product to another. Moreover, although different
advanced age, aortic insufficiency, coronary artery disease, thromboplastin reagents produce very similar results with
long cardiopulmonary bypass, left ventricular dysfunction, normal blood, they can produce very different prothrombin
endocarditis, hypertension, mechanical prosthesis, times with anticoagulated blood [29]. According Campos
preoperative pacemaker, dialysis-dependent renal failure et al. [33], only about a third of patients have adequate
and the diameter of the valve. Since the risk of late anticoagulation level in more than half of the follow-up
mortality was increased by emergency surgery and urgency visits, and the residence time within the desired range of
of the operation. All these predictors were considered in anticoagulation is directly related to the occurrence of
the model of this cohort, but only age above 70 years and complications.
aortic valve insufficiency agreed with this study as As the study of Peterseim et al. [6], whose sole predictor
predictors of mortality. of bleeding events was a mechanical prosthesis (P=0.003),
In the Mayo Clinic [31], it was studied a historical cohort the study by Hammermeister et al. [19] also showed a higher
of 440 patients undergoing aortic valve replacement incidence of bleeding events for the group with mechanical
between January 1991 and December 2000, half of whom valves compared to the group of patients with
received mechanical substitute and the other half received bioprostheses (51 ± 4% versus 30 ± 4%, P=0.0001). The
a biological substitute, with a mean follow up of 9.1 year for linearized rate of bleeding events was significantly lower
the first group and 6.2 years for the second group. The for patients who underwent implantation of bioprostheses
survival at 5 and 10 years was 87% and 68% for patients in comparison to patients undergoing implantation of
with mechanical prosthesis and 72% and 50% for patients mechanical prostheses (0.3 ± 0.1% per year versus 1.2 ±
with bioprostheses (P <0.001), respectively, in contrast to 0.3% per year; P=0.001). Oxenham et al. [26] also observed
the present study, where there were statistically different a higher incidence of bleeding events in patients receiving
in relation to survival for these two groups over time. Also, mechanical substitutes, being 2.0% to 2.5% per year with a
perioperative mortality was observed as statistically higher mechanical prosthesis and 0.9% to 2.0% per year with a
for the group of patients with bioprostheses (P=0.04), which porcine prosthesis (P=0.001).
was not confirmed in this study. The cohort studied by Kulik et al. [27] showed no
In a prospective randomized trial [24] performed on two differences from bleeding events among patients with
Italian centers, 310 patients underwent aortic valve biological and mechanical substitutes (P=0.74), as well as
replacement between January 1995 and June 2003, aged 55 the trial by Stassano et al. [24] (P = 0.08). This last attributed
and 70 years, comparing events with mechanical or this result to the possibility of low-intensity anticoagulation
biological prostheses, there were also no differences in for patients with mechanical prostheses in the sample and/
mortality between the groups at 5, 10 and 13 years of follow- or the possibility of patients with biological prostheses
up (P = 0.20), and in this cohort. In this Italian study, have received anticoagulation during follow-up. In contrast
functional class according to the New York Heart to these findings, Brown et al. [31] found a statistically
Association was an independent predictor of mortality significant difference between groups in these two types
(P=0.01), which was not observed in this cohort. of prostheses for bleeding events, occurred in 15% of
patients with mechanical prostheses and 7% of patients
Bleeding events with bioprostheses (P=0.01), although 19% of the latter
The predictors of bleeding events in this study, were receiving warfarin sodium.
statistically significant, were mechanical prostheses In our sample, 2.3% of cases had renal failure and only
(P=0.015) and serum creatinine levels above 2 mg/dL (P one patient underwent dialysis prior to surgery and
<0.001). therefore was not considered in the model. Umezu et al.
As stated by Geldorp et al. [32], patients with a [34] studied a cohort of 63 dialysis patients undergoing
mechanical prosthesis require anticoagulation throughout surgery for valve replacement in January 1990 to July 2007,
their life and the risk of bleeding events increases with at The Heart Institute of Japan, with a mean of 49 months,
advancing age, as observed in this study during the follow- and found the presence of bleeding events in 29.7 % of
up. This is often due to excessive levels of anticoagulation cases, which was much higher than found in this sample
in patients who are not subject to adequate control, (4.7%). They also observed a higher incidence of bleeding
334
events in patients with mechanical valve substitutes in Limitations of the Study

comparison to the biological. Still, a systematic review, also It is a retrospective study, performed at a single center
held in Japan [35], confirms the presence of bleeding and with insufficient sample for rare events.
complications for patients using anticoagulants in the
presence of dialysis. In addition, it is stated that the CONCLUSIONS
mechanical prosthesis seems to be the predominant choice
for hemodialysis patients in that country, because of their Based on the findings of this cohort, it is concluded
high life expectancy and because the studies did not show that:
differences in long-term follow-up when comparing 1) The mortality rate was statistically similar between
mechanical and biological prostheses. groups;
2) The baseline characteristics of patients are the most
Reoperation important determinants of late mortality after surgery;
It can be observed in most of existing publications that 3) There was a tendency to group with reoperation for
the risk of reoperation begins to grow after 10 years of bioprosthesis, especially after 10 years of follow-up;
surgery to implant of valve substitute, probably due to 4) Patients with mechanical prostheses presented more
dysfunction of the prosthesis, and increases progressively bleeding events over time, especially after 5 years of
over time, decreasing with advancing age [17,24,26,32,36- follow-up;
42]. In this cohort, a trend was observed for reoperation 5) The data in this study are consistent with the current
after 10 years of follow-up (P=0.057), which is probably at literature.
the borderline sample descriptive level.
The study by Peterseim et al. [6] showed no significant
difference for patients older than 65 years who received
bioprostheses compared to the group that received a
mechanical prosthesis (P=0.4), and, according to Cox
regression analysis, the use of bioprosthesis (P=0.01) and
the age of 65 years (P=0.0001) were the only variables
predictive of reoperation. In this cohort, only endocarditis
(P<0.001) and serum creatinine levels above 2 mg/dL REFERENCES
(P=0.044) were significantly associated with reoperation.
1. Braunwald E. Aortic valve replacement: an update at the turn
Hammermeister et al. [17] found a higher incidence of of the millennium. Eur Heart J. 2000;21(13):1032-3.
reoperation for patients with biological substitutes
compared with mechanical (29 ± 5% versus 10 ± 3%, P= 2. Cohen G, David TE, Ivanov J, Armstrong S, Feindel CM. The
0.004). Additionally, as the trial by Oxenham et al. [26], the impact of age, coronary artery disease, and cardiac comorbidity
risk of reoperation was significantly higher after 12 years on late survival after bioprosthetic aortic valve replacement. J
for all patients who received a porcine prosthesis (11.3 ± Thorac Cardiovasc Surg 1999;117(2):273-84.
3.6% for porcine versus 4.2 ± 2.1% for mechanical, P <0
0001). Stassano et al. [24] also observed that reoperation 3. Tjang YS, van Hees Y, Körfer R, Grobbee DE, van der Heijden
GJ. Predictors of mortality after aortic valve replacement. Eur
was more frequent in the group that underwent implantation
J Cardiothorac Surg. 2007;32(3):469-74.
of a bioprosthesis, in agreement with other studies
presented (P = 0.003). 4. Kuduvalli M, Grayson AD, Au J, Grotte G, Bridgewater B,
Ruel et al. [40] studied a cohort of 2348 patients Fabri BM; North West Quality Improvement Programme in
undergoing surgery for aortic valve replacement between Cardiac Interventions. A multi-centre additive and logistic risk
1970 and 2002, with a maximum follow-up period of 32.4 model for in-hospital mortality following aortic valve
years. The free time of reoperation for patients with replacement. Eur J Cardiothorac Surg. 2007;31(4):607-13.
mechanical replacement was 96.2%, 94.1% and 93.8% for
10, 15 and 20 years after surgery, respectively, similar to the 5. Ministério da Saúde. Departamento de Informática do SUS –
DATASUS. Informações de Saúde. Assistência à Saúde.
data presented in this cohort. As for biological substitutes,
Internações hospitalares. Procedimentos hospitalares por
time free from reoperation was 76.1%, 61.4% and 59.6% (P gestor - a partir de 2008. Acesso em: 30/10/ 2010.
<0.001) in the same periods of follow-up. Advancing age
was a protective factor against reoperation due to structural 6. Peterseim DS, Cen YY, Cheruvu S, Landolfo K, Bashore TM,
dysfunction of the bioprosthesis in the aortic position (HR Lowe JE, et al. Long-term outcome after biologic versus
= 0.97, P ≤ 0.001), attributed to less deterioration of the mechanical aortic valve replacement in 841 patients. J Thorac
prosthesis in older patients. Cardiovasc Surg. 1999;117(5):890-7.
335
7. Lund O, Bland M. Risk-corrected impact of mechanical versus mortality after cardiac valvular operations. J Thorac Cardiovasc
bioprosthetic valves on long-term mortality after aortic valve Surg. 1996;112(3):708-11.
replacement. J Thorac Cardiovasc Surg. 2006;132(1):20-6.
20. Akins CW, Miller DC, Turina MI, Kouchoukos NT,
8. El Oakley R, Kleine P, Bach DS. Choice of prosthetic heart Blackstone EH, Grunkemeier GL, et al. Guidelines for reporting
valve in today’s practice. Circulation. 2008;117(2):253-6. mortality and morbidity after cardiac valve interventions. Ann
Thorac Surg. 2008;85(4):1490-5.
9. Segalote RC, Pomerantzeff PMA, Brandão CMA, Stolf NAG.
Cirurgia de preservação da valva aórtica em idosos com estenose 21. Sociedade Brasileira de Hipertensão; Sociedade Brasileira de
aórtica. Rev Bras Cir Cardiovasc. 2008;23(4):519-23. Cardiologia; Sociedade Brasileira de Endocrinologia e
Metabologia; Sociedade Brasileira de Diabetes; Sociedade
10. Campos NLKL, Andrade RR, Silva MAM. Anticoagulação Brasileira de Estudos da Obesidade. I Diretriz Brasileira de
oral em portadores de próteses valvares cardíacas mecânicas. Diagnóstico e Tratamento da Síndrome Metabólica. Arq Bras
Experiência de dez anos. Rev Bras Cir Cardiovasc. Cardiol. 2005;84(Suppl I):1-28.
2010;25(4):457-65.
22. Svennevig JL, Abdelnoor M, Nitter-Hauge S. Twenty-five-
11. Andrade ING, Moraes Neto FR, Oliveira JPSP, Silva ITC, year experience with the Medtronic-Hall valve prosthesis in
Andrade TG, Moraes CRR. Avaliação do EuroSCORE como aortic position: a follow-up cohort study of 816 consecutive
preditor de mortalidade em cirurgia cardíaca valvar no Instituto patients. Circulation. 2007;116(16):1795-800.
do Coração de Pernambuco. Rev Bras Cir Cardiovasc.
2010;25(1):11-8. 23. Bose AK, Aitchison JD, Dark JH. Aortic valve replacement in
octogenarians. J Cardiothorac Surg. 2007;2:33.
12. De Bacco MW, Sartori AP, Sant’Anna JRM, Santos MF, Prates
PR, Kalil RAK, et al. Fatores de risco para mortalidade 24. Stassano P, Di Tommaso L, Monaco M, Iorio F, Pepino P,
hospitalar no implante de prótese valvar mecânica. Rev Bras Spampinato N, et al. Aortic valve replacement: a prospective
Cir Cardiovasc. 2009;24(3):334-40. randomized evaluation of mechanical versus biological valves
in patients ages 55 to 70 years. J Am Coll Cardiol.
13. Guedes MAV, Pomerantzeff PMA, Brandão CMA, Vieira 2009;54(20):1862-8.
MLC, Leite Filho OA, Silva MF, et al. Plastia valvar mitral
pela técnica do Duplo Teflon: análise do remodelamento 25. Turina MI. Future of heart valve surgery. Eur J Cardiothorac
cardíaco pela ecocardiografia tridimensional. Rev Bras Cir Surg. 2004;26(suppl 1):S8-S13.
Cardiovasc. 2010;25(4):534-42.
26. Oxenham H, Bloomfield P, Wheatley DJ, Lee RJ, Cunningham
14. Feguri GR, Macruz HMS, Bulhões D, Neves A, Castro RM, J, Prescott RJ, et al. Twenty year comparison of a Björk-
Fonseca L, et al. Troca valvar aórtica com diferentes próteses. Shiley mechanical heart valve with porcine bioprostheses.
Existem diferenças nos resultados da fase hospitalar? Rev Bras Heart. 2003;89(7):715-21.
Cir Cardiovasc. 2008;23(4):534-41.
27. Kulik A, Bédard P, Lam BK, Rubens FD, Hendry PJ, Masters
15. Grunkemeier GL, Li HH, Naftel DC, Starr A, Rahimtoola SH. RG, et al. Mechanical versus bioprosthetic valve replacement
Long-term performance of heart valve prostheses. Curr Probl in middle-aged patients. Eur J Cardiothorac Surg.
Cardiol. 2000;25(2):73-154. 2006;30(3):485-91.
16. Rahimtoola SH. Choice of prosthetic heart valve for adult 28. Edwards FH, Peterson ED, Coombs LP, DeLong ER, Jamieson
patients. J Am Coll Cardiol. 2003;41(6):893-904. WR, Shroyer ALW, et al. Prediction of operative mortality
after valve replacement surgery. J Am Coll Cardiol.
17. Hammermeister KE, Sethi GK, Henderson WG, Grover FL, 2001;37(3):885-92.
Oprian C, Rahimtoola SH. Outcomes 15 years after valve
replacement with a mechanical versus a bioprosthetic valve: 29. Butchart EG, Payne N, Li HH, Buchan K, Mandana K,
final report of the Veterans Affairs randomized trial. J Am Coll Grunkemeier GL. Better anticoagulation control improves
Cardiol. 2000;36(4):1152-8. survival after valve replacement. J Thorac Cardiovasc Surg.
2002;123(4):715-23.
18. von Elm E, Altman DG, Egger M, Pocock SJ, Gøtzsche PC,
Vandenbroucke JP; STROBE Initiative. The Strengthening the 30. Florath I, Albert A, Hassanein W, Arnrich B, Rosendahl U, Ennker
Reporting of Observational Studies in Epidemiology IC, et al. Current determinants of 30-day and 3-month mortality
(STROBE) statement: guidelines for reporting of observational in over 2000 aortic valve replacements: impact of routine laboratory
studies. J Clin Epidemiol. 2008;61(4):344-9. parameters. Eur J Cardiothorac Surg. 2006;30(5):716-21.
19. Edmunds LH Jr, Clark RE, Cohn LH, Grunkemeier GL, Miller 31. Brown ML, Schaff HV, Lahr BD, Mullany CJ, Sundt TM,
DC, Weisel RD. Guidelines for reporting morbidity and Dearani JA, et al. Aortic valve replacement in patients aged 50
336
to 70 years: improved outcome with mechanical versus biologic 37. Hammermeister KE, Sethi GK, Henderson WG, Oprian C,
prostheses. J Thorac Cardiovasc Surg. 2008;135(4):878-84. Kim T, Rahimtoola S. A comparison of outcomes in men 11
years after heart-valve replacement with a mechanical valve or
32. van Geldorp MW, Eric Jamieson WR, Kappetein AP, Ye J, bioprosthesis. Veterans Affairs Cooperative Study on Valvular
Fradet GJ, Eijkemans MJ, et al. Patient outcome after aortic Heart Disease. N Engl J Med. 1993;328(18):1289-96.
valve replacement with a mechanical or biological prosthesis:
Weighing lifetime anticoagulant-related event risk against 38. Kassaï B, Gueyffier F, Cucherat M, Boissel JP. Comparison
reoperation risk. J Thorac Cardiovasc Surg. 2009;137(4):881-6. of bioprosthesis and mechanical valves, a meta-analysis of
randomized clinical trials. Cardiovasc Surg. 2000;8(6):477-83.
33. Campos NLKL, Andrade RR, Silva MAM. Anticoagulação
oral em portadores de próteses valvares cardíacas mecânicas. 39. Carrier M, Pellerin M, Perrault LP, Pagé P, Hébert Y, Cartier
Experiência de dez anos. Rev Bras Cir Cardiovasc. R, et al. Aortic valve replacement with mechanical and biologic
2010;25(4):457-65. prostheses in middle-aged patients. Ann Thorac Surg. 2001;71(5
suppl):S253-6.
34. Umezu K, Saito S, Yamazaki K, Kawai A, Kurosawa H. Cardiac
valvular surgery in dialysis patients: comparison of surgical 40. Ruel M, Kulik A, Rubens FD, Bédard P, Masters RG, Pipe
outcome for mechanical versus bioprosthetic valves. Gen AL, et al. Late incidence and determinants of reoperation in
Thorac Cardiovasc Surg. 2009;57(4):197-202. patients with prosthetic heart valves. Eur J Cardiothorac Surg.
2004;25(3):364-70.
35. Tokunaga S, Tominaga R. Current status of the mechanical
valve and bioprosthesis in Japan. J Artif Organs. 41. Hoffmann G, Lutter G, Cremer J. Durability of bioprosthetic
2008;11(2):53-9. cardiac valves. Dtsch Arztebl Int. 2008;105(8):143-8.
36. Hammermeister KE, Henderson WG, Burchfiel CM, Sethi GK, 42. Vrandecic MO, Gontijo Filho B, Silva JAP, Fantini FA,
Souchek J, Oprian C, et al. Comparison of outcome after valve Barbosa JT, São José MC, et al. Estudo multicêntrico dos
replacement with a bioprosthesis versus a mechanical resultados das trocas valvares com o uso da bioprótese Biocor
prosthesis: initial 5 year results of a randomized trial. J Am no Estado de Minas Gerais. Rev Bras Cir Cardiovasc.
Coll Cardiol. 1987;10(4):719-32. 1988;3(3):159-68.
337
ORIGINAL ARTICLE Rev Bras Cir Cardiovasc 2011; 26.1: 69-75
Predictors of mortality in Patients over 70 years-

old Undergoing CABG or valve surgery with
cardiopulmonary bypass
Preditores de mortalidade em pacientes acima de 70 anos na revascularização miocárdica ou troca
valvar com circulação extracorpórea
Alexander John Pessoa Grant ANDERSON1, Francisco Xavier do Rêgo BARROS NETO1, Marcelo de Almeida
COSTA1, Luciano Domingues DANTAS2, Alexandre Ciappina HUEB3, Marcelo Fernandes PRATA4
RBCCV 44205-1248
Abstract death rates: mechanical Ventilation ³ 12 hours (P <0.001),

Objective: To identify risk factors in octogenarians and ICU stay (P = 0.033), re-exploration (P = 0.001), inotropic
septuagenarians submitted to cardiovascular surgery with support> 48 hours (P <0.001), use of blood components (P
cardiopulmonary bypass (CPB). <0.001).
Methods: Per-operative variables of 265 Patients over 70 Conclusion: Overall mortality justifier of the
years of age Were analyzed. 248 (93.6%) Were octogenarians interventions. CPB time 75 minutes Greater Than,
and 17 (6.4%) nonagenarians. mechanical Ventilation over 12 hours, length of ICU stay,
Results: Overall mortality did not differ Between the need for reoperation, inotropic drug support over 48 hours,
groups, nor did the type of procedure (CABG or valvular) (P and use of blood components are Associated with the higher
= 0545). Pre-operative variables did not Increase the death mortality rate.
risk, nor did the use of arterial grafts or venous (P = 0261),
or the number of grafts per patient (P = 0131). CPB and Descriptors: Aged. Cardiac Surgical Procedures.
cross-clamp time are Associated with higher mortality. The Myocardial Revascularization. Heart Valves. Risk Factors.
group Survivors Had An average CPB time of 70 ± 27 minutes
while the non-survivors group 88.8 ± 25.4 minutes (P
<0.001). Cross-clamp time in the survivors was 55.5 ± 20 Resumo
minutes, while 64.9 ± 16 minutes in the non-survivors (P = Objetivo: Identificar fatores de risco em septuagenários
0.014). Using multivariate logistic regression, CPB time is e octogenários submetidos à cirurgia cardiovascular com
Associated with death (Pearson’s c 2 = 0.0056). CPB time circulação extracorpórea (CEC).
over 75 minutes presents an Increased risk of death of 3.2 Métodos: Avaliadas variáveis peri-operatórias de 265
times (at 95% CI: 1.3-7.9) over Those With CPB time <75 pacientes com mais de 70 anos; desses, 248 (93,6%) eram
minutes. Post-operative variables Associated with Increased septuagenários e 17 (6,4%) eram octogenários.
1. Cardiovascular surgeon (Cardiovascular Surgeon Cardiovascular Work done at Department of Cardiovascular Surgery - Santa Casa de
Surgery Department, Hospital Santa Casa de Misericordia de Misericordia de Limeira.
Limeira)
2. Cardiovascular surgeon (Cardiovascular Surgeon Cardiovascular Correspondence address:
Surgery Department, Hospital Santa Casa de Misericordia de Alexander J. P. G. Anthony Anderson Av Ometto, 675 - Vila Claudia
Limeira) Limeira - SP CEP: 13480-000.
3. Doctor of Medicine (Cardiovascular Surgeon Surgery Division, E-mail: ajpganderson@gmail.com
Heart Institute, Faculty of Medicine, University of São Paulo)
4. Cardiovascular surgeon (Cardiovascular Surgeon, Chief of
Cardiovascular Surgery, Hospital Santa Casa de Misericordia de Article received on June 17th, 2010
Limeira). Article accepted on January 12th, 2011
69
ANDERSON, AJPG ET AL - Predictors of mortality in Patients over Rev Bras Cir Cardiovasc 2011; 26.1: 69-75
70 years-old Undergoing CABG or valve surgery with cardiopulmonary
bypass
Resultados: Não houve diferença de mortalidade entre apresenta 3,2 vezes (IC 95%: 1,3 - 7,9), maior chance de
eles, com mortalidade global de 22 (8,3%) pacientes. Não óbito do que os pacientes com tempo de CEC £ 75 minutos.
houve diferença em relação ao tipo de procedimento Variáveis pós-operatórias: tempo de ventilação mecânica ³
(revascularização ou tratamento valvar) (P=0,545). As 12 horas (P< 0,001), tempo de internação na UTI (P=0,033),
variáveis pré-operatórias não aumentaram o risco de morte. reoperação (P=0,001), suporte inotrópico > 48 horas
Enxerto arterial ou venoso (P=0,261) e número de enxertos (P<0,001) e necessidade de hemoderivados (P<0,001)
utilizados por paciente (P=0,131) não aumentaram a aumentam a mortalidade.
mortalidade. O grupo de sobreviventes apresentou tempo Conclusão: A mortalidade global justifica a intervenção.
médio de CEC de 70 ± 27 minutos e o grupo óbito, 88,8 ± 25,4 CEC > 75 minutos, tempo de ventilação mecânica superior a
minutos, com significância estatística (P<0,001). O tempo 12 horas, de internação em UTI, reoperação, suporte
de isquemia no grupo de sobreviventes foi de 55,5 ± 20 inotrópico por período superior a 48 horas e uso de
minutos e no grupo óbito, 64,9 ± 16 minutos, com hemoderivados estão associados a maior mortalidade.
significância (P=0,014). Na regressão logística multivariada,
o tempo de CEC é a variável que se associa a morte, com qui- Descritores: Idoso. Procedimentos Cirúrgicos Cardíacos.
quadrado de Pearson =0,0056. Tempo de CEC > 75 minutos Revascularização Miocárdica. Valvas Cardíacas. Fatores de Risco.
INTRODUCTION individuals to offer an optimal treatment, which determines

the success of the procedure [8,9].
Population aging is a worldwide phenomenon. Thus, this work aims to identify predictive variables,
According to the 2000 census, there was a 35.6% increase pre, per and postoperative related to increased morbidity
in the number of elderly (people over 65 years of age) in or mortality in patients septuagenarians and octogenarians
Brazil in relation to 1991. Estimates point to the possibility undergoing CABG or valve replacement using
that in the next 20 years, in Brazil, the number of elderly cardiopulmonary bypass.
people exceeding 30 million people, and represent almost
13% of the population [1]. METHODS
Among the causes of death more frequently in the
elderly are heart disease, cancer and cerebrovascular We evaluated 2731 patients undergoing cardiovascular
disease. According to Beaglehole [2], ischemic heart disease procedures in the Department of Cardiovascular Surgery,
is the leading cause of death in industrialized countries, Santa Casa de Limeira, SP, Sao Paulo, from January 1998 to
accounting for 30% of all deaths each year. January 2009, and 265 (9.7% of total operated) were aged
In Brazil, in the early 90’s, cardiovascular diseases more than 70 years. Of these, 248 (93.6%) were
represent the leading cause of death, accounting for septuagenarian and 17 (6.4%) were octogenarians.
approximately 34% of deaths in the country [3]. IBGE figures Over 265 patients were evaluated septuagenarian and
currently show a life expectancy of around 67.6 years. octogenarian predictive variables before, during and after
Demographic data in Brazil and in developed countries surgery, to identify which risk factors associated with
show a clear increase of the elderly population. Estimates increased mortality when undergoing artery bypass graft
place in the sixth as the elderly population in the world in surgery or valve replacement, both using cardiopulmonary
2025 [3]. bypass.
Considering these data and according to several studies
that demonstrate the effectiveness, improved quality of
life and increased survival by means of coronary artery
bypass grafting [4,5], is experiencing an increase in the
indication of surgical treatment of coronary artery disease Chart 1. Criteria for Inclusion
Aged 70 years
and degenerative valve disease in the elderly.
Cardiopulmonary bypass
Advances in surgical technique, types of oxygenators, Coronary artery bypass grafting (CABG) or Combined Coronary
myocardial protection, less invasive surgery, intensive care Artery Bypass (CABG associated) (CABG associated with
management and better physical therapy and post-operative surgical treatment of cardiac valves, CABG associated with aortic
help to reduce morbidity and mortality in the elderly [6-7] aneurysm repair, CABG associated VSD)
and is essential to know what are its characteristics Surgical treatment of cardiac valves (Valve)
70
bypass
Inclusion criteria were age less than 70 years, use of Statistical analysis
cardiopulmonary bypass, cardiovascular revascularization The various parameters studied were compared by
procedure (CABG), CABG procedures combined (CABG analysis of variance with one factor being broken, the
and heart valve surgery, CABG and aortic aneurysm, RM differences between groups by Student t test and the Fisher
and VSD after acute myocardial infarction) and heart valve exact test. For analysis of CPB and ischemia test was used
surgery. Exclusion criteria were minimized by aiming at a nonparametric Mann-Whitney. We used logistic regression
more comprehensive assessment of these patients (Chart univariate and multivariate analysis. The multivariate model
1). It was used as the sole criterion, patients referred to the selected by the process variables stepwise. Data are
operating room in cardiac arrest. presented as mean, maximum and minimum standard
Patients were grouped as follows: patients aged 70 to deviation. We established the significance level of 5%.
79 years (Group septuagenarian) 248 (93.6%) patients and
80 to 89 years (Group Octogenarian) 17 (6.4%). Regarding RESULTS
the type of procedure were grouped into isolated CABG or
combined with another procedure (CABG or CABG The overall mortality was 22 (8.3%) patients. There was
combined), 223 (84.2%) and surgical treatment of cardiac no difference in mortality between octogenarians and
valves (Valve) 42 (15.8%). septuagenarians, 20 deaths occurred among 248 (8.0%)
The preoperative variables were assessed occurrence patients septuagenarian, and two deaths among 17 (11.7%)
of: acute myocardial infarction (AMI), through enzyme octogenarian patients (P = 1), Table II. Regarding the type
elevation with or without electrocardiographic changes, of procedure, it was observed that there was no statistical
diabetes mellitus (DM) with fasting glucose ³ 126 mg / dL, difference when comparing the MR group associated with
hypertension, with systolic pressure level ³ 140 mmHg MR, with 18 deaths among 206 (8.7%) patients in the MR
(hypertension), cerebrovascular accident (CVA), smoking group versus two deaths among 17 (11.7% ) of the patients
(SMOKE), peripheral vascular disease (DPB), characterized associated with MRI, and these two were associated with
by arterial insufficiency with intermittent claudication, and MRI post-infarction VSD. Compared to normal valve, two
prior cardiovascular surgery (Reop). deaths occurred among 42 (4.7%) patients. Thus, there were
In all cases, the cardioplegia was performed using a 20 deaths out of 223 (8.9%) patients in the RM + RM and
solution Buckberg10 amid normothermic blood, with two associated deaths among 42 (4.7%) patients in the valve
intervals of 20 minutes. Table 1 shows the preoperative (P = 0.545).
variables.
Variables obtained during surgery, assessed the
presence and number of arterial and venous grafts, the time
of cardiopulmonary bypass and myocardial ischemia time. Table 2. Distribution of deaths by age and type of procedure.
The postoperative variables were length of stay in
intensive care unit, hospital stay, duration of mechanical Septuagenarian Octogenarians
ventilation, reoperation for bleeding, need for blood (n=248) (n=17)
transfusion, postoperative stroke, postoperative confusion Isolated MR February 15 2
and the need inotropic support for more than 48 hours. Associated MR 3 0
Valve 2 0
MR: Myocardial Revascularization
Table 1. Preoperative variables. N = 265.

Pre-operative variables Occurences Percent Table 3. Statistical analysis of preoperative variables.
(N=265) Pre-operative variables P
MI 53 20% IAM 0,165
DMII 70 26.4% DMII 0,682
SH 170 64.2% HAS 0,381
Stroke 5 1.9% AVC 1
SMOKE 46 17.4% FUMO 0,387
ISQ 2 0.8% ISQ 1
Reop 8 3% Reop 0,136
AMI: acute myocardial infarction; DMII: diabetes mellitus, AMI: acute myocardial infarction; DMII: diabetes mellitus,
hypertension, hypertension, stroke, stroke, TOBACCO: Smoking, hypertension, hypertension, stroke, stroke; SMOKING: smoking;
ISQ: peripheral vascular disease Reop: CABG surgery ISQ: peripheral vascular disease; Reop: CABG surgery.
71
bypass
The preoperative variables analyzed did not increase

the risk of death in this group of patients (RM, RM and
associated Valve). The statistical data are described in
Table 3.
Regarding the type of arterial or venous graft, arterial
grafts were used 95 and 187 vein in groups MR and
associated CABG. We observed an average use of arterial
grafts 0.42 per patient and 1.82 venous grafts per patient.
There was no increase in mortality when comparing the
use of arterial grafts compared to vein grafts (P = 0.261).
There was no difference in mortality when we evaluated
the number of grafts per patient (P = 0.131), which ranged
from 1 to 5 grafts.
Regarding perioperative variables, we found that the
duration of CPB and ischemic times are associated with an
increased mortality. The group of survivors had mean CPB
time 70 ± 27 minutes in the death group and the mean was Figure 3 - ROC curve showing the cut-off for the duration of CPB.
88.8 ± 25.4 minutes with statistical significance (P <0.001). (CPB> 75 minutes 3.2 times more likely death [95% CI: 1.3 to
7.9]). CPB: Cardiopulmonary Bypass, CI: Confidence Interval
The ischemia in survivors was 55.5 ± 20 minutes and in the

deceased group, 64.9 ± 16 minutes with significance (P =
0.014) (Figure 1).
From this univariate analysis, we used multivariate
logistic regression model with duration of CPB and ischemia
to evaluate whether the two variables are important
determinants of mortality. Significance was not observed.
Thus, we used a selection process to identify which one is
more important as a determinant of mortality. Where it was
Figure 1 - Time of CPB-related group of survivors and observed that the CPB is the variable that is associated
nonsurvivors. CPB: Cardiopulmonary Bypass with death, with Pearson c 2 = 0.0056. Figure 2 shows the
graph of logistic regression.
We chose to construct an ROC curve for the cut to the
CPB time. Through it, we can say that patients with CPB
time> 75 minutes have 3.2 times (95% CI: 1.3 - 7.9) greater
chance of death than patients presenting with CPB time £
75 minutes (Fig. 3).
The postoperative variables were analyzed and showed
that the variables length of hospital stay (P = 0.188),
postoperative stroke 22 (8.3%) patients (P = 0.230) and
postoperative confusion 22 ( 8.3%) patients (P = 0.082)
variables are not predictors of mortality.
The postoperative variables that are related to increased
mortality are: duration of mechanical ventilation ³ 12 hours,
which occurred in 13 of 22 (59.0%) patients who died (P
<0.001), length of stay in unit intensive care unit ³ 48 hours,
which occurred in 12 of 22 (54.5%) deaths (P = 0.033),
Figure 2 - Graphic logistic regression showing the probability of reoperation for bleeding in six of 22 (27.2%) deaths (P =
death in relation to CPB. CPB: Cardiopulmonary Bypass 0.001); inotropic support for more than 48 hours in 14 of 22
72
bypass
(63.6%) deaths (P <0.001) and need for blood products more with the daily reality in cardiac surgery. Thus, we used
than six units of packed red cells or fresh frozen plasma, patients undergoing CABG, CABG associated with surgical
which occurred in 13 of 22 (59.0%) deaths (P<0.001), as procedures such as valve replacement, aneurysm of
detailed in Table 4. ascending aorta or ventricular septal defect and patients
Regarding the use of blood products, packed red cells undergoing aortic or mitral valve replacement. Data analysis
was used in 77.7% of patients, fresh frozen plasma in 6.0% and grouping of these patients were performed separately.
of patients and platelet concentrate for 3.3% of cases. The But, analyzing data, there was no difference between them,
group of survivors received an average of 2.7 ± 1.3 units and we opted for presenting the group as such.
and the death group received 6.9 ± 13.5 units of blood We analyzed preoperative variables, which usually are
products. predictors of poor prognosis, as previous MI, or
reoperation. But he noted that these variables did not show
differences in mortality in this group of patients [15].
Table 4. Postoperative variables. In the past, the use of arterial grafts in patients
Variables P undergoing emergency situations or for critically ill or the
Time of Hospitalization 0.188 elderly was associated with a worse prognosis, especially
Postoperative stroke 0.230 in relation to the patency of the graft and wound infection.
Mental confusion Postoperative 0.082 Tyszka & Fucuda [16] observed that the use of internal
Ventil time. Mechanical > 12h 13/22 pts (59.0%) <0.001 thoracic artery brings no increase in morbidity and even
Length of Stay in ICU 12/22 pts (54.5%) 0.033 improves the early and late survival and should therefore
Review of Hemostasis 6/22 pts (27.2%) 0.001 be considered as a graft of first choice for the elderly.
Inotropic support > 48 h 14/22 pts (63.6%) <0.001
When we evaluated the perioperative variables, it was
Use of Blood Products (> 06 U) 13/22 pts (59.0%) <0.001
found that the time of cardiopulmonary bypass and
CVA: cerebrovascular accident, ICU: Intensive Care Unit; pts: myocardial ischemia time, are associated with increased
patients mortality. This information is important because based on
this premise, one can establish an operative strategy that
aims at reducing as far as possible from the time of CPB or
ischemia. It was noted that a CPB time greater than 75
DISCUSSION minutes, increases by 3.2 times the odds of death with a
confidence interval ranging from 1.3 to 7.9. The literature
Increased life expectancy of our population is well shows some parameters related to higher mortality rate in
documented, with an evident increase in the number of elderly patients, but the time of ischemia and CPB presented
individuals aged over 70 years. Thus, the incidence of as variable multivariate logistic presented in this study has
cardiovascular disease also increases, which results in our important [17-19].
midst, in a high incidence of MI prior to surgery [11]. Once Regarding postoperative variables, it was observed that
stated the surgical procedure, the added technology and the duration of mechanical ventilation longer than 12 hours,
care before, during and after the surgery, lead best outcome the time of ICU stay longer than 5 days, reoperations due
for elderly patients who require highly complex to postoperative bleeding, the use of inotropic or
cardiovascular intervention, either to improve the quality vasopressor catecholamines by more than 48 hours and
or life expectancy [12] . the use of blood products, either red blood cells, fresh
It was noted that work that, in one center in São Paulo, frozen plasma or platelet numbers greater than 6 units are
9.7% of patients operated for 11 years were septuagenarian linked to increased mortality. These variables had already
or octogenarian, translating into a significant portion of been demonstrated in other series of cases, and were also
total patients. With this fact in mind, we should prepare for highlighted by us [20-22].
the future as they grow in longevity and life expectancy in Evaluating the aspect of the use of cardiopulmonary
our population [3]. Surgery in patients over 70 years is bypass, Iglézias et al. [23] observed that the procedures
feasible and we achieved similar rates of mortality rates of performed with cardiopulmonary bypass are similar to
patients undergoing cardiovascular procedures of high those reported for patients operated without CPB. The
complexity in the population above 70 years [13,14], incidence of cerebral ischemia and death is the same in
performing all surgeries with cardiopulmonary bypass. groups with or without CPB and not appear as isolated
When the guidelines were established in this project, predictors [20]. In line with our data, Angelini et al. [24]
the initial goal was to add the greatest possible number of reported recently that the long-term results are indifferent
variables and the smallest possible number of exclusion to the use of CPB or not.
criteria. The idea was to bring the most of the data obtained Limiting factor of our work is the retrospective analysis
73
bypass
of conventional and electronic medical records. Possibly, 6. Bapat V, Allen D, Young C, Roxburgh J, Ibrahim M. Survival
some data traditionally found in previously published and quality of life after cardiac surgery complicated by
works, such as preoperative predictors of mortality have prolonged intensive care. J Card Surg. 2005;20(3):212-7.
not been manifested in our study precisely because of this
7. Collart F, Feier H, Kerbaul F, Mouly-Bandini A, Riberi A, Di
limitation.
Stephano E, et al. Primary valular surgery in octogenarians:
perioperative outcome. J Heart Valve Dis. 2005;14(2):238-42.
CONCLUSIONS
8. Scott BH, Seifert FC, Grimson R, Glass PS. Octogenarians
Septuagenarian and octogenarian patients when undergoing coronary artery bypass graft surgery: resource
undergoing cardiovascular procedures with CPB as a utilization, postoperative mortality, and morbidity. J
coronary bypass, associated CABG or valve replacement, Cardiothorac Vasc Anesth. 2005;19(5):583-8.
overall mortality is acceptable. The cardiopulmonary bypass
time greater than 75 minutes increases mortality 3.2 times. 9. Alves Júnior L, Rodrigues AJ, Évora PRB, Basseto S, Scorzoni
Filho A, Luciano PM, et al. Fatores de risco em septuagenários
Postoperative variables such as duration of mechanical
ou mais idosos submetidos à revascularização do miocárdio e
ventilation, length of stay in intensive care unit, reoperation ou operações valvares. Rev Bras Cir Cardiovasc.
for bleeding, prolonged inotropic support and use of blood 2008;23(4):550-5.
products is related to increased postoperative mortality.
10. Buckberg GD, Beyersdorf F, Allen BS, Robertson JM.
Integrated myocardial management: background and initial
application. J Card Surg. 1995;10(1):68-89.
11. Almeida AS, Manfroi WC. Peculiaridades no tratamento da

cardiopatia isquêmica no idoso. Rev Bras Cir Cardiovasc.
2007;22(4):476-83.
12. Takiuti ME, Hueb W, Hiscock SB, Nogueira CR, Girardi P,

Fernandes F, et al. Qualidade de vida após revascularização
cirúrgica do miocárdio, angioplastia ou tratamento clínico. Arq
Bras Cardiol. 2007;88(5):537-44.
13. Edwards FH, Clark RE, Schwartz M. Coronary artery bypass

grafting: the Society of Thoracic Surgeons National Database
experience. Ann Thorac Surg. 1994;57(1):12-9.
14. Almeida RMS, Lima Jr. JD, Martins JF, Loures DRR
Revascularização do miocárdio em pacientes após a oitava
REFERENCES década de vida. Rev Bras Cir Cardiovasc. 2002;17(2):116-22.
1. Fundação IBGE. Informações estatísticas e geocientíficas. 15. Ladeira RT, Jatene FB, Monteiro R, Zucato SP, Baracioli
Disponível em: http://www.ibge.gov.br LM, Hueb AC, et al. Cirurgia de revascularização na fase
aguda do infarto do miocárdio. Análise dos fatores pré-
2. Beaglehole R. International trends in coronary heart disease operatórios preditores de mortalidade. Arq Bras Cardiol.
mortality, morbidity, and risk factors. Epidemiol Rev. 2006;87(3):254-9.
1990;12(1):1-15.
16. Tyszka AL, Fucuda LS. O uso da artéria torácica interna no
3. Ministério da Saúde. Doenças cardiovasculares no Brasil - idoso: indicações e resultados imediatos. Rev Bras Cir
Sistema Único de Saúde – SUS. Brasília: Ministério da Saúde, Cardiovasc. 2001;16(4):269-74.
Coordenação de Doenças Cardiovasculares;1993. 36p.
17. Taniguchi FP, Souza AR, Martins AS. Tempo de circulação
4. Loures DRR, Carvalho RG, Mulinari L, Silva Jr AZ, Schmidlin extracorpórea como fator de risco para insuficiência renal aguda.
CA, Brommelstroet M, et al. Cirurgia cardíaca no idoso. Rev Rev Bras Cir Cardiovasc. 2007;22(2):201-5.
Bras Cir Cardiovasc. 2000;15(1):1-5.
18. Kobayashi KJ, Williams JA, Nwakanma LU, Weiss ES,
5. Deininger MO, Oliveira OG, Guedes MGA, Deininger EG, Gott VL, Baumgartner WA, et al. EuroSCORE predicts
Cavalcanti ACW, Cavalcanti MGFW, et al. Cirurgia de short- and mid-term mortality in combined aortic valve
revascularização do miocárdio no idoso: estudo descritivo de replacement and coronary artery bypass patients. J Card
144 casos. Rev Bras Cir Cardiovasc. 1999;14(2):88-97. Surg. 2009;24(6):637-43.
74
bypass
19. Gopaldas RR, Chu D, Dao TK, Huh J, Lemaire SA, Coselli JS, 22. Engoren M, Habib RH, Hadaway J, Zacharias A, Schwann
et al. Predictors of surgical mortality and discharge status after TA, Riordan CJ, et al. The effect on long-term survival of
coronary artery bypass grafting in patients 80 years and older. erythrocyte transfusion given for cardiac valve operations.
Am J Surg. 2009;198(5):633-8. Ann Thorac Surg. 2009;88(1):95-100.
20. Ricci M, Karamanoukian HL, Dancona G, Bergsland J, Salerno 23. Iglézias JCR, Lourenção Jr. A, Dallan LAO, Puig LB, Oliveira
SA. Revascularização do miocárdio no paciente idoso: com ou
TA. On-pump and off-pump coronary artery bypass grafting
sem circulação extracorpórea? Rev Bras Cir Cardiovasc.
in the elderly: predictors of adverse outcome. J Card Surg.
2003;18(4):321-5.
2001;16(6):458-66.
24. Angelini GD, Culliford L, Smith DK, Hamilton MC, Murphy
21. Busch T, Friedrich M, Sirbu H, Stamm C, Zenker D, Dalichau GJ, Ascione R, et al. Effects of on- and off-pump coronary
H. Coronary artery bypass procedures in septuagenarians are artery surgery on graft patency, survival, and health-related
justified. Short and long-term results. J Cardiovasc Surg quality of life: long-term follow-up of 2 randomized controlled
(Torino). 1999;40(1):83-91. trials. J Thorac Cardiovasc Surg. 2009;137(2):295-303.
75
Original Cardiovascular 35
Comparison between Sequential Organ Failure

Assessment Score (SOFA) and Cardiac Surgery
Score (CASUS) for Mortality Prediction after
Cardiac Surgery
A. M. A. Badreldin 1 F. Doerr 1 M. M. Ismail 1 M. B. Heldwein 1 T. Lehmann 2 O. Bayer 3 T. Doenst 1
K. Hekmat 1
1 Department of Cardiothoracic Surgery, Friedrich Schiller University Address for correspondence and reprint requests Dr. Akmal M. A.
of Jena, Jena, Germany Badreldin, M.D., Ph.D., Department of Cardiothoracic Surgery,
2 Institute of Medical Statistics, Computer Sciences and Friedrich Schiller University of Jena, Erlanger Allee, 07747 Jena,
Documentation, Friedrich Schiller University of Jena, Jena, Germany Germany (e-mail: akmalbadreldin@yahoo.com).
3 Department of Anesthesiology and Intensive Care Medicine,
Friedrich Schiller University of Jena, Jena, Germany
Downloaded by: World Health Organization ( WHO). Copyrighted material.

Thorac Cardiovasc Surg 2012;60:35–42.
Abstract Background Our purpose was to evaluate and compare the accuracy of the “Sequen-
tial Organ Failure Assessment” score (SOFA) and the “Cardiac Surgery Score” (CASUS)
for the prediction of mortality after cardiac surgery.
Methods Between January 1, 2007 and December 31, 2008 we prospectively included
all consecutive adult patients admitted to our intensive care unit (ICU) after cardiac
surgery. Both scoring systems were calculated daily from the 1st day in the ICU (day of
operation) until the 7th ICU day. We evaluated the ICU mortality prediction of both
models using calibration and discrimination statistics.
Results 2801 patients (29.6% females) were included. Mean age was 66.9 10.7
years. Intensive care unit mortality was 5.2%. The calibration of the “Sequential Organ
Failure Assessment Score” and “Cardiac Surgery Score” was reliable for all days (p
Key words 0.05). CASUS was more accurate in predicting survival and mortality compared to SOFA
► outcomes (includes for all days, as evidenced by the larger areas under the Receiver Operating Characteristic
mortality morbidity) curves.
► intensive care Conclusions Both CASUS and SOFA are reliable mortality prediction tools after cardiac
► scoring system surgery. However, CASUS was more accurate in predicting the individual patient's risk of
► cardiac surgery mortality. Thus, use of the CASUS in cardiac surgery intensive care units is
► mortality prediction recommended.
Introduction The “Sequential Organ Failure Assessment” score (SOFA)

(►Table 1) is the only postoperative scoring system that
Although most intensive care-based scoring systems have has been validated in cardiac surgery patients.3,4 It was
excluded cardiac surgery patients,1,2 many of them have concluded that SOFA can be applied to cardiac surgery
been used in these patients due to the lack of specific patients without any further modifications. Moreover, the
postoperative risk stratification tools in cardiac surgery. “SOFA” score had a good accuracy when used in the early
received Copyright © 2012 by Thieme Medical DOI http://dx.doi.org/

August 28, 2010 Publishers, Inc., 333 Seventh Avenue, 10.1055/s-0030-1270943.
accepted after revision New York, NY 10001, USA. ISSN 0171-6425.
January 17, 2011 Tel: +1(212) 584-4662.
published online
April 28, 2011
36 Comparison between SOFA and CASUS Badreldin et al.
Table 1 The Sequential Organ Failure Assessment (SOFA) scoring system.
Organ System Descriptor Score points

1 2 3 4
Respiratory PaO2/FiO2 mmHg 400 300 200, on ventilator 100, on ventilator
Renal creatinine, mg/dL 1.2–1.9 2.0–3.4 3.5–4.9 or <500 >5 or <200
or urine output, mL/d
Liver bilirubin, mg/dL 1.2–1.9 2.0–5.9 6.0–11.9 12.0
Cardiovascular catecholamines, MAP <70 dopamine 5, dopamine >5, epinephrine dopamine >15,
μg/kg/min or MAP dobutamine 0.1, norepi 0.1 epinephrine >0.1,
(any dose) norepi >0.1
Coagulation platelets 103/μL 150 100 50 20
Central nervous Glasgow coma scale 14–13 12–10 9–6 <6
PaO2: partial oxygen pressure; FiO2: fraction of inspired oxygen; MAP: mean arterial blood pressure; norepi: norepinephrine

postoperative period after cardiac surgery. On the other Hospital (approval no.: 2809–05/10). For patients who were
hand, the “Cardiac Surgery Score” (CASUS) (►Table 2) was readmitted to the ICU during the study period, only the first
proposed as a specialized scoring system which would take admission was considered. Preoperative data were collected
account of the pathophysiological characteristics of this using the QUIMS 2.0b quality control system (University
subgroup of patients.5 Accordingly, we aimed in this study Hospital of Muenster, Germany). Postoperative data were
to compare the recently introduced CASUS with the widely collected from the intensive care information system COPRA
used SOFA scoring system, using the well-established, pre- 5.2 (COPRASYSTEM GmbH, Sasbachwalden, Germany), which
operative, additive and logistic EuroSCOREs as reference is linked to the patient monitors (Philips IntelliVue MP70,
models. This comparison was performed using calibration Amsterdam, Netherlands), ventilators (Draeger Evita IV, Lü-
and discrimination analysis. beck, Germany and Hamilton Galileo, Bonaduz, Switzerland),
blood gas analyzing devices (ABL800Flex Radiometer, Copen-
hagen, Denmark) and central laboratories. The attending
Materials and Methods
physician was responsible for data collection during the first
This study evaluated the prospectively collected data of all postoperative week. Two assigned medical clerks validated
consecutive adult cardiac surgery patients admitted to our the data collection on a daily basis. A senior consultant
ICU after cardiac surgery. The study was conducted between performed a second periodical validation. Inconsistencies
January 1, 2007 and December 31, 2008, and was approved by between raters were resolved by consensus. No data were
the Institutional Review Board of Friedrich Schiller University missing. Outcome was defined as ICU mortality.
Table 2 The Cardiac Surgery Score (CASUS).
Organ System Descriptor Score points

0 1 2 3 4
Respiratory PaO2/FiO2 mmHg extubated >250 151–250 75–150 <75
Renal S. creatinine (mg/dL) <1.2 1.2–2.2 2.3–4.0 4.1–5.5 >5.5
CVVH/dialysis no – – – yes
Liver S. bilirubin (mg/dl) <1.2 1.2–3.5 3.6–7.0 7.1–14.0 >14.0
Cardiovascular PAR ¼ HR CVP/MAP <10.1 10.1–15.0 15.1–20.0 20.1–30.0 >30.0
lactic acid (mmol/l) <2.1 2.1–4.0 4.1–8.0 8.1–12.0 >12.0
intraaortic balloon pump no – – – yes
ventricular assist device no – – – yes
3
Coagulation platelets10 /μL >120 81–120 51–80 21–50 <21
Central nervous neurologic state normal – confused sedated diffuse neuropathy
CVVH: continuous venovenous hemofiltration; CVP: central venous pressure; FiO2: fraction of inspired oxygen; HR: heart rate; MAP: mean arterial
blood pressure; PAR: pressure adjusted heart rate; PaO2: partial oxygen pressure; S. bilirubin: serum bilirubin; S. creatinine: serum creatinine
Thoracic and Cardiovascular Surgeon Vol. 60 No. 1/2012

Comparison between SOFA and CASUS Badreldin et al. 37
Table 3 Types of operations in the study population.
Operation Number %
CABG 1526 54.5
Isolated valve surgery 635 22.7
Combined CABG and valve surgery 381 13.6
Ascending aorta and aortic arch surgery 60 2.1
Combined ascending aorta and valve surgery 116 4.1
Combined ascending aorta and coronary surgery 5 0.2
Cardiac transplantation 24 0.9
Congenital, cardiac tumors, pulmonary embolectomy, Assist device implantation 54 1.9
Total 2801 100
CABG, coronary artery bypass grafting.
Statistical analysis indicate the absence of significant discrepancy between

Statistical analysis was performed using the SPSS software, predicted and observed mortality. Calibration (HL test) was

version 18 (SPSS, Inc., Chicago, IL, USA). Graphics were drawn considered good when the χ2 value was low and p was high
up using Microsoft Excel software. Continuous scale data are (not <0.05). Discrimination (the ability of a scoring model to
presented as mean Â± standard deviation (SD). The means of differentiate between survival and mortality) was evaluated
continuous data were compared by two-tailed Student's t- with receiver operating characteristic (ROC) curves. The area
test in subgroups with normal data distribution. In sub- under the curve (AUC) indicates the discriminative ability of
groups with non-normal frequency distribution (large posi- parameters, which in this case meant the ability to discrimi-
tive skewness), nonparametric rank tests were used nate survivors from decedents. An AUC of 0.5 is equivalent to
(Kruskal-Wallis test). A value of p < 0.05 was considered a random chance, while AUC >0.7 indicates a moderate
significant. Calibration tests were performed using the prognostic model and AUC >0.8 indicates a good prognostic
Hosmer-Lemeshow test (goodness-of-fit test, HL test) to model. Overall correct classification (OCC; here, the ratio of
Table 4 Results of calibration, discrimination and overall correct classification for the preoperative EuroSCOREs and for CASUS and
SOFA from day 1 to day 7.
Day Scoring model Correctly predicted HL test ROC analysis

outcome
OCC (%) χ2 p-value AUC 95% CI
Preoperative (2801) Add-Euro 94.7 9.10 0.33 0.71 0.64–0.79
Log-Euro 94.6 19.75 0.01 0.71 0.63–0.78
ICU Day 1 (2801) CASUS 96.0 3.65 0.82 0.93 0.91–0.95
SOFA 95.3 7.90 0.34 0.85 0.81–0.88
ICU Day 2 (2769) CASUS 96.9 13.97 0.052 0.97 0.96–0.98
SOFA 95.3 6.75 0.56 0.91 0.88–0.93
ICU Day 3 (1234) CASUS 93.8 10.29 0.17 0.94 0.93–0.96
SOFA 90.8 6.45 0.60 0.90 0.88–0.93
ICU Day 4 (815) CASUS 92.4 3.66 0.82 0.93 0.91–0.96
SOFA 89.3 8.35 0.40 0.89 0.86–0.91
ICU Day 5 (566) CASUS 91.2 8.08 0.33 0.92 0.89–0.95
SOFA 86.9 2.46 0.96 0.89 0.85–0.92
ICU Day 6 (430) CASUS 89.5 4.71 0.79 0.90 0.86–0.94
SOFA 85.6 3.98 0.86 0.88 0.84–0.91
ICU Day 7 (338) CASUS 87.9 9.84 0.28 0.90 0.86–0.94
SOFA 84.3 8.80 0.36 0.87 0.82–0.91
ICU: intensive care unit; 95% CI: 95% confidence interval; Add-Euro: additive EuroSCORE; AUC: Area under ROC curve; HL test: Hosmer-Lemeshow test;
Log-Euro: logistic EuroSCORE; OCC: overall correct classification; ROC: receiver operating characteristic test

correctly predicted survivals and mortality to the total between CASUS and SOFA increases progressively. Howev-
number of patients) values were calculated for the collected er, the difference between both scoring models with regard
data. to the AUC was more or less stable or even declined over
time in the ICU. However, CASUS was superior in both
figures for all days.
Results
The study included 2801 patients who were admitted to the
Discussion
ICU over a period of two years. ICU mortality was 5.2% (n ¼
147). 29.6% (n ¼ 830) of patients were female. Mean age was In this study that carried out in 2801 patients, it was found
66.9 10.7 years (range 19–89 years). Types of surgical that both CASUS and SOFA can be used to predict mortality in
procedures are presented in ►Table 3. The length of stay in cardiac surgery patients. However, the mortality prediction
the ICU was 4.3 6.8 days (range 1–189 days, median: 2.0 of the CASUS proved to be more accurate. This study can
days, 75th percentile: 4.0 days). Preoperatively collected additionally be considered as an “external validation” of both
mean additive EuroSCORE was 6.3 3.6, and mean logistic the CASUS and the SOFA scoring models in cardiac surgery
EuroSCORE was 9.9 12.9 (median: 5.3, 75th percentile: patients. The SOFA score was originally developed in 1996 as a
11.3). ►Table 4 summarizes the results of calibration, dis- morbidity risk stratification model for sepsis patients.6 Be-
crimination and overall correct classification for SOFA and cause of its good performance and reliability, SOFA later
CASUS from the 1st ICU day (operative day) to the 7th ICU day became globally accepted and was widely used as a scoring

and for both preoperative EuroSCORE models. model for ICU patients in general, not only to predict morbid-
OCC values for postoperative CASUS were higher com- ity but also mortality.7 In 2003, Ceriani et al.3 suggested using
pared to those of SOFA for all days. ►Fig. 1 demonstrates SOFA for cardiac surgery patients. Based on the good results
the OCC rates of both CASUS and SOFA for each day. CASUS that they obtained in 218 patients, they concluded that SOFA
and SOFA as well as the preoperative additive EuroSCORE can be used in cardiac surgery without any need for specific
showed no significant differences between expected and modifications. In 2005, CASUS was introduced5 to take ac-
observed mortality using the HL test. The logistic Euro- count of the special pathophysiological conditions of ICU
SCORE did not perform well in this calibration test (p ¼ patients after cardiac surgery. This scoring system was evalu-
0.01, see ►Table 4). Areas under ROC curves of CASUS ated in 3230 patients operated in the same center, who were
(0.90) were larger than those of both EuroSCOREs and divided into one construction set and two validation sets. In a
SOFA for all days. Overall, the best ROC curves result was for subgroup of these same sets of patients, the CASUS model was
CASUS on the second ICU day (0.97). ►Fig. 2 shows the also tested with regard to the reliability of its derivatives
AUCs of both CASUS and SOFA for each postoperative (mean, maximum and delta CASUS).8 The results of both
day. ►Fig. 3 demonstrates the AUCs for both scoring models assessments were promising. However, our present study is
from day 1 to day 7 collectively. On examining both ►Figs. 1 the first external validation of CASUS, i.e., undertaken in
and 3, ►Fig. 1 shows that the difference in OCC values another institution.
Figure 1 Overall correct classification rates for CASUS and SOFA from day 1 to day 7.

Figure 2 ROC curves for CASUS and SOFA from day 1 to day 7.
To start with, the results of the HL test indicated that there more accurate with regard to mortality prediction, since the
were no significant differences between expected and ob- models have different maximum numbers of points. OCC and
served mortality for both CASUS and SOFA. This means that ROC analyses are more informative for this purpose.
both scores were suitable as scaled models, where the It is known that perfect discrimination (ROC test) is
mortality ratio generally increases in the study population important to evaluate the individual patient's risk using a
as the average values of the scoring systems increase. How- scoring system, while better calibration is needed for clinical
ever, the results of HL tests could not be used to compare the trials or comparison of care between ICUs.9 OCC reflects the
two models to each other or to conclude which of the two was ability of the model to correctly predict a total number of


Figure 3 Summary and results of area under the ROC curve for CASUS and SOFA from day 1 to day 7.
survivals and mortalities, which is expressed as a percentage An analysis of ►Figs. 1 and 3 shows that the differences
of the whole study population. The correctly predicted per- between the OCC values of both models progressively in-
centage was higher with the CASUS model. ►Table 5, which creased with a longer stay in the ICU, while the differences in
presents the analysis of the highest OCC values of both models AUCs progressively decreased. The patients that stay longer in
(day 2) in detail, shows that CASUS correctly predicted 56.8% the ICU after cardiac surgery are in a more critical condition
of deaths, while SOFA predicted only 12.9% of deaths. More- and the average of their scoring values and mortality rates can
over, OCC showed the ratio of predicted to observed outcome be expected to be higher. In addition, the survivors (patients
without considering whether survival or mortality actually with better organ function) are discharged from the ICU,
occurred in the specific individuals predicted by the model to whereupon they are excluded from the study population.
live or die. In other words, OCC did not demonstrate the Hence, it seems to be logical for the AUCs of both models to
ability of the model to identify the individualized risk for each gradually get closer to each other as the risk of mortality rises
patient. Correct prognosis of individualized risk depends on among those who stay in the ICU. This indicates that CASUS is
the specificity and sensitivity of the model, which can be more specific to cardiac surgery patients and more capable of
tested by ROC analysis. These ROC analyses again favored accurately predicting the individual patient's risk, as its risk
CASUS for all days. detection and good discrimination starts early after ICU
Table 5 The results of overall correct classification of both models on day 2.
Observed Predicted ICU mortality

No Yes Percentage correct
CASUS ICU mortality
No 2609 28 98.9
Yes 57 75 56.8
Overall percentage 96.9
SOFA ICU mortality
No 2623 14 99.5
Yes 115 17 12.9
Overall percentage 95.3
ICU: intensive care unit.

admission and identifies possible mortality, even when the ventricular assist devices (VAD). This was based on the results
study population consists of patients with a large variety of of Higgins et al.,16 who reported an odds ratio for the risk of
organ dysfunction and even when the immediate postopera- mortality of 4.46 for postoperative IABP usage. VAD were
tive effect of cardiopulmonary bypass (CPB) is still present. assumed to have the same impact on the incidence of
The more the variables of a scoring system match the patho- mortality.5 Modified MODS15 was the only scoring system
physiological conditions of a specific group of patients, the that considered serum lactate when its value exceeds 5
larger the AUCs in a ROC test and the greater the accuracy of mmol/L. However, Jansen et al.17 demonstrated in a systemic
its discrimination. health technology assessment that blood lactate monitoring
The major differences between CASUS and SOFA are the should have a place in the risk stratification of critically ill
variables used to represent three main systems: the renal patients. Even relative hyperlactatemia (<2 mmol/L) was
system, the central nervous system and the cardiovascular shown to be associated with an increased risk of mortality
system. The renal system was represented in both SOFA and in another study.18 Meanwhile, according to the results of
CASUS by serum creatinine levels. The SOFA score introduced different studies and based on the availability of monitoring
oliguria as a criterion for renal failure,10 which shares the serum lactate in different ICUs, its integration in scoring
same points with serum creatinine (points 3 and 4 models is beyond debate.
in ►Table 1). In contrast, CASUS divided the criteria for renal Other differences between both scores are limited to the
dysfunction into two separate variables: serum creatinine choice of the cutoff points of variables. We have to remember
and any form of renal replacement therapy. It was claimed that the construction of both models was based on retrospec-

that renal replacement therapy masks renal dysfunction and, tive analyses carried out in an attempt to identify the descrip-
accordingly, should be evaluated separately.11 tors of mortality and multiorgan dysfunction in two different
Vincent et al.6 suggested the Glasgow Coma Scale (GCS) as categories of patients (sepsis patients for SOFA and cardiac
the parameter of choice for assessing the degree of morbidity surgery patients for CASUS). This renders these small differ-
of the central nervous system (SOFA score, ►Table 1). How- ences acceptable. We chose in our study to evaluate both
ever, in the same study, they stated that the application of GCS scoring systems from the day of ICU admission (day 1) until
is limited and that it was not clear whether the actual or the day 7. It is true that patients who stayed <48 hours in the ICU
assumed GCS should be used when a patient is actively were excluded from the target population of SOFA.6,10 How-
sedated. This might be more obvious in cardiac surgery ever, most of the other scoring systems were designed to be
patients, where sedation is actively continued for several used only on ICU admission. We assumed that a reliable model
hours after ICU admission until achieving appropriate hemo- should be validated from ICU admission until discharge. We
dynamic stability.12,13 Moreover, there are different protocols had to limit our study to seven days as the number of patients
for the de-escalation of sedation that vary from one ICU to thereafter was too small for reliable statistical analysis.
another. Furthermore, accurate evaluation with the GCS
needs several minutes per patient, which also limits its
Conclusion
acceptance as a variable included in an ICU scoring system.
For these reasons, it was assumed that GCS was not the best Both the CASUS and the SOFA scoring systems are reliable as
tool for neurological assessment in cardiac surgery. Instead, mortality risk stratification models in cardiac surgery pa-
CASUS introduced a specific variable called “neurological tients for the first seven postoperative days. However, CASUS
state” (►Table 2). This variable was intended to take account proved to be more accurate in the prediction of the risk of
of the protocol for sedation and weaning from ventilators mortality for individual patients. It should be adopted in
after cardiac surgery. cardiac surgery ICUs.
The cardiovascular system, as the most important system
in cardiac surgery patients, was evaluated in CASUS by four
separate variables (►Table 2). Use of a pressure-adjusted Acknowledgements
heart rate was inspired by the original “Multiple Organ The authors wish to thank Dr. Tobias Berg for his substan-
Dysfunction Score” (MODS) system.14 Although it was re- tial technical and statistical support; and for the realiza-
placed in the modified MODS15 by a combination of heart tion of the online CASUS calculator, which can be found on
rate, inotropic agents and serum lactate >5 mmol/L, this the homepage of the Department of Cardiothoracic Sur-
modification was found by the group that constructed CASUS gery, http://www.htchirurgie.uniklinikum-jena.de or on
not to be suitable for cardiac surgery.5 They found that heart the CASUS homepage (http://www.cardiac-icu.org). An
rate per se did not correlate with mortality. They neglected to Applet (Cardiac ICU) for iPhone, iPad and iPod touch is
include inotropic agents as a predictor of mortality since the available for free from the iTunes App store: (http://itunes.
types and doses differ from hospital to hospital and even apple.com/us/app/cardiac-icu/id389965786?mt ¼ 8).
between physicians within the same hospital. In fact, Vincent
et al.10 claimed that a scoring system should avoid, if possible,
any therapeutic measures so as to avoid any biases. However,
inotropic agents were included as a variable in SOFA. In References
contrast, CASUS included two major cardiovascular thera- 1 Knaus W, Draper E, Wagner D, Zimmerman J. APACHE II a severity
peutic procedures: intra-aortic balloon pump (IABP) and of disease classification system. Crit Care Med 1985;13;818–829

2 Marshall JC, Cook DJ, Christou NV, Bernard GR, Sprung CL, Sibbald care units: results of a multicenter, prospective study. Working
WJ. Multiple organ dysfunction score: a reliable descriptor of a group on “sepsis-related problems” of the European Society of
complex clinical outcome. Crit Care Med 1995;23;1638–1652 Intensive Care Medicine. Crit Care Med 1998;26;1793–1800
3 Ceriani R, Mazzoni M, Bortone F, et al. Application of the sequential 11 Hekmat K, Doerr F, Kroener A, et al. Prediction of mortality in
organ failure assessment score to cardiac surgical patients. Chest intensive care unit cardiac surgical patients. Eur J Cardiothorac
2003;123;1229–1239 Surg 2010;38;104–109
4 Pätilä T, Kukkonen S, Vento A, Pettilä V, Suojaranta-Ylinen R. 12 Rady MY, Ryan T. Perioperative predictors of extubation failure
Relation of the sequential organ failure assessment score to and the effect on clinical outcome after cardiac surgery. Crit Care
morbidity and mortality after cardiac surgery. Ann Thorac Surg Med 1999;27;340–347
2006;82;2072–2079 13 Kollef MH, Wragge T, Pasque C. Determinants of mortality and
5 Hekmat K, Kroener A, Stuetzer H, et al. Daily assessment of organ multiorgan dysfunction in cardiac surgery patients requiring
dysfunction and survival in intensive care unit cardiac surgical prolonged mechanical ventilation. Chest 1995;107;1395–1401
patients. Ann Thorac Surg 2005;79;1555–1562 14 Marshall JC, Cook DJ, Christou NV, Bernard GR, Sprung CL, Sibbald
6 Vincent JL, Moreno R, Takala J, et al. The SOFA (Sepsis-related WJ. The multiple organ dysfunction (MOD) score: a reliable
Organ Failure Assessment) score to describe organ dysfunction/ descriptor of a complex clinical outcome. Crit Care Med 1995;
failure. On behalf of the working group on sepsis-related problems 23;1638–1652
of the European Society of Intensive Care Medicine. Intensive Care 15 Cook R, Cook D, Tilley J, et al. Multiple organ dysfunction:
Med 1996;22;707–710 baseline and serial component scores. Crit Care Med 2001;
7 Vincent JL, Ferreira F, Moreno R. Scoring systems for assessing 29;2046–2050
organ dysfunction and survival. Crit Care Clin 2000;16;353–366 16 Higgins TL, Estafanous FG, Loop FD, et al. ICU admission score for
8 Badreldin AM, Kroener A, Heldwein MB, et al. Prognostic value of predicting morbidity and mortality risk after coronary artery

daily Cardiac Surgery Score (CASUS) and its derivatives in cardiac bypass grafting. Ann Thorac Surg 1997;64;1050–1058
surgery patients. Thorac Cardiovasc Surg 2010;58;392–397 17 Jansen TC, VAN Bommel J, Bakker J. Blood lactate monitoring in
9 Sakr Y, Krauss C, Amaral A, et al. Comparison of the performance of critically ill patients: a systematic health technology assessment.
SAPS II, SAPS3, APACHE II, and their customized prognostic models Crit Care Med 2009;37;2827–2839
in a surgical intensive care unit. Br J Anaesth 2008;101;798–803 18 Nichol AD, Egi M, Pettila V, et al. Relative hyperlactatemia and
10 Vincent JL, de Mendonça A, Cantraine F, et al. Use of the SOFA score hospital mortality in critically ill patients: a retrospective multi-
to assess the incidence of organ dysfunction/failure in intensive centre study. Crit Care 2010;14(1):R25

GuaragnaSCORE satisfactorily predicts outcomes

in heart valve surgery in a Brazilian hospital
GuaragnaSCORE prediz satisfatoriamente os desfechos em cirurgia cardíaca valvar em hospital
brasileiro
Michel Pompeu Barros de Oliveira Sá1, Marcus Villander Barros de Oliveira Sá2, Ana Carla Lopes
de Albuquerque2, Belisa Barreto Gomes da Silva2, José Williams Muniz de Siqueira2, Phabllo Rodrigo
Santos de Brito2, Frederico Pires Vasconcelos2, Ricardo de Carvalho Lima3
DOI: 10.5935/1678-9741.20120003 RBCCV 44205-1344
Abstract association (P <0.001). The score presented a good accuracy,

Objective: The aim of this study is to evaluate the since the discrimination power of the model in this study
applicability of GuaragnaSCORE for predicting mortality according to the ROC curve was 78.1%.
in patients undergoing heart valve surgery in the Division Conclusions: The Brazilian score proved to be a simple
of Cardiovascular Surgery of Pronto Socorro Cardiológico and objective index, revealing a satisfactory predictor of
de Pernambuco - PROCAPE, Recife, PE, Brazil. perioperative mortality in patients undergoing heart valve
Methods: Retrospective study involving 491 consecutive surgery at our institution.
patients operated between May/2007 and December/2010.
The registers contained all the information used to calculate Descriptors: Risk. Heart valve diseases. Cardiovascular
the score. The outcome of interest was death. Association of surgical procedures.
model factors with death (univariate analysis and
multivariate logistic regression analysis), association of risk Resumo
score classes with death and accuracy of the model by the Objetivo: O objetivo deste estudo é avaliar a aplicabilidade
area under the ROC (receiver operating characteristic) curve do GuaragnaSCORE na predição de mortalidade
were calculated. perioperatória em pacientes submetidos à cirurgia cardíaca
Results: The incidence of death was 15.1%. The nine valvar na Divisão de Cirurgia Cardiovascular do Pronto
variables of the score were predictive of perioperative death Socorro Cardiológico de Pernambuco - PROCAPE, Recife,
in both univariate and multivariate analysis. We observed PE, Brasil.
that the higher the risk class of the patient (low, medium, Métodos: Estudo retrospectivo envolvendo 491 pacientes
high, very high, extremely high), the greater is the incidence consecutivos operados entre maio/2007 e dezembro/2010. Os
of postoperative AF (0%; 7.2%; 25.5%; 38.5%; 52.4%), registros continham todas as informações utilizadas para
showing that the model seems to be a good predictor of risk calcular a pontuação. O desfecho de interesse foi óbito. A
of postoperative death, in a statistically significant associação de fatores do escore com óbito (análise univariada
1 – MD, MSc. Corresponding author:

2 – MD Michel Pompeu Barros de Oliveira Sá
3 – MD, MSc, PhD, ChM. Av. Eng. Domingos Ferreira, 4172/405 – Recife, PE, Brazil – ZIP
code: 51021-040.
E-mail: michel_pompeu@yahoo.com.br
This work was carried out at Division of Cardiovascular Surgery of
Pronto Socorro Cardiológico de Pernambuco (PROCAPE). University Article received on November 8th, 2011
of Pernambuco (UPE), Recife, PE, Brazil. Article accepted on January 3rd, 2012
1
Sá MPBO, et al. - GuaragnaSCORE satisfactorily predicts outcomes Rev Bras Cir Cardiovasc 2012;27(1):1-6
Resultados: A incidência de óbito foi de 15,1%. As nove

Abbreviations, acronyms & symbols
variáveis do escore foram preditoras de morte em análise
CABG coronary artery bypass graft univariada e multivariada. Observamos que, quanto maior
CI confidence interval a classe de risco do paciente (baixa, média, alta, muito alta,
CPB cardiopulmonary bypass extremamente alta), maior é a incidência de óbito (0%;
EF ejection fraction 7,2%; 25,5%; 38,5%; 52,4%), demonstrando que o modelo
NYHA New York Heart Association criteria parece ser um bom preditor de risco de óbito, em uma
OR odds ratio associação estatisticamente significativa (P<0,001). O escore
PAH pulmonary arterial hypertension
apresentou boa acurácia, levando em consideração que a
ROC curve receiver operating characteristic curve
SPSS Statistical Package for Social Sciences área sob a curva ROC foi de 78,1%.
Conclusões: O escore brasileiro demonstrou-se um índice
simples e objetivo, revelando-se um preditor satisfatório de
óbito no período perioperatório em pacientes submetidos à
e análise de regressão logística multivariada), associação cirurgia cardíaca valvar em nossa instituição.
de classes de risco do escore com óbito e acurácia do modelo
através da área sob a curva ROC (receiver operating Descritores: Risco. Doenças das valvas cardíacas.
characteristic) foram calculados. Procedimentos cirúrgicos cardiovasculares.
INTRODUCTION surgery at the Division of Cardiovascular Surgery of Pronto

Socorro Cardiológico de Pernambuco (PROCAPE), Recife,
Currently, a total of 275.000 cardiac valve replacement PE, Brazil, from May 2007 to December 2010. We excluded
surgeries are carried out worldwide [1], with operative the following: patients whose records did not contain the
mortality ranging from 1 to 15% [2,3]. The reported mortality necessary data concerning the variables to be studied;
in Brazil is 8.9% for heart valve surgeries, according to patients undergoing surgery for tricuspid and/or pulmonary
administrative register from DATASUS [4]. valves (when isolated, due to small number of patients
Guaragna et al. [5] recently proposed a Brazilian risk undergoing these procedures); age < 18 years.
score for prediction of surgical risk after heart valve surgery
– we baptized the model as GuaragnaSCORE. However, Study design
several studies show that risk prediction scores tend to It was a retrospective study of exposed and nonexposed
have inferior performance when applied to different groups to certain factors (independent variables) with outcome
of patients which have been used to development of the (dependent variable) followed by assessment of a model
original model [6]. So the external assessment in population (the score of Guaragna et al. [5]).
of patients with new data from other institutions is always The independent variables were: gender (male/female),
important for the score has wide clinical application [7-9]. age (years), surgical priority (emergency/urgency surgery
Previously, we tested EuroSCORE in coronary artery bypass considered as a single variable and defined as the need to
graft (CABG) surgery at our institution, and this proved to undergo surgical intervention in up to 48 hours, due to
be a simple and objective index, revealing a discriminating imminent risk of death or unstable clinical-hemodynamic
satisfactory postoperative outcome, so we showed the condition), heart failure functional class according to New
importance of validating risk prediction models in local York Heart Association criteria (NYHA I, II, III, IV), ejection
institutions in order to verify its applicability [8]. fraction (EF%, measured by echocardiography), serum
The objective of this study is to evaluate the ability of creatinine (mg/dL), pulmonary arterial hypertension (PAH,
the score of Guaragna et al. [5] in predicting surgical risk in detected at the echocardiogram, defined as systolic
our institution, specifically in the group undergoing heart pressure in pulmonary artery ≥ 30 mmHg according to the
valve surgery. Brazilian Guideline of Pulmonary Arterial Hypertension of
2005), combined CABG surgery.
METHODS The dependent variable was perioperative death
(considered in the transoperative period and throughout
Source population the entire hospitalization period).
After approval by the ethics committee, in accordance Each patient was evaluated for the presence or absence of
with Resolution 196/96 (National Board of Health – Ministry the nine risk factors established by Guaragna et al. [5],
of Health – Brazil) [10,11], we reviewed the records of respecting the definition of each of them and giving them the
patients undergoing consecutive isolated heart valve correct score (Table 1). Depending on the final score, each
surgery (replacement or repair) or combined with CABG patient was placed in one of the five risk groups (Table 2).
2
Table 1. Factors associated with development of outcome (death) with a mean age of 44.6 ± 17.9 years, being 51.5% female.
after heart valve surgery and appropriate score In-hospital death occurred in 15.1% (n=74) patients.
Clinical profile Score
Age > 60 years 3 Univariate analysis
Emergency/urgency surgery 17
Analyzing the variables proposed in the score with the
Female sex 2
Ejection fraction < 45% 2 occurrence of death, we observed that all of them were
Combined CABG 3 significantly associated with this complication (Figure 1).
Pulmonary arterial hypertension 2
NYHA class III or IV 2
Creatinine 1.5 - 2.49 mg/dL 2
Creatinine > 2.5 mg/dL or dialysis 6
Table 2. Risk category according to the score

Risk category Total score
Low 0-3
Medium 4-6
High 7-9
Very High 10 - 13
Extremely high > 14
Statistical methods
Data were analyzed using percentage and descriptive
statistics measures. The following tests were used: chi-
square test or Fisher’s exact test (as appropriate, for non-
parametric variables). In the study of univariate association
between categorical variables, the values of the odds ratio
(OR) and a confidence interval (CI) for this parameter with
a reliability of 95% were obtained.
Multivariate analysis was adjusted to a logistic
regression model to explain the proportion of patients who
died that were significantly associated to the level of 5% (P
<0.05) by a backward elimination procedure. The calibration
of multivariate model was evaluated by the Hosmer-
Lemeshow goodness-of-fit test.
The accuracy (discrimination ability of the score) was
calculated using the area under the ROC curve (receiver Fig. 1 - Association of clinical characteristics with the occurrence
of death after heart valve surgery (univariate analysis). EF - ejection
operating characteristic curve), built on correct prediction
fraction; CABG - coronary artery bypass graft
of death (among high, very high and extremely high risk
categories) and correct prediction of survival (among low
and medium risk categories).
The level of significance in the decision of the statistical Multivariate analysis
tests was 5%. The program used for data entry and retrieval Applying a multivariate logistic regression model,
of statistical calculations was SPSS (Statistical Package for associations of clinical variables of the score remained
Social Sciences) version 15.0. strongly associated with death (Table 3). The model was
well accepted (P<0.001) and showed a degree of explanation
RESULTS of 88.4%. The Hosmer-Lemeshow goodness-of-fit was also
well accepted (P=0.811), indicating a good model calibration.
Incidence of death and population characteristics
Taking into account the inclusion and exclusion criteria, Analysis of the score and prediction of death
we analyzed 491 patients undergoing heart valve surgery The incidence of death according to the risk score
3
Table 3. Multivariate logistic regression model

OR / 95%CI
Variable Univariate analysis Multivariate analysis P value
Age > 60 years 5.04 (2.99 - 8.52) 3.33 (1.67 - 6.64) 0.001*
Female 2.20 (1.30 - 3.71) 2.40 (1.23 - 4.67) 0.010*
NYHA class III - IV 3.84 (2.11 - 6.99) 2.14 (1.04 - 4.43) 0.040*
Combined CABG 7.98 (4.11 - 15.50) 6.56 (2.71 - 15.89) < 0.001*
Ejection fraction < 45% 3.03 (1.73 - 5.31) 2.30 (1.09 - 4.87) 0.029*
Emergency/urgency surgery 9.01 (4.83 - 16.79) 7.48 (3.46 - 16.18) < 0.001*
Pulmonary arterial hypertension 2.65 (1.28 - 5.50) 2.77 (1.13 - 6.78) 0.026*
Creatinine 1.50 - 2.49 mg/dL 3.35 (1.81 - 6.21) 2.20 (1.07 - 7.00) 0.048*
Creatinine > 2.5 mg/dL 3.25 (1.33 - 7.96) 2.09 (1.03 - 8.00) 0.049*
(*): Significative at 5% level. Constant P < 0.001
Fig. 2 - Relationship between the risk group classification according

to the score and incidence of death. Note the upward curve as it
increases the risk class
classification is showed in Figure 2. We observed that the

higher is the risk category, the higher is the incidence of Fig. 3 - Receiver operating characteristic (ROC) curve. The graphic
death, in a statistically significant association (P<0.001). shows the good accuracy of the model
Accuracy of the proposed risk score

According to the results presented in the area under We observed that age ≥ 60 years is an independent
the ROC curve (overall capacity of the measure used to predictor of death in patients undergoing heart valve
discriminate individuals who died or survived), measured surgery. Almeida et al. [12] already demonstrated that age
by 78.1%, the score shown a good measure to identify is associated with the occurrence of death after heart valve
patients with risk of death (Figure 3). surgery regardless of type of prosthesis (biological or
mechanics).
DISCUSSION We also observed that female is an independent
predictor of death in patients undergoing heart valve
The incidence of death in our study was 15.1%. This is surgery. Andrade et al. [13] demonstrated that female gender
28% greater than that observed in the original study by increases by 2 times the chance of death in patients
Guaragna et al. [5], which was 11.8%. This is probably undergoing heart valve surgery (independent association).
because our population has surplus of 17% of female, 25% We also observed that ejection fraction ≤ 45% is an
of patients with NYHA class III/IV, 35.2% of left ventricular independent predictor of death in patients undergoing heart
dysfunction, 301.6% of pulmonary arterial hypertension, valve surgery. De Bacco et al. [14] showed that left
81.5% of creatinine 1.50-2.49 mg/dL and 213.6% of creatinine ventricular dysfunction is associated with increased
≥ 2.5 mg/dL, which makes our population as higher risk. mortality in patients undergoing heart valve replacement.
4
We also observed that combined CABG is an REFERENCES

independent predictor of death in patients undergoing
heart valve surgery. We think it happens for two reasons. 1. Rabkin E, Schoen FJ. Cardiovascular tissue engineering.
First, many studies show that combination of two or more Cardiovasc Pathol. 2002;11(6):305-17.
concomitant cardiac surgeries increases the risk of death
2. Brandão CMA. Avaliação do risco em cirurgia cardíaca valvar.
[5,8,9]. Second, the combination with CABG increases In: Grinberg M, Sampaio RO, eds. Doença valvar.
time of cardiopulmonary bypass (CPB). Many studies Barueri:Manole;2006. p.199-201.
showed the impact of CPB in increased rates of morbidity
and mortality in various situations related to cardiac 3. Ambler G, Omar RZ, Royston P, Kinsman R, Keogh BE,
surgery [15,16]. Taylor KM. Generic, simple risk stratification model for heart
We also observed that pulmonary arterial hypertension valve surgery. Circulation. 2005;112(2):224-31.
is an independent predictor of death in patients undergoing
heart valve surgery. Roques el al. [17] demonstrated the 4. Ribeiro AL, Gagliardi SP, Nogueira JL, Silveira LM, Colosimo
EA, Lopes do Nascimento CA. Mortality related to cardiac
importance of pulmonary hypertension on outcomes during
surgery in Brazil, 2000-2003. J Thorac Cardiovasc Surg.
the development of EuroSCORE. 2006;131(4):907-9.
We also observed that NYHA class III/IV is an
independent predictor of death in patients undergoing heart 5. Guaragna JC, Bodanese LC, Bueno FL, Goldani MA. Proposed
valve surgery. Despite this evidence, we must remember preoperative risk score for patients candidate to cardiac valve
that it is precisely these patients (patients more surgery. Arq Bras Cardiol. 2010;94(4):541-8.
symptomatic) who benefit from more aggressive strategies
(surgical treatment) compared with medical therapy [18]. 6. Shahian DM, Blackstone EH, Edwards FH, Grover FL,
We also observed that high levels of creatinine is an Grunkemeier GL, Naftel DC; STS workforce on evidence-
based surgery. Cardiac surgery risk models: a position article.
independent predictor of death in patients undergoing heart
Ann Thorac Surg. 2004;78(5):1868-77.
valve surgery. Volkmann et al. [19] showed that, beyond
patients with overt renal insufficiency, patients with hidden 7. Sá MP, Figueira ES, Santos CA, Figueiredo OJ, Lima RO,
renal dysfunction (normal creatinine, but with decreased Rueda FG, et al. Validation of MagedanzSCORE as a predictor
glomerular filtration rate) had a higher incidence of death of mediastinitis after coronary artery bypass graft surgery.
after cardiac surgery, demonstrating the impact of renal Rev Bras Cir Cardiovasc. 2011;26(3):386-92.
disease in outcomes.
It is worth mentioning that other scores have been tested 8. Sá MP, Soares EF, Santos CA, Figueredo OJ, Lima RO, Escobar
in cardiac surgery, although not specifically in heart valve RR, et al. EuroSCORE and mortality in coronary artery bypass
graft surgery at Pernambuco Cardiologic Emergency Medical
surgery. Recently, Mejía et al. [9] tested the EuroSCORE
Services [Pronto Socorro Cardiológico de Pernambuco]. Rev
and 2000 Parsonnet-Bernstein in Heart Institute of the Bras Cir Cardiovasc. 2010;25(4):474-82.
University of São Paulo Clinics Hospital (InCor-USP). They
showed that the similarity between both observed and 9. Mejía OA, Lisboa LA, Puig LB, Dias RR, Dallan LA,
expected mortality by the scores allows them to confirm Pomerantzeff PM, et al. The 2000 Bernstein-Parsonnet score
that the values given by the scores to the various risk and EuroSCORE are similar in predicting mortality at the Heart
factors could be applied to their patients. Therefore, they Institute, USP. Rev Bras Cir Cardiovasc. 2011;26(1):1-6.
concluded that both models were similar and adequate in
predicting the mortality of patients undergoing CABG, heart 10. Sá MP, Lima RC. Research Ethics Committee: mandatory
necessity. Requirement needed. Rev Bras Cir Cardiovasc.
valve, and associated surgeries at their institution. The
2010;25(3):III-IV.
GuaragnaSCORE stands out in the sense of having been
created specifically for heart valve surgery. 11. Lima SG, Lima TA, Macedo LA, Sá MP, Vidal ML, Gomes
The GuaragnaSCORE presented a good accuracy, since AF, et al. Ethics in research with human beings: from knowledge
the discrimination power of the model in this study to practice. Arq Bras Cardiol. 2010;95(3):289-94.
according to the ROC curve was 78.1%.
12. Almeida AS, Picon PD, Wender OC. Outcomes of patients subjected
CONCLUSIONS to aortic valve replacement surgery using mechanical or biological
prostheses. Rev Bras Cir Cardiovasc. 2011;26(3):326-37.
The risk score proposed by Guaragna et al. [5] seems to
13. Andrade IN, Moraes Neto FR, Oliveira JP, Silva IT, Andrade
be a good model for prediction of death in patients TG, Moraes CR. Assesment of the EuroSCORE as a predictor
undergoing heart valve surgery. for mortality in valve cardiac surgery at the Heart Institute of
Pernambuco. Rev Bras Cir Cardiovasc. 2010;25(1):11-8.
5
14. De Bacco MW, Sartori AP, Sant’Anna JR, Santos MF, Prates 17. Roques F, Nashef SA, Michel P, Gauducheau E, de Vincentiis
PR, Kalil RA, et al. Risk factors for hospital mortality in valve C, Baudet E, et al. Risk factors and outcome in European
replacement with mechanical prosthesis. Rev Bras Cir cardiac surgery: analysis of the EuroSCORE multinational
Cardiovasc. 2009;24(3):334-40. database of 19030 patients. Eur J Cardiothorac Surg.
1999;15(6):816-22.
15. Anderson AJ, Barros Neto FX, Costa MA, Dantas LD, Hueb
AC, Prata MF. Predictors of mortality in patients over 70 18. Kassab AK, Kassab KK. Mitral regurgitation: comparison
years-old undergoing CABG or valve surgery with among clinical and surgical treatment medium term in agreement
cardiopulmonary bypass. Rev Bras Cir Cardiovasc. with the functional class. Rev Bras Cir Cardiovasc.
2011;26(1):69-75. 2002;17(2):128-31.
16. Sá MP, Lima LP, Rueda FG, Escobar RR, Cavalcanti PE, Thé 19. Volkmann MA, Behr PE, Burmeister JE, Consoni PR, Kalil
EC, et al. Comparative study between on-pump and off-pump RA, Prates PR, et al. Hidden renal dysfunction causes increased
coronary artery bypass graft in women. Rev Bras Cir in-hospital mortality risk after coronary artery bypass graft
Cardiovasc. 2010;25(2):238-44. surgery. Rev Bras Cir Cardiovasc. 2011;26(3):319-25.
6
Saudi J Anaesth. 2012 Jul-Sep; 6(3): 242–247.
doi: 10.4103/1658-354X.101215
PMCID: PMC3498662
Perioperative predictors of morbidity

and mortality following cardiac surgery
under cardiopulmonary bypass
Ishwar Bhukal, Sohan Lal Solanki,1 Shankar Ramaswamy,2 Lakshmi Narayana Yaddanapudi,
Amit Jain, and Pawan Kumar1
Department of Anaesthesia and Intensive Care, Postgraduate Institute of Medical Education

and Research, Chandigarh, India
1
Department of Anaesthesiology, Sanjay Gandhi Postgraduate Institute of Medical Sciences,
Lucknow, India
2
Department of Anaesthesiology, Western General Hospital, Edinburgh, United Kingdom
Address for correspondence: Dr. Sohan Lal Solanki, Senior Resident, Department of
Anaesthesiology, SGPGIMS, Rae-Barreily Road, Lucknow, Uttar Pradesh, India. E-mail:
me_sohans@yahoo.co.in
Copyright : © Saudi Journal of Anaesthesia
This is an open-access article distributed under the terms of the Creative Commons
Attribution-Noncommercial-Share Alike 3.0 Unported, which permits unrestricted use,
distribution, and reproduction in any medium, provided the original work is properly cited.
Go to:
Abstract
Background:
Prediction of outcome after cardiac surgery is difficult despite a number of models

using pre-, intra- and post-operative factors. Ideally, risk factors operating in all three
phases of the patients’ stay in the hospital should be incorporated into any outcome
prediction model. The aim of the present study was to identify the perioperative risk
factors associated with morbidity, mortality and length of stay in the recovery room
(LOSR) and length of stay in the hospital (LOSH).
Methods:
Eighty-eight adults of either sex, patients undergoing elective open cardiac surgery were
studied prospectively. The ability of a number of pre-, intra- and post-operative factors
to predict outcome in the form of mortality, immediate morbidity (LOSR) and
intermediate morbidity (LOSH) was assessed.
Results:
Factors associated with higher mortality were preoperative prothrombin index (PTI),
American Society of Anesthesiology-Physical Status (ASA-PS) grade, Cardiac
Anaesthesia Risk Evaluation (CARE) score and New York Heart Association (NYHA)
class, intraoperative duration of cardiopulmonary bypass (DCPB), number of inotropes
used while coming off cardiopulmonary bypass and postoperatively, Acute Physiology
and Chronic Health Evaluation (APACHE) II excluding the Glassgow Comma Scale
(GCS) component and the number of inotropes used. Immediate morbidity was
associated with preoperative PTI, inotrope usage intra- and post-operatively and the
APACHE score. Intermediate morbidity was associated with DCPB and intra- and post-
operative inotrope usage. Individual surgeon influenced the LOSR and the LOSH.
Conclusion:
APACHE score, a general purpose severity of illness score, was relatively ineffective in
the postoperative period because of sedation, neuromuscular blockade and elective
ventilation used in a number of these patients. The preoperative and intraoperative
factors like CARE, ASA-PS grade, NYHA, DCPB and number of inotropes used
influencing morbidity and mortality are consistent with the literature, despite the small
size of our sample.
Keywords: Cardiac surgery, outcome prediction, prognostication, risk stratification
Go to:
INTRODUCTION
Cardiac surgery has remained a very complex area for outcome prediction. Although
several severity scoring systems for general intensive care unit (ICU) purposes, like the
Acute Physiology and Chronic Health Evaluation (APACHE) III score, Mortality
Prediction Models (MPM) II and III and the Simplified Acute Physiology Score (SAPS)
II, have matured through two or three generations, they still do not apply well in cardiac
surgery patients. In actual fact, such patients were deliberately excluded during the
development of many of these scoring systems.[1–5]
The outcome prediction models used specifically for cardiac surgery include Cardiac
Anaesthesia Risk Evaluation CARE score,[6] Parsonnet score,[7] Tuman score,[8] Tu
score[9] and European System for Cardiac Operative Risk Evaluation (EuroSCORE)
score,[10] which used preoperative factors to predict the outcome. Intraoperative events
such as the duration of cardiopulmonary bypass (DCPB) and cross-clamp time are
known to be associated with postoperative outcome.[11]
The postoperative course of cardiac surgical patients has been studied using various
general severities of illness scoring systems such as the APACHE versions II[11–13]
and III,[11,14] SAPS, Organ System Failure Index (OSFI) and a number of MPM[13]
and also by comparison of EuroSCORE and Parsonnet score.[15]
Ideally, risk factors operating in all three phases of the patient's stay in the hospital, i.e.,
the pre-, intra- and postoperative periods, should be assessed for their ability to predict
the outcome. This study has been planned to identify the perioperative risk factors
associated with morbidity, mortality and LOSR and LOSH.
Go to:
METHODS
Institutional ethical committee approval was taken and 88 adult patients of either sex
between 18 and 70 years of age undergoing elective open cardiac surgery (coronary
artery bypass grafting (CABG), valve replacement and correction of congenital heart
diseases) under CPB were included in this prospective observational study. A written
informed consent was obtained from all the patients for participating in this study.
Exclusion criteria were patients undergoing off-pump CABG, patients with morbid
obesity and patients who needed intubation in the preoperative period.
Preoperative data including patient's demographics, weight, body mass index (BMI),
CARE score, ASA-PS grade, NYHA functional classification grade, serum electrolytes,
hematrocit, random blood sugar, blood urea, serum creatinine, prothombin time (PTI),
activated partial thromboplastin time (aPTT), 12-lead electrocardiogram (ECG), chest
X-ray, echocardiogram, angiogram (if available), pulmonary function tests and current
medications were noted.
Recorded intraoperative data included DCPB, duration of aortic cross-clamp (DACC),

urine output during the surgery (pre-CPB, during CPB and post-CPB), inotropes used,
significant intraoperative events and their management. Postoperative data included
duration of sedation, time of extubation or tracheostomy and time of decannulation of
tracheostomy, APACHE II score for the first 24 h, significant postoperative events and
their management and the LOSR and the LOSH. The best GCS recorded in the first 24 h
after the operation after stopping the sedation was used for the calculation of the
APACHE II score. Patients who died or were sedated beyond 24 h after the surgery
were not included for the calculation of APACHE II. Patients were started to be weaned
from the ventilator once the following criteria were met: no acute ischemia,
hemodynamically stable (mean arterial pressure >65, cardiac index >2), absence of new
arrhythmia, blood loss <2 mL/kg/h, urine output ≥ 1 mL/kg/h, demonstrating signs of
awakening from anesthesia and core temp 97.0 F or greater. Patients were extubated
once the following criteria met: patient is awake, cooperative and following commands,
able to lift head off the pillow, PO 2 >80 mmHg with FIO 2 <0.40 on continuous positive
airway pressure (CPAP), spontaneous tidal volumes >5 cc/kg and respiratory rate (RR)
<30.
Discharge criteria from the hospital were stable hemodynamics, afebrile for the past 24
h, no surgical incision discharge, independence in daily living activity, oral food intake
and normal bowel function.
Details of mortality and the cause of death, morbidity as assessed by the number and the
nature of complications, LOSR and LOSH were noted.
Statistical analysis
The parametric, ordinal and nominal data were expressed as mean and standard
deviation (SD), median and interquartile range (IQR), and proportions with 95%
confidence interval (CI), respectively. Student's unpaired t, Mann-Whitney U and χ2
tests were used to compare the parametric, ordinal and nominal data between the
survivors and nonsurvivors. Linear regression was performed between LOSR and
LOSH on the one hand and the factors that affected mortality, DACC, surgeon and
surgical category on the other hand.
Go to:
OBSERVATIONS AND RESULTS

A total of 77 of 88 (87.5%; 95% CI: 80.6–94.6%) patients were discharged from the
hospital after the operation. Two patients expired a few months later during the period
of readmission. The data of these two patients were not analyzed. Eleven (12.5%; 95%
CI: 5.6–19.4%) patients expired during their stay in the hospital in the postoperative
period. The mean LOSR and LOSH of the discharged patients were 5.1±1.7 and
16.3±6.0 days, respectively. The mean LOSR for the patients who died was 17.2±2.7
days.
Of the 18 preoperative factors studied, only four factors were significantly different
between the alive and the expired groups. They were PTI, ASA-PS grade, CARE score
and NYHA class [Table 1]. There was no significant difference in the demographic
data, physical characteristics and surgical category in the patients between the alive and
the expired groups. Sixty-three percent of the patients had valvular, 16% congenital and
14% coronary heart disease. There was no major difference in the disease distribution
between the survivors and the nonsurvivors. Patients with lower preoperative PTI had
higher mortality (t test, P=0.039) and prolonged LOSR (linear regression, P=0.024).
However, preoperative PTI did not significantly prolong the LOSH [Tables [Tables11
and and22].
All three preoperative risk predictors, namely ASA-PS, CARE and NYHA, were good
predictors of mortality but were poor predictors of LOSR and LOSH [Table 1]. No
patient with ASA-PS grade 2 died; seven of 77 (9%) patients with ASA-PS grade 3 and
4 of six (67%) patients with ASA-PS grade 4 died. Higher ASA-PS grade (Mann-
Whitney, P=0.00027) was associated with higher mortality. Only two patients with
CARE score ≤2 died. However, eight out of 46 (17.4%) patients with CARE score 3
died. There was only one patient with CARE score of 5 who also died (100%). Higher
CARE score (Mann-Whitney, P=0.032) was associated with higher mortality. All
patients with NYHA class 1 and 2 survived, whereas seven of 25 (28%) patients with
NYHA class 3 and 4 out of six (67%) patients with NYHA class 4 died. Higher NYHA
class (Mann-Whitney, P=0.00023) was associated with higher mortality.
Of the 11 intraoperative factors analyzed, only DCPB and number of inotropes used
were significantly different between the alive and the expired groups. Seven out of 30
(23%) patients with DCPB >150 min died. Sixty-four percent (seven of 11) of the
patients who died had a DCPB ≥150 min. Long DCPB was also associated with
prolonged LOSH (linear regression, P=0.028). But, this did not predict the LOSR
[Tables [Tables22 and and3].3]. Patients requiring more number of inotropes in the
intraoperative period had higher mortality (χ2, P=0.004) and prolonged LOSR (linear
regression, P=0.005) and LOSH (linear regression, P=0.002). Seventy-five percent of
the patients (3/4) requiring more than two inotropes in the intraoperative period
ultimately died. All patients who did not need any inotropic agents in the intraoperative
period survived [Tables [Tables22 and and3].3]. DACC did not predict the mortality or
LOSR. But, long DACC was associated with significantly longer LOSH (linear
regression, P=0.021) [Table 2].
Of the 27 postoperative factors analyzed, eight parameters were significantly different

between the expired and the alive groups. Patients who died had more severe metabolic
acidosis as evidenced by lower pH (t test, P=0.014), lower bicarbonate (t test, P=0.002),
higher base deficit (t test, P=0.004), need for higher inspired oxygen concentration FiO 2
(t test, P=0.018), requirement of more number of inotropes (χ2, P=0.001), tendency
toward lower mean BP in the first 24-h postoperative period (t test, P=0.011) and
presence of a higher total leukocyte count (TLC) (t test, P=0.045) and aPTT (χ2,
P=0.022) [Table 4].
Three of the four patients requiring more than two inotropes ultimately died. All
patients who did not need inotropic agents in the postoperative period survived. Patients
who needed more number of inotropes in the postoperative period also had a prolonged
LOSR (linear regression, P=0.006) and LOSH (linear regression, P=0.013) [Table 2].
Thirty-one percent (95% CI: 8.3–53.7) (five/16) of patients with abnormal aPTT died
while only 7.4% (95% CI: 1.2–13.6) (five/68) of the patients with normal aPTT died
[Table 4].
GCS for the first postoperative day was available in 69 (92%) patients in the alive group
and in only one (9.1%) patient in the expired group. As the expired patients tended to be
more sedated, the GCS of these patients was not available. Hence, the APACHE II
score was calculated without the GCS score component (APACHE II– GCS). APACHE
II–GCS was associated with significantly higher mortality (Mann-Whitney, P=0.012)
and prolonged LOSR (linear regression, P=0.035). But, this score did not determine the
LOSH [Tables [Tables22 and and44].
The expired patients had significantly lower RR in the 24-h postoperative period
compared with the alive group (t test, P=0.001). These patients also had a significantly
prolonged LOSR (linear regression, P=0.016) [Tables [Tables22 and and44].
Three surgeons performed approximately 50, 25 and 25% of the surgeries. A total of
nine anesthesia consultants were involved. Neither the anesthesiologist nor the surgeon
influenced the mortality. However, surgeons influenced the LOSR (linear regression,
P=0.025) and LOSH (linear regression, P=0.0002) [Table 2].
Multivariate regression was performed with mortality as the dependant variable and
individually significant parameters as the independent variables. This showed no
parameter to be statistically significant. This could primarily be explained by the fact
that mortality was only 11 out of 88 patients, which is a small number to study the
effect of different factors in multivariate regressions model.
The patients who died in the postoperative period had multiple deranged parameters,
which, when studied in isolation, were significantly different in the two groups. But, to
study the effect of those factors in multivariate regression model requires a larger
sample size and strict noninterdependence of factors, which was not possible in this
study.
Go to:
DISCUSSION
Among the demographic factors, age and sex had been associated with increased
mortality in western studies, which predominantly included surgeries for coronary
artery diseases.[7–9,16] All traditional risk indices including Parsonnet,[7] Tuman[8]
and Tu[9] scores and the report published in 1999 by the American College of
Cardiology/American Heart Association (ACC/AHA) Task Force, incorporate
increasing age and female sex as risk factors.[16] Our study did not correlate age and
female sex with mortality [Table 1]. As there were only six patients with age more than
65 years, our study probably did not cover the entire spectrum of the age adequately
[Table 1]. The majority of our patients were operated for valvular heart disease. During
our study period, there was only one female patient (8.3%) who was operated for
CABG while 28 (50.8%) had valve replacement. This difference in the surgical mix
could probably explain the difference in the results.
The preoperative factors associated with higher mortality were PTI, ASA-PS grade,
CARE score and NYHA class [Table 1]. Of these, only PTI was also associated with
prolonged LOSR but not LOSH [Table 2]. Association of PTI with outcome could
probably be related to the underlying general condition of the patients. Also, two of the
expired patients received therapeutic doses of unfractionated and low molecular weight
heparin. Preoperative PTI has not been recognized and mentioned as a risk factor for
open heart surgery in the literature previously. Patients with poor general condition and
functional status, associated comorbid illness and its degree of control, and the
complexity of the surgery predicted the mortality after the surgery.[6–10,17–19] The
evidence from the literature shows that the basic drawback of the preoperative risk
indices had been their inability to predict morbidity and LOS as accurately as
mortality.[6] Our study has also exposed this drawback.
The adverse effects of prolonged DCPB and DACC are well known. Prolonged DCPB
and DACC have been associated with increased mortality, LOS in ICU, prolonged need
for mechanical ventilation, poor myocardial function, higher inotrope requirement, poor
neurological outcome and increased bleeding tendencies.[14,20–22] The intraoperative
factors associated with higher mortality in our study were DCPB and the number of
inotropes used while coming off CPB [Table 3]. The intraoperative inotrope usage was
associated with both prolonged LOSR and LOSH, while the DCPB predicted only the
LOSH [Table 2].
The postoperative factors associated with higher mortality were the APACHE II score
excluding the GCS component and the number of inotropes used [Table 4]. The
postoperative inotrope usage was associated with both prolonged LOSR and LOSH
while the APACHE II score predicted only the LOSR [Table 2]. The number of
inotropes used in the intraoperative and the postoperative periods has been associated
with mortality, LOSR and LOSH. Both the literature and our study confirm this
fact.[22] Patients requiring a large number of inotropes to maintain their hemodynamic
status tend to have poor myocardial performance due to the severity of the disease in the
preoperative period or due to intraoperative events. Their general condition and
immunity also tend to be poor. This can result in the occurrence of Multi Organ
Dysfunction Syndrome (MODS), septicemia, higher TLC, coagulation abnormalities
and metabolic acidosis in the postoperative period.[14] These patients are more heavily
sedated for the sake of better endotracheal tube tolerance, which itself can influence
their hemodynamic status. The RR of these patients was significantly lower for the
same reason and, hence, lower RR was associated with prolonged LOSR. Also, these
patients required higher FiO 2 to maintain their oxygenation. This explains the
association between several postoperative parameters including the patient's
hemodynamic status, laboratory tests and the mortality. As the patients are not weaned
from the mechanical ventilator unless their inotropic support level is reduced to an
acceptable level, the number of inotropes used intra- and postoperatively directly
influences the LOSR and LOSH [Tables [Tables22–4].
General purpose severity of illness scores such as APACHE scores were relatively
ineffective in the postoperative period because of sedation, NM blockade and elective
ventilation used in a number of these patients. Therefore, we were forced to calculate
the APACHE II score without the GCS score (APACHE II-GCS). Our results revealed
that the APACHE II-GCS score predicted the mortality and the LOSR but not the
LOSH [Tables [Tables22 and and4].4]. As the same information could be obtained by
simpler preoperative risk indices, the APACHE II score did not give us any extra
information.
Neither the anesthesiologist nor the surgeon influenced the mortality of the patients in
our study. However, surgeons influenced the LOSR and LOSH due to various patient-
related and other reasons not directly related to the patients’ status [Table 2].
There has been a well-known association between low-volume centers (<100 CABG
procedures per year) and low-volume surgeons (<50 CABG surgeries per year) and
outcome. The observed mortality for CABG surgeries in the hospitals performing >100
cases per year was 2–3.6% while that for the hospitals performing <100 cases per year
was 5%.[23] The corresponding data for other cardiac surgeries like those for VHD is
not available. On the basis of the number of CAD cases performed, our institute will
come under the low-volume center category. This could be an independent risk factor
for the outcome.
Despite the small size of the sample, this study has recognized several factors associated
with morbidity and mortality, which are consistent with the literature. Also, the main
limitation of our study is the small sample size, because of which we could not calculate
the mortality for different grades or ranges of the scoring systems used pre- and post-
operatively. Also, there was no statistically significant parameter in the multivariate
logistic regression analysis. Unlike the literature evidence, prolonged DACC was
associated with increased LOSH but not with increased mortality or LOSR in our study
[Table 2]. This is probably because of statistical chance or a small sample size. There
could also be several confounding or nonpatient-related factors that could have
influenced the mortality and, especially, LOSR and LOSH. These issues can only be
resolved by a larger, multicenter study in the future.
Go to:
CONCLUSIONS
The association of the various risk factors with the mortality and the LOSR and LOSH
following open heart surgery were analyzed. The time-tested preoperative scoring
systems like the ASA-PS grade, CARE score and NYHA class predicted the mortality
following open heart surgery well. However, they were poor predictors of morbidity as
assessed by the LOSR and the LOSH. The DCPB predicted the mortality and LOSH but
not LOSR. The DACC predicted the LOSH but not the mortality or the LOSR.
Postoperative hemodynamic and metabolic status, coagulation status and TLC were
associated with mortality. The intra- and pos-toperative inotrope requirement predicted
the mortality, LOSR and LOSH. The postoperative APACHE II score was difficult to
estimate due to the nonavailability of GCS in these sedated patients. The derived score
obtained by calculating the APACHE II score without the GCS score component
(APACHE II–GCS) predicted the mortality and the LOSR but not the LOSH. The
identities of the anesthesiologist and the surgeon were not associated with the mortality.
The surgeon influenced the LOSR and the LOSH.
Go to:
Footnotes
Source of Support: Nil
Conflict of Interest: None declared.
Go to:
REFERENCES
1. Knaus WA, Wagner DP, Draper EA, Zimmerman JE, Bergner M, Bastos PG, et al. The APACHE
III prognostic system.Risk prediction of hospital mortality of critically ill hospitalized adults.
Chest. 1991;100:1619–36. [PubMed: 1959406]
2. Le Gall JR, Loirat P, Alperovitch A, Glaser P, Granthil C, Mathieu D, et al. A simplified acute
physiology score for ICU patients. Crit Care Med. 1984;12:975–7. [PubMed: 6499483]
3. Le Gall JR, Lemeshow S, Saulnier F. A new simplified acute physiology score (SAPS II) based
on a European/ North American multicenter study. JAMA. 1993;270:2957–63. [PubMed:
8254858]
4. Lemeshow S, Teres D, Pastides H, Avrunin JS, Steingrub JS. A method for predicting survival
and mortality of ICU patients using objectively derived weights. Crit Care Med. 1985;13:519–
25. [PubMed: 4006490]
5. Lemeshow S, Teres D, Klar J, Avrunin JS, Gehlbach SH, Rapoport J. Mortality probability
Models (MPM II) based on an international cohort of intensive care unit patients. JAMA.
1993;270:2478–86. [PubMed: 8230626]
6. Dupuis JY, Wang F, Nathan H, Lam M, Grimes S, Bourke M. The cardiac anesthesia risk
evaluation score: A clinically useful predictor of mortality and morbidity after cardiac surgery.
Anesthesiology. 2001;94:194–204. [PubMed: 11176081]
7. Parsonnet V, Dean D, Bernstein AD. A method of uniform stratification of risk for evaluating
the results of surgery in acquired adult heart disease. Circulation. 1989;79(suppl I):3–12.
8. Tuman KJ, McCarthy RJ, March RJ, Najafi H, Ivankovich AD. Morbidity and duration of ICU
stay after cardiac surgery.A model for pre-operative risk assessment. Chest. 1992;102:36–44.
[PubMed: 1623792]
9. Tu JV, Jaglal SB, Naylor CD. Multicenter validation of a risk index for mortality, intensive care
unit stay, and overall hospital length of stay after cardiac surgery. Steering committee of the
Provincial Adult Care Network of Ontario. Circulation. 1995;91:677–84. [PubMed: 7828293]
10. Parolari A, Pesce LL, Trezzi M, Cavallotti L, Kassem S, Loardi C, et al. EuroSCORE
performance in valve surgery: A meta-analysis. Ann Thorac Surg. 2010;89:787–93. [PubMed:
20172129]
11. Turner JS, Morgan CJ, Thakrar B, Pepper JR. Difficulties in predicting outcome in cardiac
surgery patients. Crit Care Med. 1995;23:1843–50. [PubMed: 7587260]
12. Turner JS, Mudaliar YM, Chang RW, Morgan CJ. Acute physiology and chronic health
evaluation (APACHE II) scoring in a cardiothoracic intensive care unit. Crit Care Med.
1991;19:1266–9. [PubMed: 1914483]
13. Martinez-Alario J, Tuesta ID, Plasencia E, Santana M, Mora ML. Mortality prediction in
cardiac surgery patients: Comparative performance of Parsonnet and general severity systems.
Circulation. 1999;99:2378–82. [PubMed: 10318657]
14. Kollef MH, Wragge T, Pasque C. Determinants of mortality and multiorgan dysfunction in
cardiac surgery patients requiring prolonged mechanical ventilation. Chest. 1995;107:1395–
401. [PubMed: 7750337]
15. Au WK, Sun MP, Lam KT, Cheng LC, Chiu SW, Das SR. Mortality prediction in adult cardiac
surgery patients: Comparison of two risk stratification models. Hong Kong Med J.
2007;13:293–7. [PubMed: 17664534]
16. Eagle KA, Guyton RA, Davidoff R, Ewy GA, Fonger J, Gott JP, et al. ACC/AHA Guidelines for
coronary artery bypass surgery; Executive summary and recommendations (revision of 1991
guidelines) Circulation. 1999;100:1464–80. [PubMed: 10500052]
17. Jackson JM, Thomas SJ. Valvular heart disease. In: Kaplan JA, Reich DL, Konstadt SN,
editors. Cardiac Anaesthesia. 4th ed. Philadelphia, PS: W.B. Saunders Company; 1999. p. 727.
18. Nashef SA, Roques F, Hammill BG, Peterson ED, Michel P, Wyse RK, et al. Validation of
European System for Cardiac Operative Risk Evaluation (EuroSCORE) in North American cardiac
surgery. Eur J Cardiothorac Surg. 2002;22:101–5. [PubMed: 12103381]
19. Michel JP, Klopfenstein C, Hoffmeyer P, Stern R, Grab B. Hip fracture surgery: Is the pre-
operative American Society of Anesthesiologists (ASA) score a predictor of functional
outcome? Aging Clin Exp Res. 2002;14:389–94. [PubMed: 12602574]
20. Floyd RD, Sabiston DC, Jr, Lee KL, Jones RH. The effect of duration of hypothermic
cardioplegia in ventricular function. J Thorac Cardiovasc Surg. 1983;85:606–11. [PubMed:
6601216]
21. Harker LA, Malpass TW, Branson HE, Hessel EA, 2nd, Slichter SJ. Mechanism of abnormal
bleeding in patients undergoing cardiopulmonary bypass: Acquired transient platelet
dysfunction associated with selective alpha-granule release. Blood. 1980;56:824–34. [PubMed:
6448643]
22. Hugot P, Sicsic JC, Schaffuser A, Sellin M, Corbineau H, Chaperon J, et al. Base deficit in
immediate postoperative period of coronary surgery with cardiopulmonary bypass and length
of stay in intensive care unit. Intensive Care Med. 2003;29:257–61. [PubMed: 12541160]
23. Clark RE. Outcome as a function of annual coronary artery bypass graft volume. The Ad Hoc
Committee on Cardiac Surgery Credentialing of the Society of Thoracic Surgeons. Ann Thorac
Surg. 1996;61:21–6. [PubMed: 8561556]
Go to:
Figures and Tables

Table 1
Preoperative data
Table 2
Effect of parameters associated with significant increase in mortality on LOS in the
recovery room and in the hospital
Table 3
Intraoperative data
Table 4
Postoperative data
Articles from Saudi Journal of Anaesthesia are provided here courtesy of Medknow
Publications
Analysis of immediate results of on-pump versus

off-pump coronary artery bypass grafting surgery
Análise dos resultados imediatos da cirurgia de revascularização do miocárdio com e sem circulação
extracorpórea
Marcos Antonio Cantero1, Rui M. S. Almeida2, Roberto Galhardo1
DOI: 10.5935/1678-9741.20120007 RBCCV 44205-1348
Abstract off-pump group had fewer complications in relation to

Objective: The objective of this study is to compare the perioperative myocardial infarction (P = 0.02) and use of
immediate results of patients undergoing on-pump versus intra-aortic balloon pump (P = 0.01).
off-pump coronary artery bypass graft (CABG) surgery. Conclusion: The off-pump CABG is a safe procedure
Methods: From January 2007 to January 2009, 177 with hospital mortality similar to that observed in on-pump
patients underwent CABG. Of these, 92 underwent off-pump CABG, with lower rates of complications and less need for
CABG and 85 on-pump CABG. We evaluated the intra-aortic balloon.
demographics, preoperative risk factors, preoperative
functional class, and risk assessment by the EuroSCORE. A Descriptors: Extracorporeal circulation. Cardiopulmonary
comparison between both groups regarding the postoperative bypass. Coronary artery disease. Myocardial
evolution was carried out as well. revascularization.
Results: The mean number of grafts per patient was 2.48
± 0.43 in the off-pump group versus 2.90 ± 0.59 in the on-
pump group. In the off-pump group, 97.8% of patients Resumo
received an internal thoracic artery graft, while in the on- Objetivo: Comparar os resultados imediatos da cirurgia
pump group, the percentage was 94.1% (P = 0.03). The rate de revascularização do miocárdio com e sem circulação
of complete revascularization was similar in both groups. extracorpórea (CEC).
In the off-pump group, the circumflex artery (circumflex Métodos: De janeiro de 2007 a janeiro de 2009, 177
branch of the left coronary artery) was revascularized in pacientes foram submetidos a cirurgia de revascularização
48.9% of the patients versus 68.2% of the patients in the on- do miocárdio (CRM), sendo 92, sem CEC e 85 com CEC.
pump group (P = 0.01). Hospital mortality was 4.3% for off- Foram avaliados distribuição demográfica, fatores de risco
pump CABG and 4.7% for on-pump CABG (P = 0.92). The pré-operatórios, classe funcional e avaliação de risco pelo
1. Cardiovascular Surgeon; Hospital do Coração de Dourados e Corresponding author

Hospital Evangélico Sr e Sra Goldsby King, Dourados, MS, Brasil. Marcos Antonio Cantero. Rua Delmar de Oliveira, 1725
2. Doctorate degree; Associate Professor; Universidade Estadual do Vila Progresso –Dourados, MS
Oeste do Paraná (UNIOESTE); Coordinator of the Medical Course Brasil – Zip Code: 79825-115
at Faculdade Assis Gurgacz (FAG); Member of the Deliberative E-mail: marcoscantero@sbccv.org.br
Council of the Brazilian Society of Cardiovascular Surgery and
the Member of the Editorial Council of the Brazilian Journal of
Cardiovascular Surgery, Dourados, MS, Brasil.
This study was carried out at Hospital Evangélico Sr. e Sra. Goldsby
King Dourados - MS Hospital do Coração de Dourados, Dourados, Article received on July 18th, 2011
MS, Brasil. Article accepted on December 12th, 2011
38
Cantero MA, et al. - Analysis of immediate results of on-pump versus Rev Bras Cir Cardiovasc 2012;27(1):38-44
de artéria torácica interna, enquanto que no grupo com CEC

Abbreviations, acronyms & symbols
a porcentagem foi de 94,1% (P = 0,03). A taxa de
MR Myocardial revascularization revascularização completa foi similar em ambos os grupos.
ECC Extracorporeal Circulation No grupo sem CEC, a artéria circunflexa foi revascularizada
FC Functional class em 48,9% dos casos e, em 68,2%, no grupo com CEC (P =
EuroSCORE European System for Cardiac Operative Risk 0,01). A mortalidade hospitalar foi de 4,3% e 4,7%,
Evaluation respectivamente, no grupo sem CEC e com CEC (P = 0,92).
NYHA New York Heart Association Os pacientes operados sem CEC apresentaram menor índice
MAP Mean arterial pressure de complicações em relação ao infarto perioperatório (P=
CVP Central venous pressure
0,02) e ao uso de balão intra-aórtico (P= 0,01).
STS Society of Thoracic Surgeons
ACT Activating clotting time Conclusão: A cirurgia coronariana sem CEC é um
procedimento seguro, com mortalidade hospitalar similar a
dos pacientes operados com CEC, com menores taxas de
EuroSCORE. A evolução no pós-operatório foi comparada complicações e de incidência de infarto perioperatório, bem
entre os grupos. como menor necessidade de balão intra-aórtico.
Resultados: A média de enxertos por paciente foi de 2,48
± 0,43, no grupo sem CEC, e 2,90 ± 0,59, no com CEC. No Descritores: Circulação extracorpórea. Ponte cardiopulmonar.
grupo sem CEC, 97,8% dos pacientes receberam um enxerto Doença da artéria coronariana. Revascularização miocárdica.
INTRODUCTION the countryside of Brazil, the objective of this study was to

evaluate the immediate results of patients undergoing on-
During the past decades, coronary artery by-pass graft pump and off-pump CABG. We analyzed the demographics
surgery (CABG) has allowed patients with coronary of the population and the differences in morbidity and
atherosclerotic disease to improve survival, symptoms, and mortality rates at the Cardiovascular Surgery Service of the
quality of life [1]. From the mid-1990s, efforts focused on Hospital do Coração de Dourados, Mato Grosso do Sul,
ways to reduce complications and make CABG less invasive. Brazil.
Cardiopulmonary bypass (CPB) induces the systemic
inflammatory response through activation of the METHODS
complement system, mainly via the alternative pathway
induced by blood contact with the surface of the According to current guidelines, we screened 177
extracorporeal circuit, which triggers the release of patients with multiarterial coronary artery disease
inflammatory mediators such as the interleukin 1, interleukin (insufficiency), who had surgical indication to CABG from
6, and tumor necrosis factor responsible for the systemic January 2007 to January 2009 at the Hospital do Coração
inflammatory response. In an attempt to reduce the systemic de Dourados (Dourados, MS, Brazil). Of these, 92 patients
inflammatory response, the off-pump CABG has been underwent off-pump CABG surgery (Group 1) and 85
rediscovered and refined. In 1964, Kolesov performed the patients underwent on-pump CABG (Group 2).
first off-pump CABG in Leningrad [2]. Files were retrieved from hospital registry and reviewed
This technique, after the initial experiments [3] has been retrospectively through review of the medical records and
revived by Buffolo et al. [4] and Benetti et al. [5]. Since evaluation of medical examination performed preoperatively.
then, it has been recommended as a primary treatment option The study was approved by the Local Ethical Committee
for high-risk patients [6]. (committee report nº 004/2009).
Since 1990, the experience with off-pump CABG has Patients were included in the study by an agreement
increased. In 1999, data from the Society of Thoracic between the surgeons provided that the revascularization
Surgeons (STS) reveal that it represented about 10% of the might be performed reliably and in a similar manner for both
total CABG surgeries performed in the United States [7]. operative techniques. Randomization assignment was not
Since 2001, the number of all surgical revascularization provided. After a detailed explanation of the investigative
procedures in the United States increased up to 25%. purpose of the study and the results already obtained in
According to STS, this proportion was about 20% until other services, written informed consent was obtained from
2007 [9]. all participants. Exclusion criteria were the presence of
In an attempt to evaluate and demonstrate that the off- cardiogenic shock or mechanical complications of infarction,
pump CABG is feasible in our country, with results similar ejection fraction changes (<55%), and non-acceptance of
to those found in the literature, even in a service located in the method by the patient.
39
The following variables were included in the study: Patients who underwent off-pump CABG received
1. Age, type of operation (first operation or reoperation); heparin (2 mg/kg) after induction of anesthesia and
2. Clinical stratification of the heart functional class (FC) harvesting of grafts. In both techniques, heparinization was
according to the New York Heart Association (NYHA); controlled by the activated clotting time (ACT). We have
3. The presence of risk factors, such as systemic placed a point with Ethibond 2-0 attached to a cotton strip
hypertension, smoking, dyslipidemia, diabetes mellitus, with 3 cm in the pericardial deflection between the inferior
chronic obstructive pulmonary disease, and peripheral vena cava and right inferior pulmonary vein in order to
arterial insufficiency; expose completely the heart. Distal anastomoses were
4. Occurrence of complications, such as stroke, performed, and the artery occluded proximally to the
perioperative acute myocardial infarction, presence of anastomosis with a 5-0 polypropylene thread point
ventricular and atrial arrhythmias, need for mechanical anchored in Teflon pledge. The area in which the
ventilation > 24 hours, use of intra-aortic balloon pump, anastomosis was being performed was exposed and
surgical bleeding indicating exploratory mediastinotomy stabilized with a suction stabilizer (The Medtronic
and death. Octopus® System). At the completion of the distal
The study data were set with the risk index developed anastomoses, the systolic blood pressure was maintained
by the European Association for Cardiothoracic Surgery, at 100 mmHg. The aorta was partially clamped, and the
the European System for Cardiac Operative Risk Evaluation proximal anastomoses performed. Upon completion of the
(EuroSCORE) [8]. The model database used by STS [9] was anastomoses, heparin was reversed with protamine sulfate
not used due to the lack of some data on all medical records. in both groups, and the operation was completed.
After the initial trial period of service that began the off-
pump CABG in mid-2004, all patients were operated on by Statistical Analysis
the same surgeon and surgical team. In this study, mean age and number of vessels treated
Trans- and postoperative monitoring included were compared by the Student’s t test. Other variables were
continuous electrocardiogram with electrodes placed on analyzed using the Chi-square test. A p-value of less than
the posterior (dorsal) surface, mean arterial pressure (MAP) 0.05 was considered statistically significant.
through peripheral artery catheterization, central venous
pressure (CVP) by placing a double lumen catheter into the RESULTS
vena cava, pulse oximetry by placing a digital sensor,
temperature by using an esophageal thermometer, and The mean age of the patients was 63.4 ± 8.8 years in the
urinary output. Anesthesia was induced with fentanyl (5 on-pump CABG group versus 63.0 ± 9.6 years in the off-
mcg/kg) and etomidate (0.3 mg/kg) followed by pump CABG group, with a range of 29-87 years. There were
neuromuscular blocking agents to facilitate tracheal 75 men (81.6%) in the off-pump CABG and 50 men (59.4%)
intubation, or pancuronium (0.1 mg/kg intravenous bolus, in the on-pump CABG. Regarding other demographic data,
and 0.03 mg/kg in maintenance doses). Maintenance was there were no statistically significant risk factors (Table 1).
performed with sufentanil (0.02 mg/kg/min), midazolam
bolus dose depending on requirements, and pancuronium
(0.03 mg/kg/h). Both inhalation anesthetics and halogens Table 1. Preoperative risk factors and complications in both
in combination with nitrous oxide (N2O) have also been groups.
used. Vasopressors and inotropic support was administered OPCABG (92) ONCABG (85) P value
after the onset of mobilization of the heart aiming at Age 63.0 ± 9.6 63.4 ± 8.8 0.38
appropriate organic and tissue perfusion. Ejection fraction 62.3 ± 15 59.4 ± 16.3 0.53
Patients who underwent on-pump CABG received Male 81.6% 80.3% 0.82
Hypertension 56.3% 51.7% 0.51
heparin (3 mg/kg) after induction of anesthesia and
Smoking 52.7% 54.9% 0.52
harvesting of grafts. Patients underwent a median
Dyslipidemia 70.1% 68.3% 0.63
sternotomy, cardiopulmonary bypass established by Diabetes 23.7% 28.2% 0.12
cannulation of the ascending aorta and right atrium, Previous Stroke 4.8% 5.3% 0.82
cannulation of the right superior pulmonary vein with PAI 11.9% 15.3% 0.52
introduction of a catheter for aspiration and decompression COPD 10.4% 8.7% 0.13
of the left ventricle, hypothermia at 28°C, isothermic blood Renal failure 1% 2.3% 0.76
cardioplegia delivered in an antegrade manner at a ratio of Previous Operation 2.1% 1.1% 0.97
1:4, followed by distal anastomoses. The proximal CPB = cardiopulmonary bypass; PAI = peripheral arterial
anastomoses were performed with partial clamping of the insufficiency; COPD = Chronic obstructive pulmonary disease
aorta and on a beating heart.
40
Both groups showed no differences in preoperative Obstructions in the circumflex branch of the left coronary
NYHA FC (NYHA FC I: Group 1 (3.2%) vs. Group 2 (3.5%); artery were significant (> 50%) on a lesser ratio in the off-
NYHA FC II: off-pump CABG (32.6%) vs. on-pump CABG pump CABG group (P = 0.06). The right coronary artery
(31.8%); NYHA FC III: off-pump CABG (52.1%) vs. on-pump was affected in 85.8% of patients in the off-pump CABG
CABG (52.9%); NYHA FC IV: off-pump CABG (11.9%) vs. group and in 80% of the patients in the on-pump CABG
on-pump CABG (11.7%). Elective surgery was required in group (P = 0.61).
56.5% of patients in the off-pump CABG group (52 cases) The mean number of grafts per patient was 2.48 ± 0.43 in
vs. 60% in the on-pump CABG group (P = 0.70). Considering the off-pump CABG group versus 2.90 ± 0.59 in the on-
the mortality rate logistic EuroSCORE, seven patients in pump CABG group (P = 0.02). The number of grafts
the off-pump CABG group were considered low-risk performed (Table 3) ranged from 1 to 6 with a higher
patients (score 0-2), 27 medium-risk (score 3-5), and 57 high- proportion of patients with one graft (14.1% vs. 2.3%, P =
risk (score >6). Six patients in the on-pump CABG group 0.001) and two grafts (35.8% vs. 24.7%, P = 0.03) in the off-
were considered low-risk patients (score 0-2), 25 medium- pump CABG group. However, the on-pump CABG group
risk (score 3-5), and 54 high-risk (score >6). had more patients with three grafts (58.8% vs. 40.2%, P =
The extent of coronary artery disease also showed no 0.004). In off-pump CABG group, 97.8% of patients received
significant differences between both groups (11 patients internal thoracic artery bypass-graft versus 94.1% in the
(11.9%) with one impaired vessel in the off-pump CABG on-pump CABG group (P = 0.03). The rate of complete
group vs. two patients (2.3%) in the on-pump CABG group revascularization was similar in both groups (69.5% in the
(P = 0.11)). Twenty-five patients (27.1%) had two-vessel off-pump CABG group vs. 67.0% in the on-pump CABG
coronary artery disease in the off-pump CABG group vs. group, P = 0.68). The anterior interventricular branch of left
22 patients (25.8%) in the on-pump CABG group (P = 0.98). coronary artery was revascularized in 92.3% of patients in
Triple-vessel disease was present in 56 patients (60.8%) in the off-pump CABG group, the right coronary artery in
the off-pump CABG group vs. 61 patients (71.7%) in the 54.3% and the circumflex branch in 48.9%; in the on-pump
on-pump CABG group (P = 0.37). CABG group, these proportions were 90.5%, 67.0%, and
The analysis of the type of coronary lesions showed no 68.2% respectively. There was a statistically significant
differences between the two groups. Thus, 97.8% of the difference in the amount of grafts using the circumflex branch
patients in the off-pump CABG group had critical injuries of left coronary artery in 48.9% of the patients in the off-
in the anterior interventricular branch of the left coronary pump CABG group versus 68.2% of the patients in the on-
artery vs 98.8% in the on-pump CABG group (P = 0.84). pump CABG group (P = 0.02).
Table 2. NYHA functional class according to the both groups. Table 4. Immediate morbidity and mortality in both groups.
NYHA class OPCABG ONCABG OPCABG ONCABG P-value
n (%) n (%) (n = 92) (n = 85)
I 3 (3.2%) 3 (3.5%) In-hospital mortality 4.3% 4.7% 0,89
II 30 (32.6%) 27 (31.8%) Perioperative AMI 7.6% 12.9% 0,02
III 48 (52.1%) 45 (52.9%) IAB pump use 3.2% 14.1% 0,01
IV 11 (11.9%) 10 (11.7%) Ventricular arrhythmia 2.2% 3.5% 0,84
Atrial Fibrillation 12% 12.9% 0,16
NYHA = New York Heart Association; CABG = coronary artery
Ventilation > 24 h 5.4% 11.7% 0,14
bypass grafting. All P-value were higher than 0.05
Re-intervention 4.3% 4.7% 0,35
CABG = coronary artery by-pass graft. IAB = intra-aortic balloon
pump. AMI = acute myocardial infarction
Table 3. Number of coronary grafts performed in both groups.

Number of Grafts OPCABG (n = 92) ONCABG (n = 85)
Nº of Cases % Nº of Cases % The proportion of grafts in the anterior interventricular
1 graft 13 14.1% 2 2.3% branch of left coronary artery and in the right coronary
2 grafts 33 35.8% 21 24.7% artery was similar in both groups. The conversion rate to
3 grafts 37 40.2% 50 58.8% on-pump CABG was 5.4% (five cases). Hospital mortality
4 grafts 6 6.5% 8 9.4% in the off-pump CABG was 4.3% versus 4.7% in the on-
5 grafts 3 3.2% 3 3.5% pump CABG group (P = 0.92).
6 grafts __ _ 1 1.1% The most common complication was atrial fibrillation,
CABG = coronary artery by-pass grafting which occurred in 12.9% of the patients in the on-pump
41
CABG group versus 12% of the patients in the off-pump as 30-day mortality, myocardial infarction, stroke, atrial
CABG group. fibrillation, and acute renal failure
Among other complications (Table 4), the least frequent In randomized clinical trials, off-pump CABG surgery
in the off-pump CABG group was the perioperative was not associated with any significant reduction in 30-
infarction rate (7.6% versus 16.4% in the on-pump group day mortality and myocardial infarction. It showed a relevant
(Group 2) (P = 0.04)). The need for intra-aortic balloon pump reduction in the incidence of stroke and atrial fibrillation.
was 3.2% in Group 1 vs. 11.7% in Group 2 (P = 0.01). It is There was no significant reduction in acute renal failure.
noteworthy that the groups were similar in risk score, The benefits of the off-pump CABG surgery in the
ventricular function and NYHA FC. Therefore, the groups elderly [19], in patients undergoing hemodialysis [20] and,
were homogeneous, although there was no randomization. lately, in females [21] were shown in several subgroups of
patients. In Brazil, other services have recorded their
DISCUSSION experiences demonstrating that myocardial
revascularization without cardiopulmonary bypass is a
The literature worldwide shows that off-pump CABG procedure that can be performed with low surgical risk
has been a viable option for the treatment of severe coronary and with excellent results [22]. It is considered as an
insufficiency [10,11]. independent protective factor for some complications
A systemic inflammatory response can be caused by such as mediastinitis [23] and the need for blood
platelet degranulation, activation of neutrophils and transfusion [24].
monocytes, and release of cytokines, thus contributing to There was no difference between groups in the rate of
cardiac dysfunction after CPB. The inflammatory response complications. Analyzing the incidence of perioperative
impairs lung function; CPB adds lung injury and delays the infarction in our sample, we found a higher incidence in on-
recovery of respiratory function [12]. Several studies [13] pump (12.9%) versus off-pump (7.6%) CABG surgery. These
compared the inflammatory response with and without data are similar to those found by Demers et al. [25], who
cardiopulmonary bypass by measuring serum reported 5.1% of acute myocardial infarction post-CPB
concentrations of cytokines and acute-phase proteins versus 2.0% without CPB. Lima et al. [26] also reported
before and after surgery. There was a significant attenuation 8.0% and 4.3%, respectively.
of the inflammatory response during cardiopulmonary The need for intra-aortic balloon pump shows a
bypass. With the reduction of inflammatory response, the statistically significant reduction in the off-pump CABG
pathophysiological analysis may reduce organ dysfunction, group (3.2%) compared to patients undergoing on-pump
which makes off-pump surgery less harmful. CABG surgery (14.1%) (P = 0.01). There was no difference
The risks of CABG have increased in recent years due between groups in the NYHA FC, in left ventricular function,
to patients’ older age, the greater number of patients and risk score, making them homogeneous, although they
undergoing prior angioplasty and also by the expansion were not randomized.
of indications for certain groups of patients, especially The off-pump CABG surgery allows a complete
those with severe ischemic cardiomyopathy and revascularization rate similar to that of patients undergoing
comorbidities. Such conditions confirmed the distribution on-pump CABG, as well as a percentage of use of left internal
of patients in EuroSCORE, in which most of them are in thoracic artery superior in Group 1, which may be due to
the range of high risk. the need to avoid manipulation of the ascending aorta. The
A randomized meta-analysis (ROOBY Trial) showed that number of grafts per patient, however, was lower in the off-
off-pump CABG surgery was associated with worse pump CABG group. The statistically significant decrease
outcomes and lower graft patency [14]. Observational of the grafts performed (Table 3) using the circumflex branch
studies have already suggested similar results to of the left coronary artery in the off-pump CABG group
conventional CABG [15]. The effectiveness of off-pump should be hold responsible for this result.
CABG surgery has been demonstrated in patients with As in the ROOBY study [14], our study did not show
multivessel disease [22], or in those with disease in the left significant difference in mortality. Data were confirmed by
coronary artery [16], and in high-risk patients preoperatively three large meta-analyses [27]. It was reported a mortality
as well [17]. rate after off-pump CABG similar to the on-pump CABG.
The lack of a clear benefit in clinical trials that compared
patients who underwent off-pump versus on-pump CABG, THE LIMITATIONS OF THE STUDY
led to the meta-analysis of 22 observational studies and 37
randomized clinical trials [18]. This study has several limitations requiring caution in
In observational studies, off-pump CABG surgery was their interpretation:
associated with significant reductions in all points, such 1) There are no adjustments for specific risk attributed
42
to clinical characteristics. It may have biased the choice of 7. Aldea GS, Mokadam NA, Melford R Jr, Stewart D, Maynard
a particular patient to the surgical procedure; C, Reisman M, et al. Changing volumes, risk profiles, and
2) The groups are not randomised, nor are the outcomes of coronary artery bypass grafting and
prospective analysis. This compromises the conclusion, percutaneous coronary interventions. Ann Thorac Surg.
2009;87(6):1828-38.
to a certain extent;
3) The primary endpoints could not be outlined in 8. Nashef SA, Roques F, Michel P, Gauducheau E, Lemeshow S,
advance, once the analysis is based only on the database. Salamon R. European system for cardiac operative risk
evaluation (EuroSCORE). Eur J Cardiothorac Surg.
CONCLUSION 1999;16(1):9-13.
The off-pump CABG surgery is a safe procedure with a 9. Clarke RE. The STS Cardiac Surgery National Database: an
mortality rate similar to that of the on-pump CABG surgery, update. Ann Thorac Surg. 1995;59(6):1376-80.
with a lower incidence of complications and perioperative
10. Tang AT, Knott J, Nanson J, Hsu J, Haw MP, Ohri SK. A
infarction, and less need of the intra-aortic balloon pump.
prospective randomized study to evaluate the renoprotective
The technique is feasible, with similar results even in small action of beating heart coronary surgery in low risk patients.
service facilities. However, this study lacks statistical power, Eur J Cardiothorac Surg. 2002;22(1):118-23.
and it has some biases that hamper this statement from
being consistent. 11. Patel NC, Graysson AD, Jackson M, Au J, Yonan N, Hasan R,
et al; North West Quality Improvement Program in Cardiac
Interventions. The effect off-pump coronary artery bypass
surgery on in-hospital mortality and morbidity. Eur J
Cardiothorac Surg. 2002;22(2):255-60.
12. Taggart DP, el-Fiky M, Carter R, Bowman A, Wheatley

DJ. Respiratory dysfunction after uncomplicated
cardiopulmonary bypass. Ann Thorac Surg.
1993;56(5):1123-8.
REFERENCES
13. Diegeler A, Doll N, Rauch T, Haberer D, Walther T, Falk V, et
1. Yusuf S, Zucker D, Peduzzi P, Fisher LD, Takaro T, Kennedy al. Humoral immune response during coronary artery bypass
JW, et al. Effect of coronary artery bypass graft surgery on grafting: a comparison of limited approach, “off-pump”
survival: overview of 10-year results from randomised trials technique, and conventional cardiopulmonary bypass.
by the Coronary Artery Bypass Graft Surgery Trialist Circulation. 2000;102(19 Suppl 3):III95-100.
Collaboration. Lancet. 1994;344(8922):563-70.
14. Shroyer AL, Grover FL, Hattler B, Collins JF, McDonald GO,
2. Kolessov VI. Mammary artery-coronary artery anastomosis Kozora E, et al; Veterans Affairs Randomized On/Off Bypass
as method of treatment for angina pectoris. J Thorac Cardiovasc (ROOBY) Study Group. On-pump versus off-pump coronary-
Surg. 1967;54(4):535-44. artery bypass surgery. N Engl J Med. 2009;361(19):1827-37.
3. Favaloro RG. Saphenous vein autograft replacement of severe 15. Cartier R, Brann S, Dagenais F, Martineau R, Couturier AJ.
segmental coronary artery occlusion: operative technique. Ann Systematic off-pump coronary artery revascularization in
Thorac Surg. 1968;5(4):334-9. multivessel disease: experience of three hundred cases. J Thorac
Cardiovasc Surg. 2000;119(2):221-9.
4. Buffolo E, Andrade JC, Branco JN, Aguiar LR, Ribeiro EE,
Jatene AD. Myocardial revascularization without 16. Dewey TM, Magee MJ, Edgerton JR, Mathison M, Tennison
extracorporeal circulation. Seven-year experience in 593 cases. D, Mack MJ. Off-pump bypass grafting is safe in patients
Eur J Cardiothorac Surg. 1990;4(9):504-7. with left main coronary disease. Ann Thorac Surg.
2001;72(3):788-91.
5. Benetti FJ, Naselli G, Wood M, Geffner L. Direct myocardial
revascularization without extracorporeal circulation. Experience 17. Arom KV, Flavin TF, Emery RW, Kshettry VR, Janey PA,
in 700 patients. Chest. 1991;100(2):312-6. Petersen RJ. Safety and efficacy of off-pump coronary artery
bypass grafting. Ann Thorac Surg. 2000;69(3):704-10.
6. Al-Ruzzeh S, Nakamura K, Athanasiou T, Modine T, George
S, Yacoub M, et al. Does off-pump coronary artery bypass 18. Wijeysundera DN, Beattie WS, Djaiani G, Rao V, Borger MA,
(OPCAB) surgery improve the outcome in high-risk patients? Karkouti K, et al. Off-pump coronary artery surgery for reducing
A comparative study of 1398 high-risk patients. Eur J mortality and morbidity: meta-analysis of randomized and
Cardiothorac Surg. 2003;23(1):50-5. observational studies. J Am Coll Cardiol. 2005;46(5):872-82.
43
19. Silva AMRP, Campagnucci VP, Pereira WL, Rosa RF, Franken bypass grafting surgery. Rev Bras Cir Cardiovasc.
RA, Gandra SMA, et al. Revascularização do miocárdio sem 2011;26(1):27-35.
circulação extracorpórea em idosos: análise da morbidade e
mortalidade. Rev Bras Cir Cardiovasc. 2008;23(1):40-5. 24. Sá MP, Soares EF, Santos CA, Figueiredo OJ, Lima RO, Rueda
FG, et al. Predictors of transfusion of packed red blood cells in
20. Milani R, Brofman PRS, Souza JAM, Barboza L, Guimarães coronary artery bypass grafting surgery. Rev Bras Cir
MR, Barbosa A, et al. Revascularização do miocárdio sem Cardiovasc. 2011;26(4):552-8.
circulação extracorpórea em pacientes submetidos à hemodiálise.
Rev Bras Cir Cardiovasc. 2007;22(1):104-10. 25. Demers P, Cartier R. Multivessel off-pump coronary artery
bypass surgery in the elderly. Eur J Cardiothorac Surg.
21. Sá MPBO, Lima LP, Rueda FG, Escobar RR, Cavalcanti PEF,
2001;20(5):908-12.
Thé ECS, et al. Estudo comparativo entre cirurgia de
revascularização miocárdica com e sem circulação extracorpórea
26. Lima RC, Diniz R, Césio A, Vasconcelos F, Gesteira M,
em mulheres. Rev Bras Cir Cardiovasc. 2010;25(2):238-44.
Menezes AM, et al. Revascularização miocárdica em pacientes
22. Lima RC, Escobar M, Wanderley Neto J, Torres LD, Elias octogenários: estudo retrospectivo e comparativo entre
DO, Mendonça JT, et al. Revascularização do miocárdio sem pacientes operados com e sem circulação extracorpórea. Rev
circulação extracorpórea: resultados imediatos. Rev Bras Cir Bras Cir Cardiovasc. 2005;20(1):8-13.
Cardiovasc. 1993;8(3):171-6.
27. van der Heijden GJ, Nathoe HM, Jansen EW, Grobbee DE.
23. Sá MP, Soares EF, Santos CA, Figueiredo OJ, Lima RO, Escobar Meta-analysis on the effect of off-pump coronary bypass
RR, et al. Risk factors for mediastinitis after coronary artery surgery. Eur J Cardiothorac Surg. 2004;26(1):81-4.
44
Age influences outcomes in 70-year or older

patients undergoing isolated coronary artery
bypass graft surgery
A idade influencia os desfechos em pacientes com idade igual ou superior a 70 anos submetidos à
cirurgia de revascularização miocárdica isolada
Antônio Sérgio Cordeiro da Rocha1, Felipe José Monassa Pittella2, Andrea Rocha de Lorenzo3,
Valmir Barzan4, Alexandre Siciliano Colafranceschi5, José Oscar Reis Brito6, Marco Antonio de
Mattos7, Paulo Roberto Dutra da Silva8
DOI: 10.5935/1678-9741.20120008 RBCCV 44205-1349
Abstract comprised 257 (24.8%) patients G2 776 (75.2%). Patients in

Objective: To analyze the results of isolated on-pump G1 were more likely to have in-hospital mortality than in
coronary artery bypass graft surgery (CABG) in patients ≥ G2 (8.9% vs. 3.6%, respectively; P=0.001), while the
70 years old in comparison to patients <70 years old. incidence of AMI was similar (5.8% vs. 5.5%; P=0.87) in G2.
Methods: Patients undergoing isolated CABG were More patients in G1 had re-exploration for bleeding (12.1%
selected for the study. The patients were assigned into two vs. 6.1%; P=0.003). Compared to G2, G1 had more incidences
groups: G1 (age ≥ 70 years old) and G2 (age <70 years old). of respiratory complications (21.4% vs. 9.1%; P<0.001),
The endpoints were in-hospital mortality, acute myocardial mediastinitis (5.1% vs. 1.9%; P=0.013), stroke (3.9% vs.
infarction (AMI), stroke, re-exploration for bleeding, 1.3%; P=0.016), acute renal failure (7.8% vs. 1.3%; P<0.001),
intraaortic balloon pump for circulatory shock, respiratory sepsis (3.9% vs. 1.9%;P=0.003), atrial fibrillation (15.6% vs.
complications, acute renal failure, mediastinitis, sepsis, atrial 9.8%; P=0.016), and CAVB (3.5% vs. 1.2%; P=0.023). There
fibrillation, and complete atrioventricular block (CAVB). was no significant difference in the use of the intraaortic
Results: A total of 1,033 were included in the study: G1 balloon pump. In the forward stepwise multivariate logistic
1. Doctorate Degree in Cardiology at the University of Sao Paulo de Janeiro (UFRJ); Physician at National Institute of Cardiology,
(USP); Hospital Coordinator at the National Institute of Rio de Janeiro, RJ, Brazil.
Cardiology, Rio de Janeiro, RJ, Brazil. 8. Doctorate Degree in Cardiology at The University of São Paulo
2. Master Degree in Cardiology; Head of the Coronary Disease (USP); Physician at the Coronary Disease Service; National
Service at the National Institute of cardiology Rio de Janeiro, RJ, Institute of Cardiology, Rio de Janeiro, RJ, Brazil.
Brazil.
3. Doctorate Degree in Cardiology; Physician at Coronary Disease This study was carried out at the National Institute of Cardiology,
Service; National Institute of Cardiology, Rio de Janeiro, RJ, Health Ministry, Rio de Janeiro, RJ, Brazil.
Brazil.
4. Specialization in Cardiology; Physician at the Coronary Disease Corresponding author:
Service; National Institute of Cardiology, Rio de Janeiro, RJ, Antônio Sérgio Cordeiro da Rocha
Brazil. Coordenação de Pesquisa Clínica
5. Doctorate Degree in Cardiology at USP; Head of the Surgical Rua das Laranjeiras, 374/5º andar – Rio de Janeiro, RJ
Division: National Institute of Cardiology, Rio de Janeiro, RJ, Brasil – Zip Code: 22040-006.
Brazil. E-mail: ascrbr@centroin.com.br
6. Specialization in Heart Surgery; Head of the Adult Surgical Service;
National Institute of Cardiology, Rio de Janeiro, RJ, Brazil. Article received on October 7th, 2011
7. Doctorate Degree in Cardiology at the Federal University of Rio Article accepted on February 2nd, 2012
45
Rocha ASC, et al. -Age influences outcomes in 70-year or older Rev Bras Cir Cardiovasc 2012;27(1):45-51
patients undergoing isolated coronary artery bypass graft surgery
Resumo
Abreviations, acronyms & Symbols
Objetivo: Analisar os resultados da cirurgia de
ITA internal thoracic artery
revascularização miocárdica (CRVM) isolada com circulação
C VA stroke extracorpórea em pacientes com idade ≥ 70 anos em
CAVb complete atrioventricular block comparação àqueles com < 70 anos.
IAB Intraaortic balloon pump Métodos: Pacientes submetidos consecutivamente à
ECC/ extracorporeal circulation CRVM isolada. Os pacientes foram agrupados em G1 (idade
CABG coronary artery by-pass graft surgery ≥ 70 anos) e G2 (idade < 70 anos). Os desfechos analisados
CAD coronary artery disease
DM diabetes melito
foram letalidade hospitalar, infarto agudo miocárdio (IAM),
COPD chronic obstructive pulmonary disease acidente vascular encefálico (AVE), reoperação para revisão
VD vascular disease de hemostasia (RRH), necessidade de balão intra-aórtico
PVD peripheral vascular disease (BIA), complicações respiratórias, insuficiência renal aguda
AF atrial fibrillation (IRA), mediastinite, sepse, fibrilação atrial (FA) e bloqueio
G1 group of patients of age = or > 70 years atrioventricular total (BAVT).
G2 group of patients of age < 70 years Resultados: Foram estudados 1033 pacientes, 257 (24,8%)
SH systemic hypertension
do G1 e 776 (75,2%) do G2. A letalidade hospitalar foi
AMI acute myocardial infarction
IBGE/BIGS Brazilian Institute of Geography and Statistics
significantemente maior no G1 quando comparado ao G2 (8,9%
CI Confidence interval vs. 3,6%, P=0,001), enquanto a incidência de IAM foi semelhante
AKI acute kidney injury (5,8% vs. 5,5%; P=0,87). Maior número de pacientes do G1
CKD chronic kidney disease necessitou de RRH (12,1% vs. 6,1%; P=0,003). Da mesma forma,
LCATI left coronary artery trunk injury no G1 houve maior incidência de complicações respiratórias
RHR reoperation for hemostasis review (21,4% vs. 9,1%; P<0,001), mediastinite (5,1% vs. 1,9%;
ECCT/CPBT extracorporeal circulation time P=0,013), AVE (3,9% vs. 1,3%; P=0,016), IRA (7,8% vs. 1,3%,
HF heart failure
P<0,001), sepse (3,9% vs. 1,9%; P=0,003), fibrilação atrial (15,6%
vs. 9,8%; P=0,016) e BAVT (3,5% vs. 1,2%; P=0,023) do que o
G2. Não houve diferença significante na necessidade de BIA.
regression analysis, age ≥ 70 years was an independent Na análise regressão logística multivariada “forward stepwise”,
predictive factor for higher in-hospital mortality (P=0.004), a idade ≥ 70 anos foi fator preditivo independente para maior
re-exploration for bleeding (P=0.002), sepsis (P=0.002), letalidade operatória (P=0,004) e para RRH (P=0,002), sepse
respiratory complications (P<0.001), mediastinitis (P=0.016), (P=0,002), complicações respiratórias (P<0,001), mediastinite
stroke (P=0.029), acute renal failure (P<0.001), atrial (P=0,016), AVE (P=0,029), IRA (P<0,001), FA (P=0,021) e BAVT
fibrillation (P=0.021), and CAVB (P=0.031). (P=0,031) no pós-operatório.
Conclusion: This study suggests that patients of age ≥ 70 Conclusão: Este estudo sugere que pacientes com idade
years were at increased risk of death and other complications ≥ 70 anos estão sob maior risco de morte e outras
in the CABG’s postoperative period in comparison to younger complicações no pós-operatório de CRVM em comparação
patients. aos pacientes mais jovens.
Descriptors: Myocardial revascularization. Hospital Descritores: Revascularização miocárdica. Mortalidade

mortality. Postoperative complications. Aged. hospitalar. Complicações pós-operatórias. Idoso.
INTRODUCTION vascular, etc.), CABG has become a procedure with low

mortality and morbidity, due to improvements in surgical
The proportion of elderly people in Brazil has increased techniques, anesthetic and postoperative care.
considerably over recent decades. Between 1980 and 2009, The objective of this study was to analyze the results
life expectancy of the population has increased more than of isolated CABG in patients of age ≥ 70 years compared
10 years ranging from 62.57 years to 73.17 years [1]. In with patients < 70 years of age.
addition, it is estimated that in 2050, over 15% of the Brazilian
population will be 70 years or older. METHODS
Due to the increased prevalence of coronary artery
disease (CAD) with age, it is assumed that an increasing This is a historical prospective study in which we
number of elderly patients will become a candidate for analyzed all patients who consecutively underwent isolated
coronary artery bypass grafting (CABG) in the coming CABG from October 1, 2001 through August 31, 2005.
years. Although this age group is susceptible to the Exclusion criteria were patients who underwent off-pump
influence of a number of comorbidities (renal, pulmonary, CABG or associated with other cardiac surgeries (orovalvar
46
Rocha ASC, et al. - Age influences outcomes in 70-year or older Rev Bras Cir Cardiovasc 2012;27(1):45-51
diseases, ventricular aneurysms, acquired interventricular mechanical ventilation > 24 h, or pulmonary infection
communications, congenital heart defects) or vascular requiring postoperative unit stay, acute kidney injury (AKI)
surgeries. The patients were assigned into two groups: G1 requiring dialysis process, mediastinitis, sepsis from any
(age ≥ 70 years) and G2 (age < 70 years). Data were retrieved source, atrial fibrillation (AF), and complete atrioventricular
directly from the database of the adult surgery service at block (CAVb) requiring temporary or permanent pacemaker.
the National Institute of Cardiology. The fulfilling of all the Urgent or emergency surgery was defined according to
fields of the form is mandatory in order to accomplish the the criteria of the American Heart Association and American
administrative process of discharging the patient. College of Cardiology [2].
At hospital admission, we collect demographic, clinical, Continuous variables are expressed as means ± standard
and laboratory data, in addition to medical history and deviation (SD), while categorical variables are expressed
physical examination. We have also gathered the by proportions. In the statistical analysis, comparisons of
comorbidities for CABG according to the criteria of the means were assessed using the Student’s t-test. Proportions
American Heart Association and the American College of were compared using the Chi-square or Fisher’s exact test.
Cardiology [2], and data relevant to surgery, such as CPB Forward stepwise multivariate logistic regression analysis
time, number of anastomoses received per patient, and was used to determine which factors could be
number of internal thoracic artery grafts used. Patients were independently relevant to the development of the study
stratified by surgical risk of death using the European outcomes. All P values are two-tailed, and P = 0.05 was
System for cardiac operative risk evaluation (additive considered as significant.
EuroSCORE).
Hypertension (HBP) was considered present when blood RESULTS
pressure was ≥ 140/90 mmHg or the patient was under regular
antihypertensive medication. Diabetes mellitus (DM) was During the study period, 1,033 patients underwent
defined by a record of an abnormal glucose tolerance test, a isolated CABG. Of these, 257 (24.8%) comprised G1 and 776
fasting blood-glucose level ≥ 126 mg/dL on two separate (75.2%) comprised G2.
tests, or the regular use of oral hypoglycaemic agents, insulin Table 1 shows that there was no difference between the
sensitizer drugs, or insulin either alone or combined. Chronic two groups of patients related to the following: gender,
kidney disease (CKD) was considered present when DM, systemic hypertension, routine diagnostic tests of
creatinine clearance was < 60 ml/h, or the patient was stable or unstable angina, or myocardial infarction less than
undergoing dialysis. Vascular disease (VD) was considered three months of CABG, chronic obstructive pulmonary
when there was a history of intermittent claudication, ankle/ disease (COPD), previous stroke (CVA), CKD, VD, or need
brachial index < 0.9, and peripheral vascular/arterial or for urgent or emergency surgery, or previous CABG.
cerebrovascular obstruction over 50% on color Doppler, CT However, compared to patients in G2, patients in G1
angiography or conventional angiography. had a higher prevalence of peripheral vascular disease
As a routine, all patients underwent a two-dimensional (PVD) (18.3% vs. 10.7%, P = 0.002), more impairment of the
echocardiography study with color Doppler to evaluate left main coronary artery (37.7% vs 26.8%, P = 0.001), and
the cavity dimensions and left ventricle (LV) global and high-risk EuroSCORE (36.2% vs. 8.4%, P <0.001). Table 2
segmental function before both surgery and patient shows that the number of anastomoses per patient was
discharge, or at the discretion of the attending physician. significantly higher in G2 than in G1 [4 (95% CI = 1-5) vs. 2
The extent and degree of coronary stenoses were evaluated (95% CI = 1-3), P = 0.017]. However, the number of internal
on cineangiocoronariography by at least two highly skilled thoracic artery grafts used was similar (95.5% vs. 93.0%,
professional hands. respectively, P = 0.713).
The following outcomes were analyzed: death from any Table 3 presents the results of surgery. The mortality
hospital origin and other postoperative complications rate was higher in G1 than in G2 (8.9% vs. 3.6%, P = 0.001).
occurred during the same hospitalization after CABG, or The incidence of postoperative AMI was similar between
within the first 30 days postoperatively. The following the two groups of patients (5.8% vs 5.5%, P = 0.876).
postoperative complications were analyzed: non-fatal Compared to patients in G2, a greater number of patients in
diagnosed acute myocardial infarction (AMI) according to G1 required reoperation for hemostasis review (12.1% vs.
the guidelines of the European Society of Cardiology [3], 6.1%, P = 0.003) and developed more respiratory
stroke (cerebrovascular accident/CVA) characterized as any complications (21.4% vs. 9.1%, P <0.001), mediastinitis
transient or permanent neurological abnormality proven by (5.1% vs. 1.9%, P = 0.013), stroke (CVA) (3.9% vs. 1.3%, P =
CT or MRI of the brain, reoperation for hemostasis review, 0.016), AKI (7.8% vs. 1.3%, P <0.001), sepsis (3.9% vs. 1.9%,
circulatory shock requiring intraaortic balloon pump (IAB), P = 0.003), AF (15.6% vs. 9.8%, P = 0.016), and CAVb
respiratory complications characterized by the use of postoperatively (3.5% vs. 1 2%, P = 0.023).
47
Table 1. Demographic and clinical preoperative characteristics Table 3. Postoperative outcomes in both groups of patients
distributed through both groups of patients. G1 (257) G2 (776) P
G1 (257) G2 (776) P Hospital mortality (%) 23(8,9) 28 (3,6) 0,001
Male gender (%) 68.1 72.8 0.151 Hemostasis review (%) 31 (12,1) 47 (6,1) 0,003
Age 74.0 ± 3.2 58.0 ± 7.8 <0.001 Post-AMI (%) 15 (5,8) 43 (5,5) 0,876
DM (%) 29.6 28.6 0.812 Respiratory Complications (%) 55 (21,4) 71 (9,1) <0,001
SH (%) 83.3 85.6 0.366 AKI (%) 20 (7,8) 10 (1,3) <0,001
CKD (%) 3.1 1.3 0.093 CVA/Stroke (%) 10 (3,9) 10 (1,3) 0,016
COPD (%) 7.8 7.2 0.783 Sepsis (%) 10 (3,9) 7 (0,9) 0,003
VD (%) 18.3 10.7 0.002 Mediastinitis (%) 13 (5,1) 15 (1,9) 0,013
Stable angina (%) 63.8 66.6 0.448 Atrial Fibrillation (%) 40 (15,6) 76 (9,8) 0,016
Unstable angina (%) 28.0 26.5 0.685 CAVb (%) 9 (3,5) 9 (1,2) 0,023
AMI < 3 months (%) 13.6 15.1 0.612 AMI = acute myocardial infarction; AKI = acute kidney injury;
HF (%) 2.3 2.6 1.0 stroke/CVA = cerebrovascular accident; CAVb = complete
Stroke (%) 4.3 2.4 0.136 atrioventricular block
Reoperation (%) 5.1 7.2 0.252
U/E CABG sugery (%) 28.0 26.7 0.682
Aditive EuroSCORE > or
Table 4. Predictive factors of post-operative complications by
= 6 points(%) 36.2 8.4 <0.001
logistic regression analysis
SH= systemic hypertension; CKD = chronic kidney disease; DM= OR IC95% P
diabetes mellitus; COPD= chronic obstructive pulmonary disease; Hospital mortality
VD= vascular disease; AMI= acute myocardial infarction; HF= Age > 70 years 2.315 1.296 a 4.136 0.004
heart failure; Stroke/CVA cerebrovascular accident; U/E CABG = VD 2.434 1.263 a 4.689 0.007
urgent or emergency coronary artery bypass graft surgery; Homeostasis review
EuroSCORE = Additive European System for Cardiac Operative Age > 70 years 2.201 1.355 a 3.601 0.002
Risk Evaluation Sepsis
Age > 70 years 5.026 1.847 a 13.679 0.002
Respiratory complications
Table 2. Cineangiocoronariographic and surgical operation Age > 70 years 2.537 1.702 a 3.784 <0.001
characteristics in both groups of patients VD 1.998 1.216 a 3.284 0.006
G1 (257) G2 (776) P AMI > 3 months 2.302 1.428 a 3.710 0.001
Lesão de 1 vaso (%) 0,0 0,4 0,007 LACD 1.614 1.079 a 2.414 0.020
Lesão de 2 vasos (%) 7,8 8,0 0,007 Stroke/CVA
Lesão de 3 vasos (%) 54,5 64,8 0,007 Age > 70 years 2.852 1.116 a 7.290 0.029
LTCE (%) 37,7 26,8 0,001 DM 2.602 1.007 a 6.724 0.048
TCEC min 76,2±27,6 73,6±26,9 0,182 COPD 7.020 2.057 a 23.961 0.002
Anastomoses/paciente Previous stroke/CVA 20.705 3.560 a 120.433 0.001
mediana (IC95%) 4 (1 a 5) 3 (1 a 4) 0,017 Mediastinitis
Enxerto de ATI (%) 93,0 95,5 0,141 Age > 70 years 2.613 1.193 a 5.724 0.016
LCAD = left main coronary artery disease; ECCT= extracorporeal Unstable angina 3.133 1.418 a 6.922 0.004
circulation time // CPBT cardiopulmonary bypass time; CI = CKD 5.247 1.217 a 22.616 0.026
confidence interval; ITA = internal thoracic artery AKI
Age > 70 years 6.015 2.672 a 13.542 <0.001
IRC 12.918 3.009 a 55.453 0.006
IAM > three months 4.206 1.717 a 10.303 0.001
Atrial fibrillation
Table 4 shows that the multivariate logistic regression
Age > 70 years 1.646 1.075 a 2.522 0.021
analysis, age ≥ 70 years (P = 0.004), and the presence of
DM 2.046 1.355 a 3.089 0.007
PVD (P = 0.007) were factors associated with increased VD 1.963 1.181 a 3.264 0.009
hospital mortality. Age ≥ 70 years was the only factor CAVb
associated requiring reoperation for hemostasis review (P Age > 70 years 2.905 1.102 a 7.654 0.031
= 0.002) and postoperative sepsis (P = 0.002). The main CKD 9.328 1.561 a 55.739 0.014
variables associated with postoperative respiratory VD = vascular disease; AMI = acute myocardialinfarction; LCAD =
complications after surgeries were as follows: Age ≥ 70 left main coronary artery disease; Stroke/CVA = cerebrovascular
years (P <0.001), PVD (P = 0.006), myocardial infarction < accident; DM=diabetes mellitus; COPD = chronic obstructive
three months after CABG (P = 0.001), and lesion of the left pulmonar disease; CKD = chronic kidney disease; AKI = acute kidney
main coronary artery (P = 0.020). injury; CAVb = complete atrioventricular block
48
The factors associated with mediastinitis were age ≥ dysfunction, and stroke (CVA). They also required more
70 years (P = 0.016), unstable angina (P = 0.004), and CKD vasopressors than patients < 70 years of age [5]. In another
(P = 0.026). The factors associated with postoperative study at the Mount Sinai School of Medicine, Mount Sinai,
stroke were age ≥ 70 years (P = 0.029), diabetes (P = 0.048), New York, data from 2,985 patients undergoing CABG were
COPD (P = 0.002), and previous stroke (CVA) (P = 0.001). prospectively collected. It was found that the operative
AKI in the postoperative period was associated with age mortality in patients of age = or > 80 years was 4.6%, in
≥ 70 years (P <0.001), CKD (P = 0.006, OR = 12.91), and MI septuagenarians it was 2.2%, and in patients < 70 years of
< three months after CABG (P = 0.001). Factors associated age it was 2.4% [6]. Naughton et al. [7] also compared the
with postoperative FA were age ≥ 70 years (P = 0.021), results in patients aged ≥ 75 years and aged 60-74 years
DM (P = 0.006), and PVD (P = 0.009). Factors associated undergoing CABG. Operative mortality (30 days) in the
with postoperative CAVb were age ≥ 70 years (P = 0.031), patients aged > 75 years was 5% compared to 1.8% in the
and CRF (P = 0.014). There was no significant difference younger patients (aged 60-74 years). The logistic
between G1 and G2 in relation to the occurrence of regression analysis showed that an age > 75 years was an
circulatory shock requiring IAB pump (13.6% vs. 10.6%, independent factor for operative mortality. Peterson et al.
P = 0.211). [8] have analyzed the outcomes of CABG performed in
24,461 patients registered in the Medicare program in the
DISCUSSION United States. They found that the operative mortality was
11.5% in patients of age ≥ 80 years versus 4.4% in patients
The present study performed at a cardiology center, of age 65 to 70 years. On the other hand, Ng et al. [9] found
which is a reference in highly complex procedures, suggests no significant difference in hospital mortality of patients ≥
that elderly patients of age ≥ 70 years are at increased risk 70 years of age compared to those < 70 years old undergoing
of hospital mortality and postoperative complications of CABG (5.4% vs. 3.8%, respectively).
all sorts as compared to younger patients after CABG. In When analyzing the outcomes of studies that did not
spite of patients aged ≥ 70 years present more comorbidities compare elderly versus younger patients undergoing on-
preoperatively than younger ones (Table 1), data pump CABG, it appears that the operative mortality varies
adjustment by multivariate logistic regression analysis widely ranging from 1.6% to 27% [10-12]. This implies
linked them to increased risk of operative mortality and different levels of preoperative risk of these patients. An
postoperative complications. It was observed that the analysis of the preoperative characteristics of the elderly
elderly patients were two times more likely to die during the patients involved in this study shows the presence of a
procedure than patients < 70 years of age (Table 4). more severe atherosclerotic damage compared to the
Furthermore, compared to younger patients, elderly patients younger patients. This is supported by the higher
are likely to have more post-operative complications prevalence of vascular disease (cerebrovascular and
(occurrence of atrial fibrillation = 1.6 times and development peripheral) and CKD (Table 1). In Brazil, when considering
of AKI = 6 times) (Table 4). the predictors of mortality in patients aged > 70 years
In this study, the higher mortality of aged patients after undergoing CABG or valve replacement with CPB,
CABG is consistent with previous published results, in Anderson et al. [13] reported a mortality rate of 8.3% for
which the operative outcomes in elderly and younger those undergoing isolated CABG. Souza et al. [12] found a
patients were compared. In the study conducted by 30-day hospital mortality rate of 8.5%, when analyzing the
Johnson et al. [4], the influence of age alone on the outcome outcomes of CABG performed in 492 patients aged 70 years
of heart surgery performed in octogenarian patients or over. Iglézias et al. [14] reported an operative mortality
compared to younger patients was questioned. In a rate of 8.5% in a retrospective analysis of 47 octogenarians
multivariate analysis, Johnson et al. demonstrated that 522 who underwent CABG at the Heart Institute, University of
aged 80 years or older undergoing CABG had a higher risk São Paulo (INCOR) between 1978 and 1993.
of death, longer length of hospital stay, neurological In a retrospective study on the outcomes of isolated
complications, and need for reoperation to treat bleeding CABG in 144 patients aged ≥ 70 years, Deinninger et al. [15]
than non-octogenarians. Similarly, Alves, Jr. et al. [5] in a observed an operative mortality rate of 5.5%. Almeida et al.
study involving 197 patients septuagenarians or elderly [16] observed hospital mortality rate of 7.1%, when analyzing
patients undergoing CABG and valve operations observed the outcomes of 70 patients after the eighth decade of life
operative mortality of isolated CABG in septuagenarians undergoing CABG. Pivatto et al. [17] described the hospital
compared to younger patients (19% versus 6%, morbidity and mortality of 140 patients aged ≥ 80 years
respectively). These authors also demonstrated that undergoing isolated or combined CABG. They have found
septuagenarians had more postoperative bleeding, an in-hospital mortality rate of 14.3%, distributed as follows:
pulmonary complications, mediastinitis, kidney 10% for isolated CABG and 22% for CABG associated with
49
other cardiac surgeries. In this study, it was also reported not necessarily with the same biological status [22]. As in
that the most frequent complications were: low output this study, we did not assess the frailty [22] in patients
(27.9%), kidney dysfunction (10%), and prolonged undergoing CABG, there is no way to measure how many
ventilatory support (9.6%) [17]. of them were vulnerable in their psychological and
Assuming that off-pump CABG could bring benefits to biological conditions, despite not having significant
patients at higher surgical risk for both operative mortality comorbidities.
and postoperative complications, some investigators have
compared the clinical outcomes of elderly patients LIMITATIONS OF THE STUDY
undergoing on-pump versus off-pump CABG. Iglézias et
al. [18] compared the clinical outcomes in patients aged ≥ Like any other observational study, this is only a
80 years undergoing on-pump versus off-pump CABG. hypothesis generator study. However, depending on the
They reported an operative hospital mortality rate much number of patients involved, it is reasonable to assume
higher in on-pump CABG (38%) than in off-pump CABG that the results are representative of current clinical practice
(11.7%) [18]. However, other postoperative complications in our country. Because all patients underwent on-pump
were similar. Thus, the incidence of myocardial infarction CABG, it is obviously that we could not verify whether off-
was 3.4% vs 2.8%; stroke (CVA) was 0% vs. 4%; assisted pump surgery would bring any different result than
ventilation > 24 hours 27.4% vs 21.1% and reoperation 2.9% observed.
vs. 1.9%, respectively for on-pump vs. of-pump CABG [18].
In a similar analysis, Lee et al. [19] retrospectively compared CONCLUSIONS
the results of on-pump vs. off-pump CABG. They found
that on-pump CABG had a higher mortality rate compared This study suggests that patients age ≥ 70 years are at
to off-pump CABG (11.5% vs. 2.1%, respectively). However, increased risk of death and other complications after CABG
unlike Iglézias et al. [18], they observed a higher incidence compared with younger patients.
of other postoperative complications: stroke (CVA) (11.5%
vs. 0%); AF (30.8% vs. 12.8%), AKI (19.2% vs. 0%),
respiratory failure (16% vs. 2.1%), dialysis (20% vs. 0%),
mechanical ventilation > 24 h (24% vs. 4.3%) and AMI (8% REFERENCES
vs. 4.3%) respectively in on-pump vs. off-pump CABG.
In another study in which the in-hospital outcome of 87 1. Instituto Brasileiro de Geografia e Estatistica (homepage).
patients aged ≥ 70 years that underwent off-pump CABG Brasília, DF. Ministério do Planejamento, Orçamento e Gestão;
was analyzed, Silva et al. [20] observed a hospital mortality (tábuas completas de mortalidade 2008; comunicação social
em 01.12.2009; acessado em 01.05.2011). Disponível em: http:/
rate of 4.6%. Of these, the incidence of AF was observed in
/www.ibge.gov.br.
32.2% of the patients, bronchopneumonia in 10.3%, sepsis
in 3.4%, AMI in 2.3%, mediastinitis in 1.1%, transient 2. Eagle KA, Guyton RA, Davidoff R, Edwards FH, Ewy GA,
ischemic attacks (TIAs) in 1.1%, and pneumothorax in 1.1%. Gardner TJ, et al; American College of Cardiology; American
However, contrary to all these studies, Saleh et al. [21] Heart Association Task Force on Practice Guidelines; American
compared retrospectively isolated on-pump and off-pump Society for Thoracic Surgery and the Society of Thoracic
CABG in 343 octogenarian patients who were matched by Surgeons. ACC/AHA 2004 guideline update for coronary artery
the propensity score. They observed no significant bypass graft surgery: summary article: a report of the American
differences between on-pump and off-pump patients with College of Cardiology/American Heart Association Task Force
on Practice Guidelines (Committee to Update the 1999
regards in-hospital mortality rate, MI, and stroke.
Guidelines for Coronary Artery Bypass Graft Surgery).
Conceptually speaking, the definition of an elderly Circulation. 2004;110(9):1168-76.
individual based on the chronological factor is subject to
errors, even though it lacks psychic, organic, and functional 3. Task Force on Myocardial Revascularization of the European
evidence [20]. In Brazil, the Brazilian Institute of Geography Society of Cardiology (ESC) and the European Association
and Statistics (IBGE) classifies as elderly the individuals for Cardio-Thoracic Surgery (EACTS); European Association
aged 60 or older, which obviously does not accurately reflect for Percutaneous Cardiovascular Interventions (EAPCI), Wijns
the biological state of them. This difficulty in the W, Kolh P, Danchin N, Di Mario C, Falk V, Folliguet T, et al.
classification of elderly individuals extends to the moment Guidelines on myocardial revascularization. Euro Heart J.
2010;31(20):2501-55.
in which they need a high complexity procedure, such as a
heart surgery. The scores that assess the operative risk do 4. Johnson WM, Smith JM, Woods SE, Hendy MP, Hiratzha
not take into account the biological factor, which leads us LF. Cardiac surgery in octogenarians: does age alone influence
to group patients with the same chronological status but outcomes? Arch Surg. 2005;140(11):1089-93.
50
5. Alves Jr. L, Rodrigues AJ, Évora PRB, Basseto S, Scorzoni 14. Iglézias JCR, Dallan LA, Oliveira SF, Ramires JAF, Oliveira
Filho A, Luciano PM, et al. Fatores de risco em septuagenários SA, Verginelli G, et al. Revascularização do miocárdio no
ou mais idosos submetidos à revascularização do miocárdio e pacientes octogenário: 15 anos de observação. Rev Bras Cir
ou operações valvares. Rev Bras Cir Cardiovasc. Cardiovasc. 1993;8(3):237-40.
2008;23(4):550-5.
15. Deinninger MO, Oliveira OG, Guedes MGA, Deininger
6. Filsoufi F, Rahmanian PB, Castillo JG, Chikwe J, Silvay G, EDG, Cavalcanti ACW, Cavalcanti MGFW, et al. Cirurgia
Adams DH. Results and predictors of early and late outcomes de revascularização do miocárdio no idoso: estudo
of coronary artery bypass graft surgery in octogenarians. J descritivo de 144 casos. Rev Bras Cir Cardiovasc.
Cardiothorac Vasc Anesth. 2007;21(6):784-92. 1999;14(2):88-97.
7. Naughton C, Feneck RO, Roxburgh J. Early and late predictors 16. Almeida RMS, Lima Jr JD, Martins JF, Loures DRR.
of mortality following on-pump coronary artery bypass graft Revascularização do miocárdio em pacientes após a
surgery in the elderly as compared to younger population. Eur oitava década de vida. Rev Bras Cir Cardiovasc.
J Cardiothorac Surg. 2009;36(4):621-7. 2002;17(2):8-14.
8. Peterson ED, Cowper PA, Jollis JG, Bebchuck JD, DeLong 17. Pivatto Jr F, Kalil RAK, Costa AR, Pereira EMC, Santos EZ,
ER, Muhlbaier LH, et al. Outcome of coronary artery bypass Valle FH, et al. Morbimortalidade em octogenários submetidos
graft surgery in 24,461 patients aged 80 years or older. à cirurgia de revascularização miocárdica. Arq Bras Cardiol.
Circulation. 1995;92(9 Suppl):II85-91. 2010;95(1):41-6.
9. Ng CY, Ramli MF, Awang Y. Coronary bypass surgery in 18. Iglézias JCR, Chi A, Talans A, Dallan LAO, Lourenço Jr A,
patients aged 70 years and over: mortality, morbidity, length Stolf NAG. Desfechos clínicos pós-revascularização do
of stay and hospital cost. Asian Cardiovasc Thorac Ann. miocárdio no paciente idoso. Rev Bras Cir Cardiovasc.
2004;12(3):218-23. 2010;25(2):229-33.
10. Iglézias JCR, Oliveira Jr JL, Dallan LAO, Lourenção Jr A, 19. Lima RC, Diniz R, Césio A, Vasconcelos F, Gesteira M,
Stolf NAG. Preditores de mortalidade hospitalar no paciente Menezes AM, et al. Revascularização miocárdica em pacientes
idoso portador de doença arterial coronária. Rev Bras Cir octogenários: estudo retrospectivo e comparativo entre
Cardiovasc. 2001;16(2):94-104. pacientes operados com e sem circulação extracorpórea. Rev
Bras Cir Cardiovasc. 2005;20(1):8-13.
11. Krane M, Voss B, Hiebinger A, Deutsch MA, Wottke M,
Hapfelmeier A, et al. Twenty years of cardiac surgery in patients 20. Silva AMRP, Campagnucci VP, Pereira WL, Rosa RF, Franken
aged 80 years and older: risks and benefits. Ann Thorac Surg. RA, Gandra SMA, et al. Revascularização do miocárdio sem
2011;91(2):506-13. circulação extracorpórea em idosos: análise da morbidade
mortalidade. Rev Bras Cir Cardiovasc. 2008;23(1):40-5.
12. Souza JM, Berlinck MF, Moreira MG, Martins JRM, Moreira
MCS, Oliveira PAF, et al. Revascularização miocárdica em 21. Saleh HZ, Shaw M, Fabri BM, Chalmers JA. Does avoidance
pacientes com idade igual ou superior a 70 anos. Rev Bras Cir of cardiopulmonary bypass confer any benefits in
Cardiovasc. 1990;5(3):141-8. octogenarians undergoing coronary surgery? Interact Cardiovasc
Thorac Surg. 2011;12(3):435-9.
13. Anderson AJPG, Barros Neto FXR, Costa MA, Dantas LD,
Hueb AC, Prata MF. Preditores de mortalidade em pacientes 22. Sündermann S, Dademasch A, Praetorius J, Kempfert J, Dewey
acima de 70 anos na revascularização miocárdica ou troca valvar T, Falk V, et al. Comprehensive assessment of frailty for elderly
com circulação extracorpórea. Rev Bras Cir Cardiovasc. high-risk patients undergoing cardiac surgery. Eur J
2011;26(1):69-75. Cardiothorac Surg. 2011;39(1):33-7.
51
Atrioventricular block in the postoperative period

of heart valve surgery: incidence, risk factors and
hospital evolution
Bloqueio atrioventricular no pós-operatório de cirurgia cardíaca valvar: incidência, fatores de risco
e evolução hospitalar
Andres Di Leoni Ferrari1, Carolina Pelzer Süssenbach2, João Carlos Vieira da Costa Guaragna3,
Jacqueline da Costa Escobar Piccoli4, Guilherme Ferreira Gazzoni5, Débora Klein Ferreira6, Luciano
Cabral Albuquerque7, Marco Antonio Goldani8
DOI: 10.5935/1678-9741.20110010 RBCCV 44205-1291
Abstract artery bypass grafting combined with valve repair and 23

Introduction: Disturbances of the cardiac conduction (2.1%) aortic and mitral combined surgery. 187 patients
system are potential complications after cardiac valve surgery. (17%) showed clinical and electrocardiographic pattern of
Objectives: This study was designed to investigate the atrio-ventricular block requiring artificial temporary pacing.
association between perioperative factors and atrio- Of these, 14 patients (7.5%) required permanent pacemaker
ventricular block, the need for temporary cardiac artificial implantation (1.27% of the total valve surgery patients).
pacing and, if necessary, permanent pacemaker implantation Multivariate analysis showed association of the incidence
after cardiac valve surgery. of atrio-ventricular block and temporary pacing with mitral
Methods: Retrospective analysis of the Cardiac Surgery valve surgery (OR 1,76; CI 95% 1.08-2.37; P=0.002),
Database - Hospital São Lucas/PUCRS. The data are implantation of bioprosthetic devices (OR 1.59; CI 95% 1.02-
collected prospectively and analyzed retrospectively. 3.91; P=0,039), age over 60 years (OR 1.99; CI 95% 1.35-
Results: Between January 1996 and December 2008 were 2.85; P<0.001), prior use of anti-arrhythmic drugs (OR 1.86;
included 1102 valve surgical procedures: 718 aortic valves CI 95% 1.04-3.14; P=0.026) and previous use of b-blocker
(65.2%), 407 (36.9%) mitral valve and 190 (17.2%) coronary (OR 1.76; CI 95% 1.25-2.54; P=0.002). Remarkably the
1. Specialist, MD, Arrhythmia Clinic at the Hospital São Lucas da 8. Specialist in Cardiovascular Surgery, Chief of Cardiovascular
Pontificia Universidade Catolica do Rio Grande do Sul (PUC-RS), Surgery of the PRM, Hospital São Lucas da PUC-RS, Porto Alegre,
Porto Alegre, Brazil. Brazil.
2. Resident of Cardiology at PRM de Cardiologia, Hospital São
Lucas da PUC-RS, Porto Alegre, Brazil. Work performed at the Department of Cardiology, Electrophysiology
3. Doctor of Cardiology, Chief of Cardiology, PRM Hospital São Laboratory, Department of Cardiac Surgery, Hospital São Lucas da
Lucas da PUC-RS, Head of the Postoperative Cardiac Surgery, Pontificia Universidade Catolica do Rio Grande do Sul (PUC-RS),
Hospital São Lucas da PUC-RS, Porto Alegre, Brazil. Porto Alegre, Brazil.
4. Ph.D. in Cellular and Molecular Biology, Adjunct Professor,
Universidade Federal do Pampa - Campus Uruguaiana, RS, Brazil. Mailing address:
5. Cardiology Electrophysiology Service of the Hospital São Lucas Andres Di Leoni Ferrari
da PUC-RS, Porto Alegre, Brazil. Av. Ipiranga, 6690 – Sala 300 – Jardim Botânico – Porto Alegre, RS,
6. Resident of General Surgery, Hospital Municipal Miguel Couto, Brazil – Zp Code: 90610-000
Rio de Janeiro, Brazil. E-mail: andredileoni@terra.com.br
7. Doctor of Health Sciences specialist in Cardiology, Federal
University of Rio Grande do Sul, Cardiovascular Surgeon of Hospital Article received on May 26th, 2011
São Lucas da PUC-RS, Porto Alegre, Brazil. Article accepted on July 14th, 2011
364
Ferrari ADL, et al. - Atrioventricular block in the postoperative Rev Bras Cir Cardiovasc 2011;26(3):364-72
period of heart valve surgery: incidence, risk factors and hospital
evolution
presence of atrio-ventricular block did not significantly show destas, 190 (17,2%) cirurgias de revascularização miocárdica
association with increased mortality, but significantly associadas à cirurgia valvar e 23 (2,1%) cirurgias valvares
prolonged (P<0.0001) hospital length-of-stay and, therefore, combinadas (aórtica+mitral). Cento e oitenta e sete (17%)
hospital costs. pacientes apresentaram quadro clínico e eletrocardiográfico
Conclusions: Our study presents a group of predictive de BAV durante o POCC valvar, necessitando de ECAT.
factors referring to a specific patient profile by which high Quatorze (7,5%) pacientes evoluíram para implante de
risk of atrio-ventricular block and the need of temporary marcapasso definitivo (1,27% do total da amostra). A análise
cardiac pacing after cardiac valve surgery it is determined. multivariada evidenciou associação significativa de BAV com
cirurgia de valva mitral (OR=1,76; IC 95% 1,08-2,37;
Descriptors: Atrioventricular block. Pacemaker, artificial. P=0,002), implante de prótese biológica (OR=1,59; IC 95%
Cardiovascular surgical procedures. Heart valves. 1,02-3,91; P= 0,039), idade maior que 60 anos (OR = 1,99; IC
Postoperative complications. 95% 1,35-2,85; P<0,001), uso prévio de medicações
antiarrítmicas (propafenona e amiodarona) (OR = 1,86; IC
95% 1,04-3,14; P=0,026) e uso prévio de betabloqueador (OR
Resumo = 1,76; IC 95% 1,25-2,54; P=0,002). Embora a presença do
Introdução: Distúrbios do sistema de condução cardíaco BAV e necessidade de ECAT não tenham se associado a
são complicações potenciais e conhecidas dos procedimentos aumento de mortalidade, prolongaram a permanência
de cirurgia cardíaca valvar. hospitalar significativamente (P<0,0001) e, portanto, o
Objetivos: Investigar a associação entre fatores peri- consumo de recursos hospitalares.
operatórios com bloqueio atrioventricular (BAV) e a Conclusão: Esse estudo evidencia um conjunto de fatores
necessidade de estimulação cardíaca artificial temporária preditivos potenciais a um perfil de pacientes que
(ECAT) e, se necessário, implante de marcapasso definitivo determinam alto risco de bloqueio atrioventricular e
no pós-operatório de cirurgia cardíaca (POCC) valvar. necessidade de estimulação cardíaca artificial temporária
Métodos: Coorte histórica de pacientes submetidos a no pós-operatório de cirurgia cardíaca valvar.
cirurgia cardíaca valvar, sendo realizada análise de banco
de dados por regressão logística.
Resultados: No período de janeiro de 1996 a dezembro de Descritores: Bloqueio atrioventricular. Marca-passo
2008, foram realizadas 1102 cirurgias cardíacas valvares: artificial. Procedimentos cirúrgicos cardiovasculares. Valvas
718 (65,2%) na valva aórtica e 407 (36,9%) na valva mitral; cardíacas. Complicações pós-operatórias.
INTRODUCTION METHODS
Disorders of cardiac conduction system are known and Population and sample
potential complications of the procedures for heart valve From January 1996 to December 2008, 1,102 cardiac
surgery. The incidence of disorders of atrioventricular (AV) surgeries were performed at the Hospital São Lucas,
in the post-cardiac surgery (POCS) valve is located, Pontifical Catholic University of Rio Grande do Sul (PUC-
according to the literature, 10 to 15% [1]. Most patients RS), 718 (65.2%) aortic valve surgery and 407 (36.9 %) mitral
have disturbances of a temporary nature and will require valve surgery. Of these, 190 (17.2%) valve surgery (aortic
temporary cardiac pacing (DPM), but 1% to 3% of patients, or mitral) were combined with bypass surgery (CABG) and
given the irreversibility of the framework, will be subject to 23 (2.1%) multiple exchange surgery (aortic + mitral), the
a definitive pacemaker (DPM) during hospitalization [ 1-3]. latter accounted for both surgery group in the aortic and
In this study, we analyzed the experience of more than 1100 mitral valve surgery.
valve surgery procedures in order to verify the relationship
between factors pre-, intra-and postoperative Study design
(perioperative) associated with atrioventricular block (AVB) Historical cohort observational study. Data were
and the need for TAC with later implant DPM on POCS. collected prospectively and entered into the database unit
365
evolution
postoperatively in cardiac surgery at the Hospital São Lucas Hospital São Lucas da PUC-RS, as previously described
da PUC-RS. [4]. After surgery, all patients were transferred to the ICU
postoperatively in cardiac surgery, on mechanical
Inclusion criteria ventilation.
Patients aged over 18 years taken to heart valve surgery
(replacement or repair) alone or combined with myocardial Statistical analysis
revascularization surgery. The data were plotted on a Microsoft Access ®
spreadsheet and analyzed in SPSS Version 11.0.
Exclusion criteria Descriptive statistics were performed, as well as the
Tricuspid and pulmonary valve surgeries were excluded univariate tests: Chi-square for ordinal variables and was
from the analysis when isolated due to the small number of used for quantitative data analysis of variance or Student
patients undergoing these procedures. Also excluded were t test (for unpaired variables) followed by post hoc test
cases with incomplete data on the need for TCP. for Bonferroni data.
Multivariate analysis was done by logistic regression
Study variables (backward conditional method). Statistical difference was
The variables analyzed were: considered P <0.05.
• Age - the average age calculated and also divided into
groups for analysis: less than 60 years and greater than or Ethical considerations
equal to 60 years; The research project study was submitted to the Ethics
• Gender (male/female); Committee in Research of FAMED PUC-RS, under
• Left ventricle ejection fraction (EF) - calculated by registration number 06003478.
echocardiography, shared values for analysis in less than
40% and greater than or equal to 40%; RESULTS
• Chronic kidney disease (CKD) - diagnosed by serum
creatinine> 1.5 mg/dl; Valve surgery from 1102 analyzed, 718 were aortic valve
• Diabetes mellitus (DM); surgery, these 485 (67.56%), valve replacement for aortic
• Chronic obstructive pulmonary disease (COPD) - stenosis, and 233 (32.45%), exchange for aortic
diagnosed clinically and/or radiological examination and / insufficiency. Of the 407 mitral valve surgeries, 193 (47.4%)
or spirometry and / or drug therapy (corticosteroids, were mitral valve replacement and 214 (52.6%) for mitral
bronchodilators); regurgitation. One hundred and ninety (17.24%) surgeries
• Atrial fibrillation (AF); were combined with CABG, these 143 (75.3%) aortic valve
• Previous Heart surgery (CVS); surgery (112 by aortic stenosis and 31 aortic) and 47 (24.7%)
• Previous use of antiarrhythmic drugs (propafenone, mitral valve surgery (14 by 33 by mitral stenosis and mitral
and/ or amiodarone); insufficiency).
• Previous use of beta-blockers; On total cardiac surgery and valvular aortic and/or mitral
• Previous use of digoxin; valve during the period analyzed, 187 (17.0%) patients had
• Class functional New York Heart Association (NYHA); clinical and electrocardiographic atrioventricular block
• Type of cardiac procedure: mitral valve, aortic valve, during the postoperative period, requiring TCP. Table 1
including valve associated with CABG surgery and shows the profile of patients who required temporary
combined valve (aortic + mitral); pacemaker through TCP and univariate analysis of
• Type of prosthesis (biological or metallic); preoperative data of these patients compared with patients
• Calcification; who underwent surgery and did not need the
• Time of cardiopulmonary bypass (CPB); aforementioned device.
• Time of aortic clamping; The characteristics of the patients are shown in Table 1:
• In-hospital mortality. average age of 65.8 years (42% older than 60 years), the
vast majority (90%) patients had an EF greater than 40%,
Outcome 44% of cases had NYHA Class III and IV, 14% underwent
Development of AVB in the POCS and the need for TCP previous cardiac surgery (CVS), 7% were taking
and definitive. antiarrhythmic medication, 25% beta-blockers and 32%
digoxin, 7% were diabetic, 6.89% had chronic kidney disease
Procedures (CKD) (serum creatinine greater than 1.5 mg/dl) and 20%
Anesthesia, the techniques of CPB and cardioplegia had AF.
were performed according to the standardization of the Evaluation of surgical risk score of Guaragna et al. [5]
366
evolution
stratified the risk of patients studied in the following need for TCP in POCS in patients over the age of 60 years
frequency: 36.7% of low risk, medium risk 33.7%, 16.6% (OR = 2.01, 95% CI 1.46 to 2.77, P <0.0001); CKD (OR = 2.12,
high risk, 6.3% very high risk and 6.7% extremely high 95% CI 1.26 to 3.58, P = 0.004), presence of AF (OR = 1.68,
risk. Data analysis showed no statistical significance 95% 1.17 to 2.41, P = 0.004), antiarrhythmic drugs ( (OR =
between the surgical risk of mortality and the need for 2.03, 95% CI 1.22 to 3.38, P = 0.005), beta-blockers (OR = 1.66,
TCP. 95% CI 1.18 to 2.33, P = 0.003) and cases of heart surgery (OR
Univariate analysis, described in Table 1, revealed a greater = 1.54, 95% CI 1.01 to 2.33, P = 0.04).
Table 1. Preoperative characteristics of the groups and univariate analysis

Variable Total TCP Ñ TCP OR IC 95% P
1102 (100%) 187 (17%) 915 (83%)
age
>60 465 (42.2%) 106 (22.8%) 359 (77.2%) 2.01 1.46 – 2.77 <0.0001
<60 634 (57.6%) 81 (12.8%) 553 (87.2%)
gender
male 619 (56.17%) 95 (15.3%) 524 (84.7%) 0.76 0.56 – 1.05 0.101
female 482 (43.83%) 92 (19.1%) 390 (80.9%)
FE
<40% 103 (9.34%) 21 (20.4%) 82 (79.6%) 1.27 0.76 – 2.12 0.343
>40% 994 (90.66%) 166 (16.7%) 828 (83.3%)
CKD (Creat >1,5)
yes 76 (6.89%) 22 (28.9%) 54 (71.1%) 2.12 1.26 – 3.58 0.004
not 1026 (93.11%) 165 (16.1%) 861 (83.9%)
DM
yes 79 (7.16%) 13 (16.5%) 66 (83.5%) 0.96 0.51 – 1.78 0.900
not 1023 (92.84%) 174 (17.0%) 849 (83.0%)
COPD
yes 124 (11.25%) 24 (19.4%) 100 (80.6%) 1.2 0.74 – 1.93 0.45
not 978 (88.75%) 163 (16.7%) 815 (83.3%)
FA
yes 227 (20.59%) 53 (23.3%) 174 (76.7%) 1.68 1.17 – 2.41 0.004
not 875 (79.41%) 134 (15.3%) 741 (84.7%)
CCV
yes 154 (13.97%) 35 (22.7%) 119 (77.3%) 1.54 1.01 – 2.33 0.040
not 948 (86.03%) 152 (16.0%) 796 (84.0%)
Antiarrhythmics
yes 82 (7.44%) 23 (28.0%) 59 (72.0%) 2.03 1.22 – 3.38 0.005
not 1020 (92.56%) 164 (16.1%) 856 (83.9%)
BB
yes 282 (25.58%) 64 (22.7%) 218 (77.3%) 1.66 1.18 – 2.33 0.003
not 820 (74.42%) 123 (15.0%) 697 (85.0%)
Digoxin
yes 353 (32.04%) 65 (18.4%) 288 (81.6%) 1.16 0.83 – 1.61 0.380
not 749 (67.96%) 122 (16.3%) 627 (83.7%)
IC NYHA
3e4 485 (44.01%) 90 (18.6%) 395 (81.4%) 1.22 0.89 – 1.67 0.213
1e2 617 (55.99%) 97 (15.7%) 520 (84.3%)
BB: Beta-blockers, CCV: Cardiovascular Surgery prior, DM: Diabetes Mellitus, CKD: chronic kidney disease, COPD: Chronic Obstructive
Pulmonary Disease, TCP: temporary cardiac pacing (pacemaker transient), AF: atrial fibrillation, EF: left ventricle ejection fraction, FC:
functional class, CI: confidence interval, NCAT: did not use temporary cardiac pacing, OR: odds ratio, NYHA: New York Heart Association,
P: statistical significance
367
evolution
Table 2. Characteristics of the surgical groups and univariate analysis

Variable Total ECAT ÑECAT OR IC 95% P
1102 (100%) 187 (17%) 915 (83%)
VALVES
aortic V
yes 718 (65.16%) 111 (15.5%) 607 (84.5%) 0.74 0.53 – 1.02 0.068
not 384 (34.84%) 76 (19.8%) 308 (80.2%)
mitral V
yes 407 (36.93%) 83 (20.4%) 324 (79.6%) 1.45 1.05 – 2.00 0.020
not 695 (63.07%) 104 (15%) 591 (85%)
V + CRM
yes 190 (17.24%) 41 (21.6%) 149 (78.4%) 1.44 0.97 – 2.12 0.063
not 912 (82.76%) 146 (16%) 766 (84%)
Prosthesis
biological
yes 198 (17.96%) 52 (26.3%) 146 (73.7%) 2.02 1.40 – 2.92 <0.0001
not 904 (82.04%) 135 (14.9%) 769 (85.1%)
calcification
yes 147 (13.33%) 33 (22.4%) 114 (77.6%) 1.50 0.98 – 2.30 0.057
not 955 (86.67%) 154 (16.1%) 801 (83.9)
death
yes 126 (11.43%) 25 (19.8%) 101 (80.2%) 1.24 0.78 – 1.99 0.361
not 976 (88.57%) 162 (16.6%) 814 (83.4%)
CABG: coronary artery bypass grafting, ECAT: temporary cardiac pacing (pacemaker transient), CI: confidence interval, NCAT: did not use
temporary cardiac pacing, OR: odds ratio, P: Statistical Significance, POCC: post-cardiac surgery, V: valve
Table 2 shows the surgical characteristics of patients In this analysis, we observed a higher risk of TCP in
studied with univariate analysis of these data. About 187 POCS in patients undergoing mitral valve replacement (OR =
surgeries requiring TCP, 111 (15.5%) had aortic valve 1.45, 95% CI 1.05 to 2.00, P = 0.02) and patients who received
surgery, 82 (43.8%) for aortic stenosis and 29 (15.5%) bioprosthetic (OR = 2.02, 95% CI 1.4 to 2.92, P <0.0001). The
surgeries for aortic insufficiency, 83 (20. 4%) were mitral 198 patients who received bioprosthetic had higher average
valve surgery, 40 of these (21.4%) for mitral stenosis and 43 age (69.4 ± 13.1 years) compared to the population that did
(23%) for mitral surgery, seven (3.7%) by double not use (52.3 ± 14.8 years). The 52 patients using prosthesis
replacement (aortic + mitral). and required TCP greater mean age (74.1 ± 8.3 years).
Regarding importance, the occurrence of death in the
POCS and the need for TCP showed no statistically
Table 3. Multivariate analysis of factors predisposing to the BAV significant association in univariate analysis (OR = 1.244,
and use of temporary cardiac pacing in POCS valve 95% CI 0.77 to 1.98, P = 0.361).
VARIABLE OR CI 95% P Conducted the data obtained from the multivariate
Age 1.99 1.35 – 2.85 <0.001 analysis (Table 3) showed a significant association of AVB
Atrial fibrillation 1.32 0.86 – 2.01 0.19
with mitral valve surgery (OR = 1.76, 95% CI 1.08 to 2.37, P
CKD 1.67 0.96 – 2.98 0.075
Antiarrhythmics 1.86 1.04 – 3.14 0.026 = 0.002), implantation of a prosthesis biological (OR = 1.59,
Beta-blockers 1.76 1.25 – 2.54 0.002 95% CI 1.02 to 3.91, P = 0.039), age over 60 years (OR = 1.99,
Bioprosthetic 1.59 1.02 – 3.91 0.039 95% CI 1.35 to 2.85, P <0.001), prior use of antiarrhythmic
Mitral valve surgery 1.76 1.08 – 2.37 0.002 drugs (OR = 1.86, 95% CI 1.04-3.14, P = 0.026) and previous
Heart surgery 1.49 0.94 – 2.32 0.080 use of beta-blockers (OR = 1.76, 95% CI 1 0.25 to 2, 54, P =
Length of hospital 1.03 1.01 – 1.04 <0.0001 0.002). Patients with AF, CKD and no prior CVS therefore
CKD: chronic kidney disease, CI: confidence interval, OR: odds presented significant risk for AVB in the POCS.
ratio, P: statistical significance, POCC: post-cardiac surgery Multivariate analysis also revealed that the length of
368
evolution
hospitalization was higher in patients requiring TCP by sinus node artery before and internodal pathways. This
AVB, with a mean hospital stay was 13.59 days compared technique, however, is rarely used in our service.
to 10.88 days who did not need TCP (OR = 1.03 95% CI 1.01 It is remarkable that in our series, aortic valve surgery,
to 1.04, P <0.0001). we found no increased risk of AVB in POCS. This finding
The subgroup of 187 patients with AVB and underwent differs from data in the literature, where no description of
TCP, 14 (7.5%) required implantation of DPM, accounting the incidence of AVB in up to 26%, and these cases need to
for 1.27% of the cohort analyzed. The average time from DPM of 8.5% [7]. Still, these patients were reported factors,
surgery to implantation of DPM was 11.33 days. singly or in combination, could explain a potential increase
in need for TCP [12,13]. The origin of the atrioventricular
DISCUSSION disorder may be the known age-associated aortic valve
disease, mechanical factors (high pressure in the left
The TCP may be necessary in a post-operative cardiac ventricle), histological abnormalities in the conduction
intervention because of the manifest after the AVB system, etc. Clinicopathologic study demonstrated that
procedure. They come as causes of the metabolic block, there is an area particularly at risk near the His bundle region
the residual effect of cardioplegia, edema, inflammation and comprised of non-coronary cusp and its portion adjacent
bleeding near the conduction tissue, anoxia, support to the right coronary artery [14].
therapeutic drug, the iatrogenic injury of the conduction The procedure used in bioprosthetic valve replacement
tissue and fibrosis. The AVB may provide temporary or also showed a statistically significant association in our
permanent. There are no criteria that allow predicting the analysis (OR = 1.59, 95% CI 1.02 to 3.91, P = 0.039). The
evolution of the blockade on its reversibility [6]. average size (median value) in the service of the
Patients who develop the AVB in the POCS generally prostheses used for aortic valve replacement is 23 mm
require TCP and some of DPM to maintain hemodynamic and the mitral valve, 29 mm, and all biological valves are
stability and physiological parameters. In our study, valves used in the service supported. In the literature, is
incidence of AVB with TCP in 17% (187 cases) of the an analysis of type of prosthesis and the risk of permanent
total of 1102 patients undergoing heart valve surgery cardiac pacing for aortic valve replacement, in which the
during the period. However, most AVBs proved to be risk factors found in the type of prosthesis to prosthesis
transient and reversible: only 1.27% of these patients size was smaller than 21 mm [15]. A plausible relation to
developed DPM need to implant this hospital stay, increased risk of TCP with bioprosthesis may be the type
incidence similar to the literature (1.3% to 9.7%) this of implant used in older patients, since the age proved to
association POCS [1-3,7-9]. be a risk factor for TCP.
There is an anatomical proximity to valve structures with The extent of coronary artery disease and the CPB time
the atrioventricular conduction system. Thus, we find could compromise the myocardial protection during
reversible causes for AVB, and the most frequent local edema surgery, facilitating the ischemic injury and / or metabolic
caused by surgical manipulation. This can cause temporary damage by the intrinsic properties of the atrioventricular
changes due to edema of the atrioventricular node, which conduction tissue (differs from cardiac myocytes and
can occur during surgery and provide spontaneous showed less tolerance to the effect of ischemia, to
reversion hours or days after surgery. On the other hand, if hyperkalemia, hypothermia and / or cardioplegia).
there is direct injury of the conduction system (prolonged Specifically, the use of cold potassium cardioplegic solution
ischemia, damage by removal of calcium from the valve may cause temporary blockage of the conduction system
structures, or deep stitches, etc.) Disorder is expected to [2]. These data were not confirmed in our analysis, where
drive greater likelihood of permanent and irreversible [1-3]. the CPB and aortic clamping were not associated with
Specifically, the association between mitral valve higher incidence of need for TCP.
replacement surgery and AVB is still subject to debate, as Age older than 60 years represented a significant risk
well as the mechanism that produces [2,3]. The risk of using factor (OR = 1.99, 95% CI 1.35 to 2.85, P <0.001). The origin
TCP these patients was 20.4 in this casuistic with statistical of this association is likely to encounter the known higher
significance (OR = 1.76, 95% CI 1.08 to 2.37, P = 0.002). As incidence of coronary obstructive component (possibly
already mentioned, the anatomy would be a relevant factor, ischemic) associated with old age and also the fact that
especially the proximal part of the posterior commissure of degenerative diseases of the conduction system are more
the mitral valve structures of the conduction system. frequent in this age group, increasing the likelihood damage
Gaudino et al. [10] and Garcia-Villarreal et al. [11] reported to the conduction system and AVB [1,7,16,17].
that 20% of patients who have replaced the mitral valve The univariate analysis showed patients with prior CVS
using the transseptal biatrial approach required DPM. This risk of AVB (22.7%). However, multivariate analysis did not
surgical approach would relate to the involvement of the confirm this relationship, perhaps by the small number of
369
evolution
patients with this feature in our series. In the literature, the rhythm or nodal escape and good chronotropic response,
risk found for this subgroup is around 5.2% [17]. or subsequent to cardiac surgery with no prospect of
It is unclear the real role of antiarrhythmic drugs in an reversal (< 15 days) as class IIa, level C [19,20].
increased incidence of AVB in the POCS and the literature In the guideline of the American College of Cardiology
is conflicting [2,3]. In our group of patients, the preoperative / American Heart Association, the implementation of the
use of propafenone and / or amiodarone (OR = 1.86, 95% CI DPM is indicated (class I, level C) for total AVB and AVB of
1.04 to 3.14, P = 0.026), as well as beta-blockers (OR = 1.76 the second advanced degree unresolved. The decision to
95% CI 1.25 to 2.54, P = 0.002), proved to be a risk factor for deploy, as well as the wait interval, depends on the discretion
AVB and need to TCP. We justify this association because of the treating physician [21].
this group of drugs able to produce pro-arrhythmic Likewise, the directive of the European Society of
bradicardizanting effects and thus influence the normal Cardiology / Task Force for Cardiac Pacing and Cardiac
function of the conduction system. Since the previous use Resynchronization Therapy recommends (class I, level C)
of digoxin for patients undergoing valve surgery showed implantation of DPM in patients who develop complete
no risk. The long-acting beta-blockers such as atenolol are atrioventricular block or second-degree Mobitz I or II after
associated with higher incidence of blockages in the valve surgery, when it is not expected to resolve the
postoperative period [3]. blockade [22].
As for other elements under review, both the CRD To meet these guidelines, decision-making in most
preoperatively for the presence of previous AF valvular implants performed in the patients, our policy to implant a
surgery showed a trend, but without statistical significance, permanent pacemaker is doing it in cases of AVB
raising the risk of AVB in POCS. presumably irreversible, with an average of 11 days of POCS
The presence of AVB and need for TCP have not valve surgery. Conduct which is in agreement with the
increased mortality in patients POCS valve (OR = 1.244, Brazilian guidelines (class IIa level of evidence C) and
95% CI 0.77 to 1.98, P = 0.361) but significantly prolong the according to AHA / ACC and ESC (class I level of evidence
hospital stay (P <0.0001) at the expense of longer C) [15-19].
hospitalization in the ICU (need to monitor clinical and The risk assessment used in this study is a score
hemodynamic restraint and care, among others) for the developed in the service of post-operative cardiac surgery
normal operation of the temporary pacemaker. This finding at the Hospital São Lucas da PUC-RS, and has been
is relevant, since most patient’s stay in ICU favors the validated in a previous study [5]. Although the
emergence of infections and carries risks of prolonged EuroSCORE is the most widespread, their patient
immobilization [1,3,16,18]. population differs from the Brazilian. The profile of patients
The time elapsed after surgery safer and more necessary undergoing cardiac surgery has changed compared to the
to indicate implant DPM remains uncertain [1]. The literature 70, with the highest percentage of elderly and women,
studies showing early implantation of the definitive higher prevalence of poor cardiac condition and
pacemaker, as Berdajs et al. [3] who studied 391 patients in associated comorbidities [23]. In addition, only 30% of
the period 1990 to 2003 who underwent mitral valve surgery the EuroSCORE were valve surgery. We chose to use the
and found an incidence of 4% of AVB in up to 4 days after score developed in the service.
surgery. In this work, we chose to deploy the DPM when
the patient is able to be discharged from the ICU, and not CONCLUSION
wait for a week. Also the work of Kim et al. [18], we observe
a similar behavior, where in a series of patients undergoing This work highlights the risk factors associated with
valve surgery that developed AVB in the first days after the development of AVB in POCS valve and the need for
surgery, during long-term monitoring, found 56% of patients TCP. However, it should be noted that the potential risk of
depending on cardiac pacing permanently. Thus, if the AVB AVB does not significantly increase the mortality of these
is not resolved within 48 hours, recommended permanent patients when undergoing heart valve surgery, but results
pacemaker implantation up to five days, whereas, in prolonged hospitalization.
presumably, irreversible injury in the conduction system. This study shows the limitations of a retrospective study,
The Brazilian Guidelines for implantable devices, as well although reflecting the “real world” of a large academic
as the Consent to Permanent Cardiac Pacemaker Implant, center. Within this limitation, we should mention the
and implantable defibrillators, recommends (class I level C) potential presence of non-measurable random variables
implantation of a cardiac stimulation device in patients with because of the inherent lack of access to complete data.
asymptomatic AVB, with wide QRS after heart surgery when We must also take into account the results are from a single
persistent (> 15 days), AVB and consequent cardiac surgery, center series, which may represent some degree of bias in
asymptomatic, persistent (> 15 days), with a narrow QRS or treatment. However, another factor to consider is the
370
evolution
absence of definitive data as to the height of the 8. Del Rizzo DF, Nishimura S, Lau C, Sever J, Goldman BS.
atrioventricular conduction disturbance, escape rhythm, Cardiac pacing following surgery for acquired heart disease. J
etc. Card Surg. 1996;11(5):332-40.
Still, the decisions of indication and the time of device
9. Lewis JW Jr, Webb CR, Pickard SD, Lehman J, Jacobsen G.
implantation in our series were not uniform over time, which
The increased need for a permanent pacemarker after
may have led to any deviation from the recommended reoperative cardiac surgery. J Thorac Cardiovasc Surg.
guidelines on the subject [19-22]. 1998;116(1):74-81.
Regarding the results of analysis of implant DPM, we
found only a small number of patients who required the 10. Gaudino M, Alessandrini F, Glieca F, Martinelli L, Santarelli
implantation of this device, making further analysis of data P, Bruno P, et al. Conventional left atrial versus superior septal
for this group of affected patients. However, the results for approach for mitral valve replacement. Ann Thorac Surg.
this group of patients are comparable to data in the literature 1997;63(4):1123-7.
[1-3,7,16,18].
11. García-Villarreal OA, González-Oviedo R, Rodríguez-González
H, Martínez-Chapa HD. Superior septal approach for mitral
valve surgery: a word caution. Eur J Cardiothorac Surg.
2003;24(6):862-7.
12. Fukuda T, Hawley RL, Edwards JE. Lesions of conduction

tissue complicating aortic valvar replacement. Chest.
1976;69(5):605-14.
REFERENCES 13. Davies M, Harris A. Pathological basis of primary heart block.

Br Heart J. 1969;31(2):219-26.
1. Merin O, Ilan M, Oren A, Fink D, Deeb M, Bitran D, et al.
Permanent pacemarker implantation following cardiac surgery: 14. Gannon PG, Sellers RD, Kanjuh VI, Edwards JE, Lillehei CW.
indicatons and long-term follow-up. Pacing Clin Electrophysiol. Complete heart block following replacement of the aortic valve.
2009;32(1):7-12. Circulation. 1966;33(4 Suppl):I152-61.
2. Meimoun P, Zeghdi R, D’Attelis N, Berrebi A, Braunberger E, 15. Elahi M, Usmaan K. The bioprosthesis type and size influence
Deloche A, et al. Frequency, predictors, and consequences of the postoperative incidence of permanent pacemaker
atrioventricular block after mitral valve repair. Am J Cardiol. implantation in patients undergoing aortic valve surgery. J Interv
2002;89(9):1062-6. Card Electrophysiol. 2006;15(2):113-8.
3. Berdajs D, Schurr UP, Wagner A, Seifert B, Turina MI, Genoni 16. Gordon RS, Ivanov J, Cohen G, Ralph-Edwards AL. Permanent
M. Incidence and pathophysiology of atrioventricular block cardiac pacing after a cardiac operation: predicting the use of
following mitral valve replacement and ring annuloplasty. Eur permanent pacemarkers. Ann Thorac Surg. 1998;66(5):1698-704.
J Cardiothorac Surg. 2008;34(1):55-61.
17. Koplan BA, Stevenson WG, Epstein LM, Aranki SF, Maisel
4. Guaragna JCV. Cirurgia cardíaca e hipertensão arterial no pós- WH. Development and validation of a simple risk score to
operatório imediato: fatores pré e transoperatórios [Dissertação predict the need for permanent pacing after cardiac valve
de Mestrado]. Porto Alegre: Pontifícia Universidade Católica surgery. J Am Coll Cardiol. 2003;41(5):795-801.
do Rio Grande do Sul;1999.
18. Kim MH, Deeb M, Eagle KA, Bruckman D, Pelosi F, Oral H,
5. Guaragna JCV, Bodanese LC, Bueno FL, Goldani MA. et al. Complete atrioventricular block after valvular heart surgery
Proposta de escore de risco pré-operatório para pacientes and the timing of pacemarker implantation. Am J Cardiol.
candidatos à cirurgia cardíaca valvar. Arq Bras Cardiol. 2001;87(5):649-51, A10.
2010;94(4):541-8.
19. Lorga AF, Fagundes AA, Barros ARC, De Paola AV, Pedrosa
6. Nascimento CS, Viotti Junior LA, Silva LHF, Araújo AM, A, Grupi CJ et al. Diretrizes Brasileiras de Dispositivos
Bragalha AMLA, Gubolino LA. Bloqueio atrioventricular de Cardíacos Eletrônicos Implantáveis (DCEI). Sociedade
alto grau induzido pela cirurgia cardíaca: estudo de critérios de Brasileira de Arritmias Cardíacas – SOBRAC/SBC,
reversibilidade. Rev Bras Cir Cardiovasc. 1997;12(1):56-61. Departamento de Estimulação Cardíaca Artificial – DECA/
SBCCV. Arq Bras Cardiol. 2007;89(6):e210-37.
7. Dawkins S, Hobson AR, Kalra PR, Tang AT, Monro JL,
Dawkins KD. Permanent pacemarker implantation after 20. Kormann DS, Gauch PRA, Takeda RT, Andrade JCS, Galvão
isolated aortic valve replacement: incidence, indications, and Filho SS, Greco OT, et al. Consenso para implante de
predictors. Ann Thorac Surg. 2008;85(1):108-12. marcapasso cardíaco permanente e desfibrilador – cardioversor
371
evolution
implantável – 1995. Departamento de Estimulação Cardíaca and Antiarrhythmia Devices): developed in collaboration with
Artificial da Sociedade Brasileira de Cirurgia Cardiovascular. the American Association for Thoracic Surgery and Society of
Rev Bras Cir Cardiovasc. 1995;10(2):107-14. Thoracic Surgeons. Circulation. 2008;117(21):e350-408.
21. Epstein AE, DiMarco JP, Ellenbogen KA, Estes NA 3rd, 22. Vardas PE, Auricchio A, Blanc JJ, Daubert JC, Drexler H,
Freedman RA, Gettes LS, et al; American College of Cardiology/ Ector H, et al. Guidelines for cardiac pacing and cardiac
American Heart Association Task Force on Practice Guidelines resynchronization therapy: The Task Force for Cardiac Pacing
(Writing Committee to Revise the ACC/AHA/NASPE 2002 and Cardiac Resynchronization Therapy of the European
Guideline Update for Implantation of Cardiac Pacemakers and Society of Cardiology. Developed in collaboration with the
Antiarrhythmia Devices); American Association for Thoracic European Heart Rhythm Association. Eur Heart J.
Surgery; Society of Thoracic Surgeons. ACC/AHA/HRS 2008 2007;28(18):2256-95.
Guidelines for Device-Based Therapy of Cardiac Rhythm
Abnormalities: a report of the American College of Cardiology/ 23. Cadore MP, Guaragna JCV, Anacker JFA, Albuquerque LC,
American Heart Association Task Force on Practice Guidelines Bodanese LC, Piccoli JCE, et al. Proposição de um escore em
(Writing Committee to Revise the ACC/AHA/NASPE 2002 pacientes submetidos à cirurgia de revascularização miocárdica.
Guidelines Update for Implantation of Cardiac Pacemakers Rev Bras Cir Cardiovasc. 2010;25(4):447-56.
372
The impact of blood transfusion on morbidity and

mortality after cardiac surgery
O impacto da hemotransfusão na morbimortalidade pós-operatória de cirurgias cardíacas
Camila de Christo DORNELES1, Luiz Carlos BODANESE2, João Carlos Vieira da Costa GUARAGNA3, Fabrício
Edler MACAGNAN4, Juliano Cé COELHO5, Anibal Pires BORGES6, Marco Antonio GOLDANI7, João Batista
PETRACCO8
RBCCV 44205-1270
Abstract forward). Were considered significant variables with P <0.05.

Objectives: To analyze the impact of blood transfusion on Results: Patients who received blood transfusions had
the incidence of clinical outcomes postoperatively (PO) from more infectious episodes as mediastinitis (4.9% vs. 2.2%, P
cardiac surgery. <0.001), respiratory infection (27.8% vs 17.1%, P <0.001)
Methods: Retrospective cohort study. We analyzed 4028 and sepsis (6.2% vs. 2.5%, P <0.001). There were more
patients undergoing coronary artery bypass grafting episodes of atrial fibrillation (AF) (27% vs. 20.4%, P <0.001),
(CABG), valve (TV), or both, in Brazilian tertiary university acute renal failure (ARF) (14.5% vs 7.3%, P <0.001) and
hospital between 1996 and 2009. We compared the stroke (4.8% vs. 2.6%, P = 0.001). The length of PO hospital
postoperative complications between patients with blood stay was higher in transfused (13 ± 12.07 days vs. 9.72 ± 7.66
transfusion (n = 916) and non-blood transfusion (n = 3112). days, P <0.001). However, mortality didn’t differ between
Univariate analysis was performed using the Student t test, groups (10.9% vs. 9.1%, P = 0.112). The transfusion was shown
and multivariate logistic regression bivariate (stepwise to be a risk factor for: respiratory infection (OR: 1.91, 95%
1. Physiotherapist, Master’s degree in Health Sciences at PUCRS, 8. Cardiovascular Surgeon at Sao Lucas Hospital, PUCRS, Full
Professor of Physiotherapy at the Lutheran University of Brazil Professor at PUCRS Medical School, PUCRS, Porto Alegre, Brazil.
- Santa Maria Campus / RS, Santa Maria, RS, Brazil.
2. PhD in Clinical Medicine, Full Professor of Cardiology at the Study conducted at the Department of Cardiology, Catholic University
Department of Internal Medicine, PUCRS Medical School, Chief of Rio Grande do Sul (PUCRS) Medical School, Porto Alegre, Brazil.
of Cardiology at Sao Lucas Hospital, PUCRS, Porto Alegre RS,
Brazil. The authors declare they have no funding source or potential conflict
3. PhD in Clinical Medicine; Cardiologist Physician. Professor at of interest that might interfere in the impartiality of this scientific
PUCRS Medical School, Chief of the Postoperative Cardiac work.
Surgery at Sao Lucas Hospital, PUCRS, Porto Alegre, Brazil.
4. Physiotherapist, PhD in Health Sciences at PUCRS, Professor of
Physiotherapy at PUCRS, Porto Alegre, RS, Brazil. Correspondence address:
5. Medicine Undergraduate Student at PUCRS, Porto Alegre, RS, Camila de Christo Dorneles – Departament of Cardiology at Sao
Brazil. Lucas Hospital, PUCRS. Av. 6690 Ipiranga Avenue, room 300 - Porto
6. Medical Internist, Resident at the Department of Cardiology, Alegre, RS, Brazil. Zipcode: 90610-000
Sao Lucas Hospital, PUCRS, Porto Alegre, Brazil. E-mail: camilacdorneles@yahoo.com.br
7. Specialist in Cardiovascular Surgery, Professor at the Department
of Surgery, PUCRS Medical School - Department of Cardiovascular
Surgery, Chief of Cardiovascular Surgery, Sao Lucas Hospital, Article received on November 15th, 2010
PUCRS, Porto Alegre, Brazil. Article accepted on February 21st, 2011
222
DORNELES, CC ET AL - The impact of blood transfusion on Rev Bras Cir Cardiovasc 2011;26(2):222-9
morbidity and mortality after cardiac surgery
CI 1.59-2.29, P <0.001), AF (OR: 1.35, 95% CI 1.13-1.61, P = logística Bivariada (Stepwise Forward). Foram consideradas
0.01), sepsis (OR: 2.08, 95% CI 1.4-3.07, P <0.001), significativas as variáveis com P<0,05.
mediastinitis (OR: 2.14, 95% CI: 1.43-3.21, P <0.001), stroke Resultados: Os pacientes que receberam hemotransfusão
(OR: 1.63, 95% CI 1.1-2.41, P = 0.014) and ARF (OR 1.8, 95% apresentaram mais episódios infecciosos como mediastinite
CI: 1.39-2.33, P <0.001). (4,9% vs. 2,2%, P<0,001), infecção respiratória (27,8% vs
Conclusion: The blood transfusion is associated with 17,1%, P<0,001), e sepse (6,2% vs. 2,5%, P<0,001). Ocorreram
increased risk of infectious events, episodes of AF, ARF and mais episódios de fibrilação atrial (FA) (27% vs. 20,4%,
stroke, as well as the increased length of hospital stay but P<0,001), insuficiência renal aguda (IRA) (14,5% vs. 7,3%,
not mortality. P<0,001) e acidente vascular cerebral (AVC) (4,8% vs. 2,6%,
P=0,001). O tempo de internação hospitalar no PO foi maior
Descriptors: Blood Transfusion. Postoperative nos transfundidos (13±12,07 dias vs. 9,72±7,66 dias, P<0,001).
Complications. Hospital Mortality. Cardiac Surgical Porém, a mortalidade não apresentou diferença entre os
Procedures. grupos (10,9% vs. 9,1%, P=0,112). A transfusão mostrou-se
como fator de risco para: infecção respiratória (OR: 1,91;
IC95%: 1,59-2,29; P<0,001), FA (OR:1,35; IC95%: 1,13-1,61;
Resumo P=0,01), sepse (OR: 2,08; IC95%: 1,4-3,07; P<0,001),
Objetivos: Analisar o impacto da hemotransfusão mediastinite (OR: 2,14; IC95%: 1,43-3,21; P<0,001), AVC
sanguínea na incidência de desfechos clínicos no pós- (OR: 1,63; IC95%: 1,1-2,41; P=0,014) e IRA (OR: 1,8; IC95%:
operatório (PO) de cirurgias cardíacas. 1,39-2,33; P<0,001).
Métodos: Estudo de coorte retrospectiva. Foram analisados Conclusão: A hemotransfusão está associada ao aumento
4.028 pacientes submetidos à cirurgia de revascularização do risco de eventos infecciosos, episódios de FA, IRA e AVC,
miocárdica (CRM), troca valvar (TV) ou ambas, em hospital bem como aumentou o tempo de permanência hospitalar,
terciário universitário brasileiro, entre 1996 e 2009. Foram mas não a mortalidade.
comparadas as complicações no PO entre os pacientes
hemotransfundidos (n=916) e não-hemotransfundidos Descritores: Transfusão de Sangue. Complicações Pós-
(n=3112). Foi realizada análise univariada através do teste t Operatórias. Mortalidade Hospitalar. Procedimentos
de Student, e análise multivariada com o uso de regressão Cirúrgicos Cardíacos.
INTRODUCTION occurrence of renal failure and infection, as well as

respiratory, cardiac and neurological complications in
Blood transfusion and hemocomponents are important transfused patients compared to those who were not
technologies in modern therapy. If it is appropriately used transfused after cardiac surgery [7-9].
in significant morbidity or mortality conditions, and when The occurrence of bleeding is a frequent complication
its not effectively prevented or controlled in any other way, of cardiovascular surgery [10]. Approximately one-third of
it can save lives and improve patients’ health. However, as the operated patients require blood transfusion, being
any other therapeutic interventions, it can lead to acute or responsible for the consumption of 10% to 25% of the
late complications, such as the risk for transmission of hemoderivatives in the United States [10-13].
infectious agents, among other clinical complications [1]. The aim of this study was to recognize the risk factors
Blood transfusion is related to the occurrence of a associated with blood transfusion in the postoperative
transfusion reaction, transmission of infection, increased period of cardiac surgery, thus, it allows strategies to be
postoperative morbidity and mortality, risk of drawn to reduce these comorbidities in this group of
immunosuppression and the cost of hospitalization [2-4]. patients.
Cardiac surgery is associated with high rates of blood
transfusion, ranging from 40% to 90% in most studies [5- METHODS
7]. Transfusions have been associated with high morbidity
and mortality in patients and some recent studies have We conducted a retrospective cohort study using data
demonstrated worse outcomes, including increased from the Cardiovascular Surgery Department of Sao Lucas
223
Hospital at PUCRS, from 4028 patients who underwent patients receiving hemoderivatives or not were also
coronary artery bypass graft (CABG), valve replacement analyzed.
(VR) or combined surgeries (CABG and VR) between The quantitative variables were described by mean and
January 1996 and December 2009. Patients who received standard deviation and the categorical variables were
two or more units of red packed blood cells in the described by means of absolute and relative frequencies.
postoperative period (PO) of the surgeries. The patients In addition to the descriptive analysis, univariate analysis
received blood transfusions when there was excessive was performed from the Student t test for the quantitative
bleeding after surgery (above 400 ml in the first variables and Pearson’s chi-square test and / or Fisher exact
postoperative hours or above 100 ml / hour for the first 6 test for the categorical variables. To assess the possible
postoperative hours) requiring surgical intervention. intervening variables, multivariate analysis was performed
The pre and perioperative variables analyzed were: by the bivariate logistic regression (stepwise forward), in
female gender, age over 65 years, smoking, obesity (BMI which all pre-and perioperative factors were adjusted as
e” 30kg / m²), chronic obstructive pulmonary disease possible confounding factor, with P <0.1 for entry logistic
(COPD) clinically diagnosed and / or through chest X-ray regression. The variables with P <0.05 were considered
and / or spirometry, and / or drug therapy (corticosteroids significant. The software SPSS (Statistical Package for the
or bronchodilators), chronic renal failure (creatinine> 1.5 Social Sciences, version 18.0) were used for statistical
mg / dL or prior hemodialysis), active infectious process analysis.
(through laboratory tests), prior cardiac surgery, history of This study research project was submitted and approved
previous diseases as diabetes (DM), systemic arterial by the PUCRS Research Ethics Committee.
hypertension (SAH), peripheral vascular disease (PVD),
cerebrovascular accident (CVA), atrial fibrillation (AF) and RESULTS
acute myocardial infarction (AMI), unstable angina (UA),
functional class 3 and 4 of heart failure according to criteria In this retrospective study, 4.028 patients were included,
of the New York Heart Association (NYHA), ejection fraction 62.9% (n = 2,533) were male and 37.1% (n = 1,495) were
below 40% (measured by echocardiography), surgical female. Regarding the type of surgery, 67.2% (n = 2706)
priority (emergency surgery / emergency as a single variable underwent CABG, 27.4% (n = 1,102) CTV and 5.6% (n = 225)
and defined as the need for intervention up to 48 hours due underwent combined surgery (CABG + VR). Considering
to imminent risk of death or unstable hemodynamic these patients, 22.7% (n = 916) required blood transfusion
condition), extracorporeal circulation time over 120 minutes, in the postoperative period of cardiac surgery. Only 139
use of intra-aortic balloon pump (IABP). (3.45%) patients in the study sample underwent surgery
The postoperative variables analyzed as possible without extracorporeal circulation. In Table 1, the patients
complications related to transfusion of hemoderivatives pre-and perioperative data are shown.
were: sepsis and respiratory infection diagnosed by isolated Ten postoperative complications were evaluated in
organisms in culture associated with fever and elevated relation to the use of packed red blood cells (RBC) (Table
white blood cell count; mediastinitis diagnosed from the 2). Patients who received blood transfusions had
presence of pain, warmth, redness, and pus in the wound significantly more infectious episodes as mediastinitis
of the sternum, presence of sternal instability and fever, as (4.9% vs. 2.2%; P = 0.001) respiratory infection (27.8% vs.
well as through a tomography of the thorax, postoperative 17.1%; P = 0.001), and sepsis (6.2% vs. 2.5%; P = 0.001).
AMI, diagnosed by the evidence of subepicardial current This group of patients also had more episodes of AF (27%
with the onset of Q wave, subendocardial injury current vs. 20.4%; P = 0.001), ARF (14.5% vs. 7.3%; P = 0.001) and
with increased markers of myocardial necrosis (troponin I CVA (4.8% vs. 2.6%; P = 0.001). AMI and ARDS were not
above 10ìg/dL and CK-MB five times higher than the statistically significant, but showed a tendency. Moreover,
reference value or above 10% of total CK) or new bundle the length of hospital stay after surgery was higher in the
branch block with elevation of markers, atrial fibrillation group that received blood transfusion (13 ± 12.07 days vs.
(AF) confirmed by electrocardiographic analysis, acute renal 9.72 ± 7.66 days; P = 0.001). However, mortality did not
failure (ARF), with an increase of creatinine and need for differ between patients who received RBC as compared to
dialysis, acute respiratory distress syndrome (ARDS), with those who were not transfused (10.9% vs. 9.1%; P = 0.112).
chest radiograph showing alveolar-interstitial infiltrates, The postoperative complications were included in
micro and / or macronodular, bilateral and asymmetric and multivariate analysis (Table 3), and adjusted for pre-and
an oxygenation index (PaO2/FiO2) <200 mmHg, perioperative factors mentioned above, and were related
cerebrovascular accident (CVA), diagnosed by clinical signs as risk factors due to blood transfusion: respiratory
of neurological deficit and CT compatible; the death rate infection (OR: 1.91; CI95%: 1.59-2.29; P<0.001), ARDS (OR:
within 30 days after surgery and hospitalization time in 2.35; CI95%: 0.97-5,67; P=0.058), (NRAF (OR:1.35; CI95%:
224
Table 1. Baseline characteristics of patients in pre and peri- 1.13-1.61; P<0.01), sepsis (OR: 2.08; CI95%: 1.4-3.07;
operative period of cardiac surgery. P<0.001), mediastinitis (OR: 2.14; CI95%: 1.43-3.21; P<0.001),
Variables Patients who Patients who did P CVA (OR: 1.63; CI95%: 1.1-2.41; P=0.014) and ARF (OR: 1.8;
underwent not undergo
hemotransfusion hemotransfusion CI95%: 1.39-2.33 days; P <0.001). AMI and death rate were
n=916 n = 3112 not statistically significant.
n(%) n(%)
Female gender 393 (43) 1102 (35.4) <0.001
Age > 65 years 375 (40.9) 1074 (34.5) <0.001
Smoking 255 (27.8) 1027 (33) 0.004
Obesity 94 (10.3) 300 (9.6) 0.622
DM 237 (25.9) 680 (21.8) 0.012 Table 3. Multivariate analysis of postoperative complications
COPD 138 (15.1) 524 (16.8) 0.222 due to blood transfusion.
CRF 108 (11.8) 210 (6.7) <0.001 Complications OR 95% CI P
CVA 52 (5.7) 180 (5.8) 0.967 Respiratory infection 1.91 1.59 – 2.29 <0.001
Pre-hemodialysis 20 (2.2) 26 (0.8) 0.001 Post AMI 0.93 0.74 – 1.18 0.567
Preinfection 58 (6.3) 185 (5.9) 0.724 AF 1.35 1.13 – 1.61 0.01
UA 285 (31.1) 832 (26.7) 0.01 Sepsis 2.08 1.40 – 3.07 <0.001
AMI 303 (33.1) 939 (30.2) 0.103 Mediastinitis 2.14 1.43 – 3.21 <0.001
AF 69 (7.5) 240 (7.7) 0.913 ARF 1.8 1.39 – 2.33 <0.001
SAH 616 (67.2) 1864 (59.9) <0.001 ARDS 2.35 0.97 – 5.67 0.058
Ef < 40% 768 (83.8) 2685 (86.3) 0.072 CVA 1.63 1.1 – 2.41 0.014
EC>120min 196 (21.4) 357 (11.5) <0.001 Death 0.85 0.64 – 1.13 0.259
PVD 95 (10.4) 230 (7.4) 0.004
IABP 80 (8.7) 216 (6.9) 0.079 OR: odds ratio, CI 95%: 95% confidence interval, P: statistical
Cir. Emergency surgery 16 (1.7) 26 (0.8) 0.028 significance; AF: Atrial Fibilation; ARF: Acute Renal Failure;
CHF 3 and 4 (NYHA) 234 (25.5) 667 (21.4) 0.03 ARDS: Acute Respiratory Distress Syndrome; Post AMI: AMI up
to 30 days after surgery; CVA? Cerebrovascular Accident
P: statistical significance; DM: Diabetes Mellitus, COPD: Chronic
Obstructive Pulmonary Disease; CRF: Chronic Renal Failure;
CVA: Cerebrovascular Accident; UA: Unstable angina; AMI: Acute
Myocardial Infarction; AF: Atrial Fibilation; SAH: Systemic Arterial
Hypertension; Ef: Ejection fraction; EC: Extracorporeal
Circulation; PVD: Peripheral Vascular Disease; IABP: Intra-aortic
Balloon Pump; CHF: Congestive Heart Failure; NYHA: New York
DISCUSSION
Heart Association
The study results suggest that blood transfusion in the
studied sample is related to the risk of infectious events
(such as mediastinitis, respiratory infection, sepsis). The
transfused patients had also more episodes of AF, ARF
and CVA. ARDS appears as a tendency to risk factor and
Table 2. Complications presented in the postoperative period. can not be dismissed.
Complications Patients who Patients who did not P The association between packed red cells transfusion
underwent undergo and postoperative complications (PO) for cardiac surgery
hemotransfusion hemotransfusion is often described in the literature. In a study conducted by
n=916 n=3112
n(%) n(%) Koch et al. [14], when analyzing the ratio of units of packed
AF 247 (27) 635 (20.4) <0.0001 red cells administered to adult patients and the risk of
Sepsis 57 (6.2) 77 (2.5) <0.0001 morbidity and mortality in patients undergoing CABG, it
Mediastinitis 45 (4.9) 68 (2.2) <0.0001 was concluded that from the 11,963 patients analyzed, 5,184
Respiratory infection 255 (27.8) 532 (17.1) <0.0001
ARF 133 (14.5) 227 (7.3) <0.0001 (49%) received at least one unit of RBC.
ARDS 10 (1.1) 12 (0.4) 0.022 The authors also concluded that there is a relationship
Post AMI 123 (13.4) 366 (11.8) 0.193 between each RBC unit transfused and the risk of
CVA 44 (4.8) 81 (2.6) 0.001 morbimortality in the postoperative period. According to
Death 100 (10.9) 283 (9.1) 0.112
HS* 13±12.07 9.72±7.66 <0.0001 the study, each administered unit increases by 77% the
postoperative mortality risk and 100% the risk of developing
*Days. Mean ± Standard deviation.
P: statistical significance; AF: Atrial Fibilation; ARF: Acute Renal any renal complication, 76% the risk of developing
Failure; ARDS: Acute Respiratory Distress Syndrome; Post AMI: postoperative infections, cardiac complications in 55% and
AMI up to 30 days after surgery; CVA: Cerebrovascular Accident; 37% the risk of neurological complications. These
HS: Hospital Stay associations remained strong even after the risk adjustment
225
factors causing postoperative complications. The authors respiratory, digestive tract, skin and subcutaneous tissue,
also reported the importance of stratifying high-risk groups as well as sepsis. In this study, the risk of infections during
for transfusion and then, modulate the pre-and hospitalization in transfused patients was two times higher
perioperative interventions that help to reduce blood compared to those who did not receive hemoderivatives.
transfusion in the postoperative period, and, consequently, In this group of patients, the risk of death was five times
the risk of complications. higher in the case of elective surgery and four times higher
In a study conducted by Möhnle et al. [15] in patients in case of non-elective surgeries, as well as the risk of death
who had low risk of developing postoperative complications after 30 days of hospital discharge was three times higher
of cardiac surgery, it was observed that, despite the low in patients undergoing elective surgery and four times
risk patients who received blood transfusion had an higher in non-elective surgeries.
increased risk of developing cardiac events, as well as an In Magedanz study et al. [18], in which a risk score for
increase in surgical site infections. However, the authors mediastinitis was created, an association between blood
report that the study’s findings related to infectious, renal transfusion and risk of postoperative mediastinitis was
complications and hospital mortality should be carefully found (OR: 2.5; CI95%: 1.5 - 4.1, P = 0.001), that also agrees
analyzed, mainly due to the fact that the group of transfused with a study that found a strong association between
patients differs in several characteristics from the group transfusion and risk of infections, which reflects the
that did not received blood transfusions in the immunosuppressive effect of blood transfusion [19].
postoperative period. The study conducted shows that transfusion of
Karkouti et al. [16] also reported that the need for blood hemocomponents is directly related to increased risk of
transfusion in the postoperative period of cardiac surgeries infectious processes in the postoperative period of cardiac
is directly associated with complications such as sepsis, surgery and the onset of AF during this period.
ARDS, ARF and death. Studies have reported that blood transfusion,
In this study, a tendency toward higher pulmonary particularly platelets, is associated with the development
complication rates in transfused patients was noted, of comorbidities such as AF, low cardiac output syndrome,
reaching statistical significance for lower respiratory tract AMI, CVA, renal failure and sepsis [14,16].
infection, but not for ARDS. The design of the study, Atrial fibrillation is a common complication after cardiac
however, did not allow to differentiate ARF due to TRALI surgery, ranging from 10% to 43% of the operated patients,
(Transfusion-related Acute Lung Injury) from other causes contributing to morbidity and increased hospital stay of
such as ARDS and lower respiratory tract infection. these patients. Koch et al. [20] in a study that evaluated the
Specifically in relation to risk of pulmonary risk of developing AF in patients undergoing blood
complications, Koch et al. [17], in another study reported transfusion after cardiac surgery, it was concluded that
that blood transfusion is associated with high risk of ARF, transfusion is associated with the risk of AF in the
ARDS, need for reintubations due to pulmonary causes, postoperative period, and the risk increases for each unit
long-term intubations and longer hospitalization period in transfused (OR: 1.2; 95% CI: 1.1-1.3 days; P <0.0001).
the intensive care unit. The same study concludes that the Although the precise mechanism is not known, the authors
use of plasma is associated with more pulmonary speculate that the inflammatory response associated with
complications after surgery when compared to other transfusion may be exacerbated, contributing to the
hemoderivatives. occurrence of AF. This may be due to leukocyte activation,
Another acute pulmonary complication caused by leading to injury of the atrial tissue.
blood transfusion, known as TRALI, the receiver of Some studies report that the ischemic process, blood
hemoderivatives develops signs and symptoms of loss and blood transfusion are known to cause severe
respiratory compromise due to a non-cardiogenic systemic inflammatory response. Blood transfusion can also
pulmonary edema. Evidence of hypoxemia with a ratio of initiate a secondary inflammatory response by modifying
PaO2/FiO2 <300, associated with bilateral pulmonary the systemic inflammatory response of the patients and
infiltrates and without fluid overload, are diagnostic criteria the direct introduction of bioactive substances into
for TRALI. But the most common criterion for diagnosis is circulation, in addition to the primary inflammatory response
the temporal criterion between the transfusion and the onset initiated by the extracorporeal circulation [21-23].
of signs and symptoms. They usually occur 1 or 2 hours In this study, patients who received hemoderivatives
after the transfusion, with 100% of the patients presenting required longer hospital stays, however, there was no
the symptoms 6 hours after transfusion. difference in mortality compared to patients who were not
Rogers et al. [6] found a significant increase in infections transfused.
during hospitalization in patients who received blood In the study conducted by Hajjar et al. [7], 502 patients
transfusion, and the most prevalent are the genitourinary, who underwent cardiac surgery with extracorporeal
226
circulation (EC) were randomized to receive blood also reduce bleeding during surgery and in the
transfusion in a liberal manner (to maintain a hematocrit postoperative period. The autologous blood donation is
e” 30%) or in a more restrictive way (hematocrit e” 24%). another method used, but still requires further studies, when
This study demonstrates that a more conservative blood the patient’s blood is collected prior to surgery in order to
transfusion does not alter the rates of comorbidities (such be used after the procedure. It is considered a simple,
as ARDS, ARF or cardiogenic shock) and mortality inexpensive, safe and effective method to reduce
between the groups, demonstrating the importance of a homologous blood transfusion in a wide variety of elective
more careful selection of patients who must undergo a surgical procedures, including cardiac surgeries [4.29].
blood transfusion. New hemoconcentration techniques during EC have
In a study conducted by McGrath et al. [24], with 32,298 also been studied. Souza & Braille [30] propose a new
patients, no association between transfusion and method of hemoconcentration, which revealed a reduction
increased morbidity and mortality after cardiac surgery of blood and plasma used during and after EC, as well as a
was found. However, this study analyzed only the platelet reduced water balance compared to the group that did not
transfusion, regardless of other hemoderivative use the hemoconcentrator.
components. In this study, however, there were some limitations. The
According to the study conducted by van Straten et al. number of units that were transfused in each patient or the
[21], the number of units transfused may be a predictor of effects of other hemoderivatives in the postoperative period
early mortality (up to 30 days in the postoperative period), was not taken into consideration. Another factor that was
but not of late mortality. Comparing the expected survival, not taken into account was the storage time of the blood
the authors concluded that in patients who received three transfusion bags. Thus, we conclude that the data collected
or more units of RBC, the survival rate drops significantly revealed a tendency demonstrating that more detailed
when compared to patients who did not receive studies are needed to show the effects of each
hemoderivative products. These patients have also high hemocomponent transfused in patients undergoing cardiac
incidence of postoperative complications, which could surgery.
explain the high early mortality in this study.
By analyzing the relationship between blood CONCLUSION
transfusion and cardiac surgery in an older population,
Veenith et al. [25] showed that blood transfusion is The analysis of this study results shows that blood
associated with a significant increase in mortality and transfusion is associated with increased risk of infectious
length of hospital stay. The authors also report that events (such as mediastinitis, respiratory infection, sepsis),
this possible association is a result of the poor health risk of developing AF, ARF, CVA and ARDS. The blood
status of patients in the preoperative period. Another transfusion increased the length of hospital stay, but not
factor that increases the risk of postoperative mortality in these patients.
morbimortality would be that the blood transfusion may
increase the risk of ischemic processes and infectious
complications. This would also explain the long-term
ICU and hospital stay.
Some studies also report strategies for reducing the use
of hemoderivatives in postoperative cardiac surgery. Some
authors report the importance of a careful preoperative
evaluation, because it can reduce the risk of bleeding and
the need for blood transfusion during the postoperative
period. The measurement of serum iron and iron
administration and preoperative erythropoietin may reduce
the need for hemoderivatives. The perioperative
interventions play an important role in reducing the risk of
bleeding during and after surgery [4]. The use of
antifibrinolytic, such as aprotinin, attenuates the systemic
inflammatory response in patients undergoing EC [26]. The REFERENCES
use of miniEC (reduction of the prime, the tubes and vacuum
1. Ministério da Saúde (Brasil). Secretaria de Atenção à Saúde.
drainage) and the use of centrifugal pumps (Bio Pump) can Departamento de Atenção Especializada. Guia para o uso de
reduce the excessive use of heparin [27,28]. Hypothermia hemocomponentes. Série A. Normas e Manuais Técnicos.
(between 30° and 32° C) during the surgical procedure may Brasília: Ministério da Saúde;2008. 140p.
227
2. Alghamdi AA, Davis A, Brister S, Corey P, Logan A. 14. Koch CG, Li L, Duncan AI, Mihaljevic T, Cosgrove DM,
Development and validation of Transfusion Risk Loop FD, et al. Morbidity and mortality risk associated with
Understanding Scoring Tool (TRUST) to stratify cardiac red blood cell and blood-component transfusion in isolated
surgery patients according to their blood transfusion needs. coronary artery bypass grafting. Crit Care Med.
Transfusion. 2006;46(7):1120-9. 2006;34(6):1608-16.
3. Ferraris VA, Ferraris SP. Limiting excessive postoperative blood 15. Möhnle P, Snyder-Ramos SA, Miao Y, Kulier A, Böttiger BW,
transfusion after cardiac procedures. A review. Tex Heart Inst Levin J, et al; Multicenter Study of Perioperative Ischemia
J. 1995;22(3):216-30. (McSPI) Research Group. Postoperative red blood cell
transfusion and morbid outcome in uncomplicated cardiac
4. Souza HJB, Moitinho RF. Estratégias para redução do uso de surgery patients. Intensive Care Med. 2011;37(1):97-109.
hemoderivados em cirurgia cardiovascular. Rev Bras Cir
Cardiovasc. 2008;23(1):53-9. 16. Karkouti K, O’Farrell R, Yau TM, Beattie WS; Reducing
Bleeding in Cardiac Surgery Research Group. Prediction of
5. Stover EP, Siegel LC, Parks R, Levin J, Body SC, Maddi R, et massive blood transfusion in cardiac surgery. Can J Anaesth.
al. Variability in transfusion practice for coronary artery bypass 2006;53(8):781-94.
surgery persists despite national consensus guidelines: a 24-
institution study. Institutions of the Multicenter Study of 17. Koch C, Li L, Figueroa P, Mihaljevic T, Svensson L, Blackstone
Perioperative Ischemia Research Group. Anesthesiology. EH. Transfusion and pulmonary morbidity after cardiac
1998;88(2):327-33. surgery. Ann Thorac Surg. 2009;88(5):1410-8.
6. Rogers MA, Blumberg N, Saint S, Langa KM, Nallamothu 18. Magedanz EH, Bodanese LC, Guaragna JC, Albuquerque LC,
BK. Hospital variation in transfusion and infection after cardiac Martins V, Minossi SD, et al. Risk score elaboration for
surgery: a cohort study. BMC Med. 2009;7:37. mediastinitis after coronary artery bypass grafting. Rev Bras
Cir Cardiovasc. 2010;25(2):154-9.
7. Hajjar LA, Vincent JL, Galas FR, Nakamura RE, Silva CM,
Santos MH, et al. Transfusion requirements after cardiac 19. Friedman ND, Bull AL, Russo PL, Leder K, Reid C, Billah B,
surgery: the TRACS randomized controlled trial. JAMA. et al. An alternative scoring system to predict risk for surgical
2010;304(14):1559-67. site infection complicating coronary artery bypass graft
surgery. Infect Control Hosp Epidemiol. 2007;28(10):1162-8.
8. Leal-Noval SR, Rincón-Ferrari MD, García-Curiel A, Herruzo-
Avilés A, Camacho-Laraña P, Garnacho-Montero J, et al. 20. Koch CG, Li L, Van Wagoner DR, Duncan AI, Gillinov AM,
Transfusion of blood components and postoperative infection Blackstone EH. Red cell transfusion is associated with an
in patients undergoing cardiac surgery. Chest. increased risk for postoperative atrial fibrillation. Ann Thorac
2001;119(5):1461-8. Surg. 2006;82(5):1747-56.
9. Engoren MC, Habib RH, Zacharias A, Schwann TA, Riordan 21. van Straten AH, Bekker MW, Soliman Hamad MA, van Zundert
CJ, Durham SJ. Effect of blood transfusion on long-term survival AA, Martens EJ, Schönberger JP, et al. Transfusion of red blood
after cardiac operation. Ann Thorac Surg. 2002;74(4):1180-6. cells: the impact on short-term and long-term survival after
coronary artery bypass grafting, a ten-year follow-up. Interact
10. Whitlock R, Crowther MA, Ng HJ. Bleeding in cardiac surgery: Cardiovasc Thorac Surg. 2010;10(1):37-42.
its prevention and treatment: an evidence-based review. Crit
Care Clin. 2005;21(3):589-610. 22. Murphy GJ, Reeves BC, Rogers CA, Rizvi SI, Culliford L,
Angelini GD. Increased mortality, postoperative morbidity,
11. Karkouti K, Cohen MM, McCluskey SA, Sher GD. A and costs after red blood cell transfusion in patients having
multivariable model for predicting the need for blood cardiac surgery. Circulation. 2007;116(22):2544–52.
transfusion in patients undergoing first-time elective
coronary bypass graft surgery. Transfusion. 23. Ranucci M. Allogeneic blood transfusions and infections after
2001;41(10):1193-203. cardiac surgery. Am Heart J. 2007;153(5):e21.
12. Johnson RG, Thurer RL, Kruskall MS, Sirois C, Gervino EV, 24. McGrath T, Koch CG, Xu M, Li L, Mihaljevic T, Figueroa P,
Critchlow J, et all. Comparison of two transfusion strategies et al. Platelet transfusion in cardiac surgery does not confer
after elective operations for myocardial revascularization. J increased risk for adverse morbid outcomes. Ann Thorac Surg.
Thorac Surg. 1992;104(2):307-14. 2008;86(2):543-53.
13. Magovern JA, Sakert T, Benckart DH, Burkholder JA, Liebler 25. Veenith T, Sharples L, Gerrard C, Valchanov K, Vuylsteke A.
GA, Magovern GJ Sr, et al. A model for predicting transfusion Survival and length of stay following blood transfusion in
after coronary artery bypass grafting. Ann Thorac Surg. octogenarians following cardiac surgery. Anaesthesia.
1996;61(1):27-32. 2010;65(4):331-6.
228
26. Greilich PE, Brouse CF, Whitten CW, Chi L, Dimaio JM, Bergh K. Heparin-coated circuit during cardiopulmonary
Jessen ME. Antifibrinolytic therapy during cardiopulmonary bypass. A clinical study using closed circuit, centrifugal pump
bypass reduces proinflammatory cytokine levels: a and reduced heparinization. Acta Anaesthesiol Scand.
randomized, double-blind, placebo-controlled study of epsilon- 1994;38(4):372-9.
aminocaproic acid and aprotinin. J Thorac Cardiovasc Surg.
2003;126(5):1498-503. 29. Yoda M, Nonoyama M, Shimakura T. Autologous blood
donation before elective off-pump coronary artery bypass
27. Perthel M, El-Ayoubi L, Bendisch A, Laas J, Gerigk M. Clinical grafting. Surg Today. 2004;34(1):21-3.
advantages of using mini-bypass system in terms of blood
product use, postoperative bleeding and air entrainment: an in 30. Souza DD, Braile DM. Avaliação de nova técnica de
vivo clinical perspective. Eur J Cardiothorac Surg. hemoconcentração e da necessidade de transfusão de
2007;31(6):1070-5. hemoderivados em pacientes submetidos à cirurgia cardíaca
com circulação extracorpórea. Rev Bras Cir Cardiovasc.
28. Sellevold OF, Berg TM, Rein KA, Levang OW, Iversen OJ, 2004;19(3):287-94.
229
Rev. Chilena de Cirugía. Vol 64 - Nº 3, Junio 2012; pág. 319-320
REVISTA DE REVISTAS
Tendencia en volumen hospitalario y morta- concentración de estos casos en menos hospitales

lidad operatoria para cirugías de alto riesgo. pero para los RVA el aumento se debió al mayor nú-
Finks JF et al. N Engl J Med 2011; 364: 2128-37 mero nacional. Esto significa que los centros de alto
volumen pasaron a concentrar la cirugía de esófago
En Estados Unidos de Norteamérica, en la última de un 39% a un 45%; para páncreas de 42% a 51%;
década, se han desarrollado tendencias orientadas a en caso de pulmón de 39% a 41% y de cistectomías
concentrar ciertas cirugías de alto riesgo (CAR), en de 35% a 47%. Al contrario, en el mismo período
centros de alto volumen (CAV), al haberse obser- disminuyó el volumen hospitalario de RC y EC.
vado una relación inversamente proporcional entre Para todos los procedimientos quirúrgicos ana-
ambos factores: a mayor volumen menor mortalidad. lizados la mortalidad disminuyó significativamente
Sin embargo, existen barreras para lograr este propó- a lo largo de los 10 años de observación. El mayor
sito o, y no se ha analizado el alcance que ha tenido volumen hospitalario explica una gran parte de la
esta iniciativa ni el impacto que ha tenido sobre la disminución de la mortalidad por cirugía de pán-
mortalidad operatoria. creas, vejiga y esófago, en cifras de 67%, 37% y
Los autores, sobre la base del registro de Medicare, 32% respectivamente. Para estos tres casos, este
analizaron qué sucedió en este sentido, entre los años efecto de declinación de la mortalidad es explicado
1999 y el 2008, para cirugía resectiva por cáncer de por la concentración de las cirugías. Una pequeña
esófago, pulmón, páncreas y vejiga y en la cirugía proporción de esta disminución es atribuible al
electiva por un aneurisma aórtico abdominal (AAA), aumento del volumen hospitalario en el caso de
revascularización coronaria (RC), endarterectomía resección pulmonar, reparación de AAA y RVA.
carotídea (EC) y reemplazo valvular aórtico (RVA). Para la mortalidad asociada a RC y EC el volumen
Para estos efectos se midió la mortalidad operatoria hospitalario no jugó un rol.
definida como la muerte ocurrida durante la hospi- Los autores concluyen que los adelantos técnicos
talización o hasta los 30 días del postoperatorio. Se en general, los mejores cuidados intensivos, los
analizaron datos de más de 3,2 millones de pacientes. cuidados perioperatorios y políticas de check-list
Durante el período, el volumen medio hospitala- explican la disminución global de la mortalidad
rio de casos operados por los cánceres seleccionados quirúrgica, pero en el caso de cirugía de páncreas,
aumentó sustancialmente, como también el de repa- esófago y vejiga hay evidencia fuerte que asocia me-
ración de AAA y en menor extensión el volumen de jor resultado con mayor volumen. Con todo, existe
RVA. Para la cirugía de esófago, aumentó de 4 a 7; un amplio espacio para mejorar.
para páncreas de 5 a 13; para pulmón de 18 a 23; en
el caso de cistectomías de 5 a 8. La media de repa- Dr. Ricardo Espinoza G.
raciones de AAA aumentó de 22 a 33. Las razones Universidad de los Andes
del aumento para las esofagectomías se debió a la Santiago, Chile
319
Documento descargado de http://www.elsevier.es el 14/01/2013. Copia para uso personal, se prohíbe la transmisión de este documento por cualquier medio o formato.
An Pediatr (Barc). 2012;77(6):366---373
www.elsevier.es/anpediatr
ORIGINAL
Marcadores pronósticos de mortalidad en el postoperatorio de las

cardiopatías congénitas
J.A. García-Hernández a,∗ , I.L. Benítez-Gómez a , A.I. Martínez-López b ,
J.M. Praena-Fernández c , J. Cano-Franco a y M. Loscertales-Abril a
a
Unidad de Gestión Clínica de Cuidados Críticos y Urgencias, Hospital Infantil Universitario Virgen del Rocío, Sevilla, España
b
Equipo Básico de Atención Primaria, Centro de Salud de Alcosa, Sevilla, España
c
Unidad de Apoyo a la Investigación, Hospital Virgen del Rocío, Sevilla, España
Recibido el 7 de noviembre de 2011; aceptado el 21 de marzo de 2012

Disponible en Internet el 15 de junio de 2012
PALABRAS CLAVE Resumen

Cardiopatías Introducción: Identificar los factores de riesgo de mortalidad postoperatoria para establecer
congénitas; indicaciones de oxigenación con membrana extracorpórea en niños cardiópatas.
Cirugía; Pacientes y métodos: Entre abril del 2007 y junio del 2009 fueron intervenidos 186 niños con
Circulación circulación extracorpórea. Se determinaron en sangre arterial y venosa, al ingreso en UCIP y a
extracorpórea; las 22 horas, el pH, la pCO2 , la SatO2 y el exceso base, y el CO2 en aire espirado. El lactato se
Asistencia midió en quirófano, al ingreso en UCIP y durante el postoperatorio, para determinar el tiempo
circulatoria en que se mantuvo elevado, su pico máximo, y la velocidad de variación. Se calculó además,
la diferencia arteriovenosa de la saturación de oxígeno, su extracción tisular, la fracción de
espacio muerto y el shunt intrapulmonar.
Resultados: La mortalidad hospitalaria fue del 13,4%. Se identificaron como factores de riesgo
de mortalidad, la edad, el tiempo de extracorpórea, el score inotrópico; el lactato al ingreso,
su pico máximo, la velocidad de variación y el tiempo en que estuvo elevado; la saturación
venosa, el exceso de base, el espacio muerto, la extracción de oxígeno, y el shunt intrapul-
monar. Las variables que mostraron mayor valor predictivo de mortalidad fueron el tiempo de
extracorpórea, el lactato al ingreso y su pico máximo. En el análisis multivariante se detectaron
como variables independientes de mortalidad, un pico de lactato de 6,3 mmol/l y un tiempo
hiperlactacidemia de 24 h.
Conclusiones: La elevación máxima del lactato posee una alta capacidad predictiva de morta-
lidad y nos permitirá iniciar precozmente la oxigenación con membrana extracorpórea.
© 2011 Asociación Española de Pediatría. Publicado por Elsevier España, S.L. Todos los derechos
reservados.
∗ Autor para correspondencia.

Correo electrónico: garcier@gmail.com (J.A. García-Hernández).
1695-4033/$ – see front matter © 2011 Asociación Española de Pediatría. Publicado por Elsevier España, S.L. Todos los derechos reservados.
http://dx.doi.org/10.1016/j.anpedi.2012.03.021
La mortalidad en el postoperatorio de las cardiopatías congénitas 367
KEYWORDS Prognostic markers of mortality after congenital heart defect surgery

Congenital heart
Abstract
disease;
Introduction: Our aim is to identify risk factors for mortality after surgery for congenital heart
Surgery;
disease in children, in order to establish indications for extracorporeal membrane oxygenation
Extracorporeal
(ECMO).
circulation;
Patients and methods: One hundred and eighty six children underwent cardiac surgery with
Circulatory assistance
extracorporeal circulation from April 2007 to June 2009. The following parameters were mea-
sured serially during their stay in Paediatric Intensive Care (PICU): Arterial and venous blood
pH, pCO2, base excess, oxygen saturation, arterio-venous oxygen saturation difference, oxygen
extraction ratio, ventilatory dead space and intrapulmonary shunting.
Results: Hospital mortality was 13,4%. The following risk factors for mortality were identified:
age, bypass time, inotropic score, lactate level upon arrival in PICU including its peak value
and its rate of variation, mixed venous saturation, base excess, ventilatory dead space, oxygen
extraction ratio, and intrapulmonary shunting. However, the strongest predictors of mortality
were bypass time, lactate levels upon admission on PICU, and the peak lactate level. Multiva-
riate analysis showed a lactate level of 6.3 mmol/l and a high blood lactate for 24 hours to be
independent predictors of mortality.
Conclusions: The peak lactate level is a strong predictor of mortality. As such, it would be a
useful indicator of the need for ECMO support.
© 2011 Asociación Española de Pediatría. Published by Elsevier España, S.L. All rights reserved.
Introducción El estudio se basó en la realización de gasometrías en

sangre arterial y venosa, para determinar el pH, la presión
En el postoperatorio de las cardiopatías congénitas, la apa- parcial de anhídrido carbónico (pCO2 ), la saturación de oxí-
rición de un bajo gasto es frecuente, asociándose a un geno (SatO2 ) y el exceso de bases (EB). En el aire espirado a
aumento de la mortalidad. La identificación precoz de los través del tubo endotraqueal se analizó el CO2 al final de la
pacientes que presentan este estado puede ser de utilidad espiración. Las determinaciones se hicieron en 2 momentos
para predecir una mala evolución. puntuales: al ingreso en UCIP y a las 22 horas. La primera
Si no se consigue una mejoría con las medidas de nos valora el estado del niño al final de la intervención tras
tratamiento convencionales, estaría indicado iniciar la la corrección quirúrgica y la segunda, la evolución clínica en
oxigenación con membrana extracorpórea (ECMO) para las primeras horas del postoperatorio.
aumentar la supervivencia postoperatoria1 , ya que con la Con los datos gasométricos obtenidos se calcularon los
ECMO esta última oscila entre el 35 y el 60%2,3 . siguientes índices: el shunt intrapulmonar (QS /QT ), la dife-
Distintos marcadores pronósticos postoperatorios han rencia arteriovenosa de la saturación de oxígeno (Da-v O2 ),
sido identificados, los más conocidos son la saturación la extracción tisular de oxígeno (ET O2 ), y la fracción de
venosa de oxígeno y el lactato sérico. El objetivo de este espacio muerto pulmonar (Vd/Vt). Para diferenciar en el
estudio fue determinar, de entre todas las variables ana- texto y en las tablas los parámetros obtenidos en el primer
lizadas, las que posean un mayor poder predictivo de tiempo respecto a los obtenidos en el segundo, se designaron
mortalidad. Esto nos ayudará a establecer de forma precoz con el número «1» a los primeros y con el «2» a los segundos.
las indicaciones de ECMO. Las determinaciones del lactato se hicieron en quirófano,
antes y durante la CEC; al ingreso en UCI-P y a las 22 horas
del primer día; a las 7 y las 18 horas. del segundo y el tercer
Pacientes y métodos días, y a las 7 horas del cuarto día. El nivel de lactato que
se estudió fue el venoso, pues es el que nos informa de una
El diseño del estudio fue de carácter prospectivo, obser- forma más fiable del estado metabólico celular6 .
vacional y descriptivo. Se aprobó por el comité de ética e El QS /QT se calculó mediante la siguiente fórmula:
investigación sanitaria del hospital, obteniéndose consenti- QS /QT = (1 --- Sat a O2 ) ÷ (1 --- Sat v O2 )7 , considerándose
miento informado de los padres. valores patológicos los superiores al 10%.
En el periodo comprendido entre abril del 2007 y junio La Da-v O2 y el ET O2 estiman la perfusión tisular sistémica
del 2009, se intervinieron consecutivamente en un hospital y el consumo de oxígeno. La primera se calcula hallando la
infantil de tercer nivel, e ingresaron posteriormente en la diferencia entre la saturación arterial de oxígeno (SaO2 ) y
UCI pediátrica, 186 niños con diversos tipos de cardiopatías la saturación venosa de oxígeno (SvO2 ) según la siguiente
congénitas. fórmula: Da-v O2 = SaO2 --- SvO2 , y consideramos patológica
Se agruparon las cardiopatías según el método interna- una diferencia mayor del 25%. La ET O2 fue calculada obte-
cional de estratificación de riesgo de la cirugía RACHS-14,5 . niendo la diferencia anterior y dividiéndola entre la SaO2
Se incluyó a todos los niños que necesitaron la técnica de {ETO2 = (SaO2 --- SvO2 ) ÷ SaO2 }8 . Los valores normales osci-
circulación extracorpórea (CEC). lan entre 0,24 y 0,289 , algunos autores consideran que una
368 J.A. García-Hernández et al
cifra mayor de 0,5 posee una alta capacidad predictiva de modelo predictor de mortalidad, obteniendo su odds ratio e
muerte10 . intervalos de confianza del 95%.
El Vd/Vt nos relaciona el CO2 arterial con el CO2 espi- Se realizaron curvas COR (curvas de características
rado. Cuando el flujo pulmonar está reducido, como en la operativas para el receptor) para localizar puntos de
estenosis pulmonar, en la insuficiencia cardiaca, y en el discriminación de la mortalidad con su sensibilidad y especi-
tromboembolismo pulmonar, se produce un descenso del ficidad. Se describe el área bajo la curva, que es una medida
CO2 espirado, debido a la dificultad para poder exhalarlo, del poder discriminativo del test y su intervalo de confianza
y esto hace que aumente el valor del Vd/Vt. La Vd/Vt se del 95%.
calcula mediante la modificación de Enghoff de la ecuación El nivel de significación estadística se estableció en
de Bohr11 : Vd/Vt = (PaCO2 --- EtCO2 ) ÷ PaCO2 , siendo su valor p < 0,05. La imputación de los datos y el análisis estadís-
normal inferior a 0,312 . (PaCO2 : presión parcial arterial de tico se realizó con el paquete Stadistical Package for Social
anhídrido carbónico). Sciences (SPSS Inc., Chicago, IL, EE. UU.) versión 18.0, de la
El estudio seriado del lactato nos permitió conocer su Unidad de Apoyo a la Investigación del Hospital Universitario
cifra al ingreso, el pico máximo y el tiempo en el que se Virgen del Rocío y G-stat 2.0 de la compañía GlaxoSmithK-
mantuvo elevado (≥ 2 mmol/l). Se determinó además, la line.
velocidad de ascenso o de descenso, en un periodo com-
prendido entre el ingreso y las 22 horas, según la siguiente
fórmula: lactato al ingreso --- lactato a las 22 horas ÷ horas Resultados
intervalo. Los pacientes se clasificaron en 2 grupos: los que
tuvieron elevado el lactato menos de 24 horas, y los que lo Se incluyó a 186 niños con edades comprendidas entre menos
tuvieron más de 24 horas. de un mes y 18 años con distintos tipos de intervenciones
La valoración de la necesidad de drogas inotrópicas se quirúrgicas (tabla 1). Los valores medianos de edad, peso,
hizo mediante una fórmula modificada del score o puntua- tiempos quirúrgicos, horas de ventilación mecánica y días
ción inotrópica13 (dopamina + dobutamina) × 1 + (milrinona) de estancia en UCIP quedan reflejados en la tabla 2.
× 10 + (adrenalina + noradrenalina + isoproterenol) × 100. En La moda de categoría de riesgo quirúrgico fue la R3 con 94
este estudio se multiplicó cada una de las drogas por las niños (50,5%). En 118 niños (63,4%) la categoría de riesgo fue
horas en que fue perfundida. igual o mayor que 3. En el grupo de neonatos, la categoría
Para realizar las gasometrías se utilizó un analizador de de riesgo quirúrgico igual o mayor de 3 la presentaron 31
gases marca Radiometer ABL 725 (Radiometer A/S, Copen- niños (91%) (tabla 3).
hagen, Dinamarca). La sangre venosa central se extrajo de En el postoperatorio fallecieron 25 niños, siendo la mor-
un catéter situado en la vena yugular interna y la sangre talidad hospitalaria del 13,4%. Este resultado podría estar
arterial se tomó de un catéter colocado en la arteria radial condicionado por el cambio generacional en el servicio de
o en la arteria cubital. cirugía cardiaca. La distribución de la mortalidad, según la
En aire espirado a través del tubo endotraqueal, se ana- categoría de riesgo quirúrgico, queda reflejada en la tabla 3,
lizó el valor máximo del CO2 al final de la espiración (Et en donde también queda representada la mortalidad espe-
CO2 ) mediante un módulo de capnografia. El Et CO2 fue ana- rada.
lizado de forma continua, eligiéndose los valores medidos en En el análisis bivariante de las variables cuantitativas con
aquellos momentos puntuales descritos anteriormente para respecto a la mortalidad hospitalaria, resultó que tanto la
realizar las gasometrías. edad como el peso fueron más reducidos en los fallecidos.
También en ellos, el tiempo de CEC y el valor medio del score
inotrópico resultaron ser más elevados (tabla 4).
Análisis estadístico En relación con el lactato sérico, su valor al ingreso
y el pico máximo fueron más elevados en los fallecidos
Se realizó estadística descriptiva, utilizándose frecuencias (tabla 5). La velocidad resultó ser levemente positiva en
absolutas y relativas en el caso de las variables cualitati- los supervivientes y ligeramente negativa en los fallecidos.
vas. Las variables cuantitativas se representaron mediante Esto se debe a que en los primeros la segunda determi-
la mediana y rango intercuartílico), tras haber comprobado nación ha sido ligeramente superior a la primera, estando
que no seguían una distribución normal. ambas muy cercanas a la normalidad. En los fallecidos en
Se hizo una comparación de los grupos de estudio cambio, la segunda determinación ha sido inferior a la
mediante test de la ␹2 (chi al cuadrado de Pearson) o el test primera, pero estando ambas muy por encima del rango
exacto de Fisher. La comparación de las variables cuantita- normal.
tivas según los dos grupos de estudio se efectuó mediante Con respecto a las variables gasométricas e índices ana-
la U de Mann-Whitney. Para analizar las variables relaciona- lizados, se encontraron diferencias significativas en los
das o apareadas (ET O2 en los dos instantes, para niños con tiempos 1 y 2, en la SvO2 , en el QS /QT , en la ET O2 , y en el
hiperlactacidemia > 24) se utilizó el test de Wilcoxon. Vd/Vt (tablas 5 y 6). En el EB solo se obtuvieron diferencias
Sobre aquellas variables estadísticamente significativas, significativas en el tiempo 1 (tabla 5).
se realizó una regresión logística bivariante para determinar En la comparación de Vd/Vt entre supervivientes y falle-
los factores predictivos de mortalidad. Entre las variables cidos, los valores fueron más elevados en los fallecidos,
identificadas, se elaboró un análisis multivariante de regre- tanto en el tiempo 1 como en el 2 (0,19 vs. 0,38; p < 0,01) y
sión logística por el método de «pasos hacia delante», para (0,14 vs. 0,4; p < 0,001), respectivamente. Lo mismo suce-
encontrar aquellos factores que podían considerarse mar- dió con la ET O2 en las 2 mediciones (0,3 vs. 0,44; p < 0,02)
cadores independientes de mortalidad y con ellos crear un y (0,35 vs. 0,56; p < 0,01).
El análisis bivariante que comparó el tiempo de hiperlac-

Tabla 1 Diagnósticos de cardiopatías y tipo de
tacidemia con la ET O2 reveló que ambas se correlacionaban
intervenciones
de forma significativa. En los niños que tuvieron un tiempo
Intervenciones quirúrgicas Frecuencia de hiperlactacidemia > 24 horas, los valores de ET O2 1 y
Cierre CIA ostium primum 14 ET O2 2 fueron mas elevados y aumentaron en el tiempo 2
Cierre CIA ostium secundum 13 con respecto al 1 (p < 0,001). En cambio, en los niños con
Corrección Canal AV completo 17 un tiempo de hiperlactacidemia < 24 horas, las 2 determina-
Cierre CIV 16 ciones fueron normales. Por tanto, podemos afirmar que la
DVPAP 1 elevación mantenida del lactato se acompaña también de
DVPAT 7 un aumento de la ET O2 .
Arteria coronaria de origen anómalo 1 Se realizó un análisis bivariante mediante un test de
Anillo vascular corrección 1 regresión logística sobre aquellas variables estadísticamente
Resección tumor auricular 1 significativas (tabla 7), no encontrándose variables que
Corrección síndrome Shone 1 hayan producido confusión. Se identificaron como factores
Glenn bidireccional 7 de riesgo de mortalidad la edad, el tiempo de CEC y el score
Fontan 20 inotrópico. En relación con el lactato: su valor al ingreso, el
Switch arterial 20 pico máximo, su velocidad de variación y el tiempo en que
Damus-Kaye-Stansel 2 se mantuvo elevado, y en cuanto a los parámetros gasomé-
Rastelli 2 tricos: la SvO2 1 y 2, el EB 1, el Vd/Vt 1, la ET O2 1 y el QS /QT
Norwood variante Sano 5 1.
Ross 1 A continuación, se calcularon la curva COR y el área bajo
Corrección tetralogía Fallot 14 la curva para cada una de estas variables, considerándose
Corrección Fallot con ausencia de válvula 2 como significativa un área bajo la curva ≥ 0,8. Se deter-
pulmonar minaron los puntos de corte con la máxima sensibilidad y
Corrección de truncus arterioso común 5 especificidad, así como los valores predictivos positivos y
Comisurotomía válvula aórtica 3 negativos. Las variables que mostraron mayor valor predic-
Prótesis válvula aórtica 6 tivo de mortalidad fueron: el tiempo de CEC, el lactato al
Plastia válvula mitral 1 ingreso y su pico máximo (tabla 8).
Prótesis válvula mitral 4 En el análisis multivariante, se incluyeron aquellas varia-
Comisurotomía válvula pulmonar 2 bles que resultaron ser estadísticamente significativas en
Prótesis válvula pulmonar 3 el análisis bivariante y que presentan un intervalo de con-
Corrección ventrículo derecho de doble 4 fianza más adecuado. Estas fueron el tiempo de CEC, la
salida puntuación inotrópica, el lactato venoso al ingreso, la
Atresia pulmonar corregida con tubo de 5 velocidad de lactato, el pico de lactato y el tiempo de
Contegra hiperlactacidemia.
Corrección VD de doble salida con estenosis 2 Se detectaron como variables independientes predictoras
pulmonar de mortalidad, el pico de lactato y el tiempo de hiper-
Ampliación del tracto de salida del 6 lactacidemia. La curva COR para este modelo estadístico
ventrículo derecho tuvo un área bajo la curva de 0,95, con un intervalo de
Total 186 confianza del 95% de 0,9-1. El punto de esta curva que
corresponde a una sensibilidad del 96% y una especificidad
Canal AV: canal aurículo-ventricular; CIA; comunicación inter- del 92,5% es el 0,2, que corresponde a un pico de lactato
auricular; CIV: comunicación interventricular; DVPAP: drenaje de 6,3 mmol/l (OR: 1,6; IC del 95%, 1,2-2; p = 0,002) y un
venoso pulmonar anómalo parcial; DVPAT: drenaje venoso pul-
tiempo de hiperlactacidemia > 24 horas (OR: 11; IC del 95%,
monar anómalo total; VD: ventrículo derecho.
1-114; p = 0,044).
Discusión
Tabla 2 Estadística descriptiva de la población estudiada El objetivo de un estado hemodinámico óptimo es conseguir
una adecuada perfusión de los tejidos. La medida del gasto
P50 [P25 -P75 ]
cardiaco, sin conocer las auténticas demandas tisulares de
Edad (meses) 11 [2-61] oxígeno, puede conducir a extraer conclusiones erróneas. De
Peso (kg) 7 [4,2-17,2] igual forma, la presión arterial es un parámetro engañoso,
CEC (min) 125 [85-167] pues desciende de una forma tardía en situaciones de shock
PA (min) 70 [41-102] cardiogénico14 .
Ventilación mecánica (h) 24 [3,75-96] Un gasto cardiaco normal puede ser insuficiente en situa-
Estancia en UCIP (días) 5 [3-11] ciones con un incremento en las necesidades de oxígeno
PA: pinzamiento aórtico; CEC: circulación extracorpórea; kg: y un bajo gasto puede ser suficiente en estados con bajos
kilogramo; min: minutos; UCIP: unidad de cuidados intensivos requerimientos, como sucede en la hipotermia terapéutica.
pediátricos. Por estos motivos, necesitamos identificar otros parámetros
de fácil medida, que nos informen del flujo tisular y del
Tabla 3 Distribución frecuencias por categorías según el método RACHS-1 y mortalidad obtenida
RACHS-1 Frecuencia Mortalidad esperadaa Mortalidad obtenida
R1 17 (9,1%) 0,4% 0% (0/17)
R2 51 (27,4%) 3,8% 5,9% (3/51)
R3 94 (50,5%) 9,5% 11,7% (11/94)
R4 17 (9,1%) 19,2% 35,3% (6/17)
R5 1 (0,5%) NA NA (1/1)
R6 6 (3,2%) 47% 66,7% (4/6)
Total n = 186 n = 25 (13,4%)
NA: no aplicable.
a Mortalidad esperada según publicación original de Jenkins et al.5 .
Tabla 4 Análisis bivariante de las variables cuantitativas respecto a la mortalidad. Datos descriptivos
Variables Supervivientes Fallecidos p
(n = 161) (n = 25)
P50 [P25 -P75 ] P50 [P25 -P75 ]
Edad (meses) 14 [4-71] 2 [0-15] 0,001*
Peso (kg) 8 [4,5-18,2] 4,1 [3,4-7,6] 0,001*
CEC (min) 120 [82,2-160] 200 [152-260] 0,001*
PA (min) 69 [40,5-100] 88 [44-120] 0,268
Score inotrópico 840 [240-2.250] 2.940 [1.370-7.440] 0,001*
CEC: circulación extracorpórea; PA: pinzamiento aórtico; kg: kilogramos; min: minutos.
* p estadísticamente significativa.
balance entre los aportes y el consumo de oxígeno15 . Los es el causante del descenso del CO2 espirado en relación al
más estudiados hasta ahora han sido la SvO2 y el lactato CO2 arterial y esto es lo que da lugar al aumento del Vd/Vt.
sérico. En el bajo gasto, la ET O2 aumenta, pero el consumo per-
Los resultados obtenidos en este estudio (tabla 7) nos manece constante. El punto donde se produce la máxima
demuestran que los pacientes fallecidos presentaban un extracción se denomina valor crítico y coincide con una ET
déficit en el aporte de oxígeno a los tejidos, producido por O2 de 0,5. Si esta situación persiste, futuras reducciones del
una reducción del gasto cardiaco. Este bajo gasto produciría aporte de oxígeno conducen a un descenso del consumo y,
un descenso del flujo sistémico que se manifiesta, por una a partir de ese momento, el consumo se vuelve patológica-
elevación del lactato y por un aumento de la ET O2 , que pro- mente dependiente del aporte. Es en ese momento donde
duce un descenso de la SvO2 en relación a la SaO2. El bajo se produce un descenso de la SvO2 , una elevación del lac-
gasto también produce una reducción del flujo pulmonar que tato y un descenso del pH y del EB. Si este déficit persiste
Tabla 5 Análisis bivariante de las variables cuantitativas respecto a la mortalidad. Lactato, parámetros gasométricos y del
equilibrio ácido base
Variables Supervivientes(n = 161)P50 [P25 -P75 ] Fallecidos(n = 25)P50 [P25 -P75 ] p
Lactato ingreso (mmol/l) 2,8 [1,9-4,6] 6,8 [4,02-10,12] 0,001*
Pico lactato (mmol/l) 3,5 [2,7-4,9] 10,7 [7,6-16,2] 0,001*
Velocidad lactato 0,12 [0,04-0,27] -0,03 [-0,4.-0,13] 0,001*
pH arterial 1 7,36 [7,28-7,42] 7,34 [7,28-7,46] 0,069
pH arterial 2 7,39 [7,35-7,44] 7,37 [7,2-7,42] 0,06
SvO2 1 65,6 [51,1-77,9] 46 [35,3-54,9] 0,001*
SvO2 2 60,8 [47,6-70,2] 37,6 [25,5-47,6] 0,001*
EB 1 -1,6 [-3,1-0,6] -3,8 [-6,7–1,2] 0,002*
EB 2 -1 [-2,85-1,4] -2,9 [-6,5-0,7] 0,085
Mmol/l: mili moles por litro; pH arterial 1: pH arterial al ingreso; pH arterial 2: pH arterial a las 22 horas; SvO2 1: saturación venosa de
oxígeno al ingreso; SvO2 2; saturación venosa de oxígeno a las 22 horas; EB 1: exceso de base al ingreso; EB 2: exceso de base a las 22
horas.
Tabla 6 Análisis bivariante de las variables cuantitativas respecto a la mortalidad. Índices gasométricos calculados
Variables Supervivientes Fallecidos p
(n = 161) (n = 25)
QSP /QT 1 (%) 7,54 [2,9-17,3] 29,5 [4,35-51,93] 0,004*
QSP /QT 2 (%) 6,69 [2,63-19,1] 48,46 [6,13-64,96] 0,002*
ET O2 1 0,30 [0,19-0,42] 0,44 [0,35-0,56] 0,02*
ET O2 2 0,35 [0,26-0,47] 0,56 [0,30-0,70] 0,015*
Da-v O2 1 (%) 29,2 [18,9-40,1] 34 [26,3-51,2] 0,071
Da-v O2 2 (%) 33,9 [25,1-43,4] 27,4 [18,1-62,9] 0,848
VD /VT 1 0,19 [0,08-0,33] 0,38 [0,22-0,5] 0,01*
VD /VT 2 0,14 [0,009-0,297] 0,4 [0,3-0,63] 0,001*
QSP /QT 1: efecto shunt intrapulmonar al ingreso; QSP /QT 2: efecto shunt intrapulmonar a las 22 horas; ET O2 1: extracción tisular de
oxígeno al ingreso; ET O2 2: extracción tisular de oxígeno a las 22 horas; Da-v O2 1: diferencia en la saturación de oxígeno arterial y
venosa al ingreso; Da-v O2 2: diferencia en la saturación de oxígeno arterial y venosa a las 22 horas; VD /VT 1: espacio muerto pulmonar
al ingreso; VD /VT 2: espacio muerto pulmonar a las 22 horas.
que es la causante del descenso simultáneo de las saturacio-

Tabla 7 Regresión logística con factores predictivos de
nes arterial y venosa. Por lo tanto, la medida en estos niños
mortalidad
de la SvO2 aislada no estima adecuadamente el gasto car-
Variables Odds ratio IC del 95% p diaco, siendo necesario relacionarla con la SaO2 mediante
Edad 0,98 0,973-0,999 0,041* la ETO2 18 .
Tiempo CEC 1,01 1,009-1,024 0,001* La medición seriada de la SvO2 nos proporciona una infor-
Score inotrópico 1,001 1,0-1,0 0,02* mación muy sensible del gasto cardiaco14 . En la actualidad,
Lactato ingreso 1,4 1,178-1,69 0,001* nosotros medimos la SvO2 continua en la vena cava supe-
Pico lactato 11,79 2,403-57,875 0,002* rior a través de una fibra óptica incorporado a un catéter de
Velocidad lactato 0,23 0,063-0,838 0,02* doble luz (Vigileo® , Edwards Lifesciences), habiéndose rea-
SvO2 1 0,94 0,911-0,969 0,001* lizado ya estudios durante y después de la cirugía cardiaca
SvO2 2 0,91 0,875-0,952 0,001* en niños19,20 .
EB a ingreso 0,79 0,697-0,913 0,001* Las 3 variables estudiadas que han tenido un mayor poder
VD /VT 1 32,08 1,593-646,2 0,024* discriminatorio para predecir la mortalidad han sido: el
ETO2 1 100,7 5,391-1881,9 0,002* tiempo de CEC, los valores de lactato al ingreso y el pico
QSP /QT 1 1,04 1,016-1,064 0,001* máximo (tabla 8).
Tiempo elevación 0,15 16,1-962,159 0,001* Según nuestros resultados, podemos afirmar que un niño
lactato que haya precisado un tiempo de CEC mayor de 149 min,
un lactato al ingreso ≥ 5,8 mmol/l y un pico de lactato de
CEC: circulación extracorpórea; EB: exceso de base; ET O2 1: 7 mmol/l tiene altas probabilidades de fallecer y debemos
extracción tisular de oxígeno al ingreso; QSP /QT 1: efecto shunt estar alerta para iniciar precozmente la técnica de ECMO
intrapulmonar al ingreso; SvO2 1: saturación venosa de oxígeno
antes de que la situación sea irreversible. El análisis mul-
al ingreso; SvO2 2: saturación venosa a las 22 horas; VD /VT 1:
espacio muerto al ingreso.
tivariable realizado también coincide con el bivariante, en
* p estadísticamente significativa. que el lactato, es el parámetro que mejor nos predice la
mortalidad. Además, este último análisis nos ha permitido
completar la indicación de ECMO, al establecer que una ele-
se produce una reducción en la producción de energía, y vación del lactato durante un periodo ≥ 24 horas, con un
seguidamente la muerte celular16 . pico máximo de 6,3 mmol/l, nos predice la mortalidad con
La ET O2 es un índice muy sensible para valorar el gasto una sensibilidad del 96%, y una especificidad del 92,5%.
cardiaco, pues nos informa del consumo de oxígeno17 . Es Otros autores también han encontrado que la ele-
muy útil en niños con corazón univentricular, ya que en vación del lactato se ha asociado a un incremento de
ellos se produce una mezcla de sangre arterial y venosa, la mortalidad21---23 . No obstante, el valor predictivo del
Tabla 8 Variables con mayor poder predictivo de mortalidad

Variables Punto corte ABC IC del 95% Sensi. Espe. VPP VPN
Tiempo CEC 149 0,80 0,7-0,9 79% 69% 27% 96%
Lactato ingreso 5,8 0,83 0,7-0,9 60% 89% 41% 95%
Pico lactato 7 0,95 0,9-1 88% 94% 69% 98%
ABC: área bajo la curva; CEC: circulación extracorpórea; Espe.: especificidad; IC: intervalo de confianza; Sensi.: sensibilidad; VPN: valor
predictivo negativo; VPP: valor predictivo positivo.
lactato, no ha sido comparado con el correspondiente a surgery for congenital heart disease. J Thorac Cardiovasc Surg.
otras variables postoperatorias, de ahí la originalidad de 2002;123:110---8.
nuestro estudio. Coincidimos, por tanto, con ellos, en que 6. Cheifetz IM, Kern FH, Schulman SR, Greeley WJ, Ungerleider
los cambios producidos del nivel de lactato es el mejor RM, Meliones JN. Serum lactates correlate with mortality after
marcador pronóstico de mortalidad24 . operations for complex congenital heart disease. Ann Thorac
Surg. 1997;64:735---8.
En la intervención de switch arterial para la transposi-
7. Shapiro BA, Peruzzi WT, Templin RK. Evaluación de la oxigena-
ción de grandes arterias, hemos constatado que el tiempo ción celular. En: Shapiro BA, Peruzzi WT, Templin RK, editores.
de CEC es considerado un factor de riesgo de mortalidad25 . Manejo clínico de los gases sanguíneos. 5a ed Buenos Aires:
En el presente estudio, hemos coincidido con la literatura Paramericana; 1996. p. 89---97.
revisada, en que un tiempo ≥ 149 min es un factor de riesgo 8. Astiz ME, Rackow EC. Assessing perfusion failure during circula-
de mortalidad. tory shock. Crit Care Clin. 1993;9:299---312.
En cuanto al índice Vd/Vt, las reducciones del gasto 9. Whyte RK. Mixed venous oxygen saturation in the newborn. Can
cardiaco son la causa más frecuente de su aumento. Su we and should we measure it? Scand J Clin Lab Invest. 1990;50
determinación en adultos con síndrome de dificultad res- Suppl 203:203.
piratoria aguda ha demostrado ser un factor de riesgo de 10. Barry P, Morris K, Ali T. Vascular access and clinical monito-
ring. Chapter 6. En: Paediatric Intensive Care. New York: Oxford
mortalidad26 . En niños intervenidos de cardiopatías congé-
University Press; 2010. p. 88.
nitas, su elevación se asoció a mayor tiempo de ventilación 11. West JB, Wagner PD. Ventilation, blood flow, and gas exchange.
y de estancia en la UCIP27 . Sus valores fueron superiores En: Murray JF, Nadel JA, editors. Textbook of respiratory
en niños intervenidos con la técnica switch arterial y con medicine, 1, 3rd ed. Philadelphia: W.B. Saunders; 2000.
la de Norwood28,29 . En la actualidad, son los únicos estu- p. 55---89.
dios realizados en el postoperatorio de niños intervenidos 12. López-Herce J. Monitorización de la función respiratoria en
de cardiopatías congénitas. el niño con ventilación mecánica II: complianza, resistencia,
atrapamiento aéreo, espacio muerto, trabajo respiratorio. En:
Grupo de respiratorio de la Sociedad Española de Cuidados
Limitaciones del estudio Intensivos Pediátricos, editor. Manual de ventilación mecánica
en pediatría. 2 ed. Madrid: Publimed; 2009. p. 171---92.
Se necesitarían estudios multicéntricos para corroborar 13. Wernosvsky G, Wypij D, Jonas RA, Mayer JE, Hanley FL, Hickey
PR, et al. Postoperative course and hemodynamic profile
estos hallazgos. El pico máximo de elevación del lactato en
after the arterial switch operation in neonates and infants. A
los fallecidos antes del fallecimiento no tiene valor puesto comparison of low-flow cardiopulmonary bypass and circulatory
que está determinado por la bradicardia y acidosis. No se arrest. Circulation. 1995;92:2226---35.
han correlacionado los factores de riesgo con las escalas 14. Wo CC, Shoemaker WC, Appel PL, Bishop MH, Kram HB,
habituales de valoración de gravedad (PRISM, PIM o PELOD). Hardin E. Unreliability of blood pressure and heart rate to eva-
En resumen, la prolongación del tiempo de CEC (≥ 149 luate cardiac output in emergency resuscitation and critical
min) se asocia a un aumento en la mortalidad. La determina- illness. Crit Care Med. 1993;21:218---23.
ción seriada del lactato posee un alto valor pronóstico. La 15. Tibby SM, Murdoch IA. Monitoring cardiac function in intensive
elevación progresiva del lactato tiene una alta capacidad care. Arch Dis Child. 2003;88:46---52.
predictiva de mortalidad y nos permitirá indicar precoz- 16. Barry P, Morris K, Ali T. Vascular access and clinical monitoring.
En: Paediatric intensive care. New York: Oxford University Press;
mente técnicas como la ECMO.
2010, 62---101.
17. Hoffman GM, Tweddell JS, Ghanayem NS, Mussatto KA,
Stuth EA, Jaquis RD, et al. Alteration of the critical arterio-
Conflicto de intereses venous oxygen saturation relationship by sustained after load
reduction after the Norwood procedure. J Thorac Cardiovasc
Los autores declaran que no existe ningún conflicto de inte- Surg. 2004;127:738---45.
reses. 18. Jeff Martin, Shekerdemian Lara S. The monitoring of venous
saturations of oxygen in children with congenitally malformed
hearts. Cardiol Young. 2009;19:34---9.
Bibliografía 19. Ranucci M, Isgrò G, De La Torre T, Romitti F, De Benedetti D,
Carlucci C, et al. Continuous monitoring of central venous
1. Nido PJ. Extracorporeal membrane oxygenation for cardiac sup- oxygen saturation (Pediasat) in pediatric patients undergoing
port in children. Ann Thorac Surg. 1996;61:336---9. cardiac surgery: a validation study of a new technology. J Car-
2. Kolovos NS, Bratton SL, Moler FW, Bove EL, Ohye RG, Bartlett diothorac Vasc Anesth. 2008;22:847---52.
RH, et al. Outcome of pediatric patients treated with extra- 20. Ranucci M, Isgrò G, Carlucci C, De La Torre T, Enginoli S,
corporeal life support after cardiac surgery. Ann Thorac Surg. Frigiola A, Surgical and Clinical Outcome Research Group. Cen-
2003;76:1435---42. tral venous oxygen saturation and blood lactate levels during
3. Morris MC, Ittenbach RF, Godinez RI, Portnov JD, Tabbutt S, cardiopulmonary bypass are associated with outcome after
Hanna BD, et al. Risk factors for mortality in 137 pediatric car- pediatric cardiac surgery. Crit Care. 2010;14:R149.
diac intensive care unit patients managed with extracorporeal 21. Siegel LB, Hauser GJ, Hertzog JH, Hopkins RA, Hannan RL,
membrane oxygenation. Crit Care Med. 2004;32:1061---9. Dalton HJ. Initial postoperative serum lactate predicts outcome
4. Mildh L, Pettilä V, Sairanen H, Rautiainen P. Predictive value in children after open heart surgery [abstract]. Crit Care Med.
of paediatric risk of mortality score and risk adjustment for 1995;23:A205.
congenital heart surgery score after paediatric open-heart sur- 22. Siegel LB, Dalton HJ, Hertzog JH, Hopkins RA, Hannan RL,
gery. Interact Cardiovasc Thorac Surg. 2007;6:628---31. Hauser GJ. Initial postoperative serum lactate levels predict
5. Jenkins KJ, Gauvreau K, Newburger JW, Spray TL, Moller JH, survival in children after open heart surgery. Intensive Care
Iezzoni LI. Consensus-based method for risk adjustment for Med. 1996;22:1418---23.
23. Munoz R, Laussen PC, Palacio G, Zienko L, Piercey G, for death in the acute respiratory distress syndrome. N Engl J
Wessel DL. Changes in whole blood lactate levels during cardio- Med. 2002;346:1281---6.
pulmonary bypass for surgery for congenital cardiac disease: an 27. Ong T, Stuart-Killion RB, Daniel BM, Presnell LB, Zhuo H,
early indicator of morbidity and mortality. J Thorac Cardiovasc Matthay MA, et al. Higher pulmonary dead space may predict
Surg. 2000;119:155---62. prolonged mechanical ventilation after cardiac surgery. Pediatr
24. Charpie JR, Dekeon MK, Goldberg CS, Mosca RS, Bove EL, Pulmonol. 2009;44:457---63.
Kulik TJ. Serial blood lactate measurements predict early out- 28. García Hernández JA, Montero Valladares C, Martínez López
come after neonatal repair or palliation for complex congenital AI, Romero Parreño A, Grueso Montero J, Gil-Fournier Carazo
heart disease. J Thorac Cardiovasc Surg. 2000;120:73---80. M, et al. Risk factors associated with arterial switch opera-
25. García-Hernández JA, Montero-Valladares C, Martínez-López AI, tion for transposition of the great arteries. Rev Esp Cardiol.
Gil-Fournier M, Praena-Fernández JM, Cano-Franco J, et al. 2005;58:815---21.
Prognostic evaluation of arterial switch in the trans- 29. García-Hernández JA, González-Rodríguez JD, Martínez-López
position of great arteries. An Pediatr (Barc). 2011;74: AI, Canalejo-González D, Romero-Parreño A, Santos de Soto
174---81. J, et al. Resultados de la intervención de Norwood para el
26. Nuckton TJ, Alonso JA, Kallet RH, Daniel BM, Pittet JF, síndrome del corazón izquierdo hipoplásico. Rev Esp Cardiol.
Eisner MD, et al. Pulmonary dead-space fraction as a risk factor 2007;60:732---8.
British Journal of Anaesthesia 107 (3): 344–50 (2011)
Advance Access publication 16 June 2011 . doi:10.1093/bja/aer166
Mixed venous oxygen saturation predicts short- and

long-term outcome after coronary artery bypass grafting
surgery: a retrospective cohort analysis
J. Holm, E. Håkanson, F. Vánky and R. Svedjeholm *
Department of Cardiothoracic Surgery and Cardiothoracic Anaesthesia, Linköping University Hospital, Linköping University,
SE-581 85 Linköping, Sweden
* Corresponding author. E-mail: rolf.svedjeholm@lio.se
Downloaded from http://bja.oxfordjournals.org/ at Centro Nacional de Informacion de Ciencias Medicas on January 14, 2013
Background. Complications of an inadequate haemodynamic state are a leading cause of
Editor’s key points morbidity and mortality after cardiac surgery. Unfortunately, commonly used methods to
† This large retrospective assess haemodynamic status are not well documented with respect to outcome. The
study analysed the aim of this study was to investigate SvO2 as a prognostic marker for short- and long-term
association between SvO2 outcome in a large unselected coronary artery bypass grafting (CABG) cohort and in
and outcome after subgroups with or without treatment for intraoperative heart failure.
coronary artery bypass Methods. Two thousand seven hundred and fifty-five consecutive CABG patients and
grafting (CABG) surgery. subgroups comprising 344 patients with and 2411 patients without intraoperative heart
† SvO2 ,60% on intensive failure, respectively, were investigated. SvO2 was routinely measured on admission to the
care unit (ICU) admission intensive care unit (ICU). The mean (SD) follow-up was 10.2 (1.5) yr.
was associated with Results. The best cut-off for 30 day mortality related to heart failure based on receiver-
worse short- and operating characteristic analysis was SvO2 60.1%. Patients with SvO2 ,60% had higher
long-term outcome after 30 day mortality (5.4% vs 1.0%; P,0.0001) and lower 5 yr survival (81.4% vs 90.5%;
CABG. P,0.0001). The incidences of perioperative myocardial infarction, renal failure, and stroke
† This cut-off value of SvO2 were also significantly higher, leading to a longer ICU stay. Similar prognostic
,60% is higher than information was obtained in the subgroups that were admitted to ICU with or without
previous reports. treatment for intraoperative heart failure. In patients admitted to ICU without treatment
† SvO2 ,60% on ICU for intraoperative heart failure and SvO2 ≥60%, 30 day mortality was 0.5% and 5 yr
admission merits survival 92.1%.
increased attention and Conclusions. SvO2 ,60% on admission to ICU was related to worse short- and long-term
monitoring after CABG outcome after CABG, regardless of whether the patients were admitted to ICU with or
and could influence without treatment for intraoperative heart failure.
postoperative
Keywords: assessment, patient outcomes; coronary artery bypass grafting; patient
decision-making.
monitoring; postoperative complications; survival rates
Accepted for publication: 4 May 2011
Complications from postoperative heart failure are the organ function. In a small cohort of patients undergoing cor-
leading cause of adverse outcomes after cardiac surgery1 2 onary artery bypass grafting (CABG), it has been shown that
and adequate monitoring of haemodynamic state in this SvO2 correlates with short-term outcome.8 These patients
setting is therefore essential. Unfortunately, commonly were managed according to a metabolic strategy whereby
used methods to assess haemodynamic state are not well inotropic drugs were largely replaced by metabolic
documented with regard to their association with support.8 9 The cohort was highly selected, and therefore,
outcome.3 4 Uncertainty regarding how to use haemo- these results may not be applicable to more general patients
dynamic data obtained for decision-making could partially undergoing CABG. Furthermore, the additional value of SvO2
explain reports of worse outcome or lack of benefit associ- measurements remains obscure given that many patients
ated with the use of pulmonary artery catheters in critically admitted to intensive care unit (ICU) already have treatment
ill patients and in cardiac surgery.5 6 for known intraoperative heart failure.
Heart failure, in physiological terms, reflects a cardiac Therefore, we wanted to test the hypothesis that SvO2 has
output insufficient to meet systemic requirements.7 Evidence predictive value for short-term outcome in an unselected
of such mismatch between supply and demand are low cohort of CABG patients. We also wanted to study the predic-
mixed venous oxygen saturation (SvO2 ) and inadequate tive value of SvO2 in patients admitted to ICU with or without
& The Author [2011]. Published by Oxford University Press on behalf of the British Journal of Anaesthesia. All rights reserved.
For Permissions, please email: journals.permissions@oup.com
SvO2 ,60% predicts worse outcome after CABG BJA
intraoperative treatment for heart failure. As postoperative and the abdominal wall. A 4-0 prolene purse string suture
complications remain an important determinant of long- was gently tightened around the puncture site at the right
term outcome, we also wanted to study SvO2 with regard ventricular outflow tract to minimize risk for bleeding at with-
to long-term survival. drawal, which usually was done the next morning before
withdrawal of the chest tubes.
Methods From this catheter, blood samples were drawn for the
measurement of SvO2 after weaning from CPB, on admission
Patients
to ICU, the first postoperative morning and whenever other
The University Hospital in Linköping is the only referral centre clinical variables raised questions about haemodynamic ade-
in the southeast region of Sweden, serving a population of quacy. Only SvO2 measurements obtained on admission to
1 million. All patients at this department undergoing iso- ICU were routinely recorded in the computerized institutional
lated CABG between 1995 and 2000 were included in the database.
study. During this period, a total of 2774 patients were oper- On the basis of previous experience, it was our practice to
ated on with isolated CABG. In this cohort, data on SvO2 were repeat sampling for SvO2 in the ICU not only in patients with
lacking in 19 patients leaving 2755 patients to be followed in low SvO2 but also in patients with a negative trend and when-
the study. Furthermore, these patients were divided into two ever other clinical variables raised questions about haemo-
subgroups of patients who did (n¼344) or did not (n¼2411) dynamic adequacy. If benign causes such as hypovolaemia
receive intraoperative treatment for heart failure before and shivering could be excluded, more serious conditions
admission to ICU. such as tamponade or myocardial pump failure were con-
The study was performed according to the principles of sidered. In this respect, SvO2 was used to identify patients
the Helsinki Declaration of Human Rights and approved by who might benefit from echocardiography in the ICU and
the ethics committee for medical research at the Faculty of more meticulous haemodynamic monitoring. Pulmonary
Health Sciences University Hospital of Linköping. Owing to artery thermodilution catheters were used in 7.9% of the
the nature of the study, the ethics committee waived the patients and mainly selected for patients with pronounced
need for written informed consent (2003-12-16; Dnr 03-596). circulatory problems in need of pharmacological support.
Patient characteristic and periprocedural data were regis-
tered prospectively in a computerized database and analysed Definitions
retrospectively. All fields were defined in a data dictionary.
Definitions for variables presented in Table 1 are given in
The cause of death within 30 days was analysed specifi-
Supplementary material.
cally from each patient chart and in most cases supported
by autopsy. The cause of death was categorized as death
related to heart failure or death due to other causes. Data Statistics
on late mortality were retrieved from the Swedish Civil Regis- A receiver-operating characteristic (ROC) analysis was carried
try. Eight patients emigrated during follow-up, four of them out to calculate the area under the curve (AUC) and to evalu-
within 5 yr of surgery. The mean (SD) follow-up was 10.2 ate prognostic performance of SvO2 with regard to all-cause
(1.5) yr (range 0.9– 12.7 yr). mortality, and mortality related to cardiac failure. The x 2
test or Fisher’s exact test was used when appropriate for
Clinical management comparison of dichotomous variables. Student’s t-test or
After an overnight fast, and administration of their b-blocker the Mann –Whitney U-test was used when appropriate for
medication, patients were premedicated with oxicone 4–10 comparison of continuous variables. The Kaplan –Meier esti-
mg and scopolamine 0.2– 0.5 mg i.m. Anaesthesia was mator was used for the assessment of long-term survival.
induced with thiopental 1– 2 mg kg21 body weight and fen- As the analyses were exploratory, no formal adjustment
tanyl 5 –10 mg kg21 body weight. Pancuronium bromide or was made for multiple testing, but a more conservative
rocuronium bromide was used for neuromuscular block. P-value of ,0.01 was required to be judged statistically sig-
Anaesthesia was maintained with intermittent doses of fen- nificant. Results are given as percentages or mean (SD). Stat-
tanyl and isoflurane. istical analyses were performed using Statistica 8.0, StatSoft
All patients underwent surgery using standard techniques Inc., Tulsa, OK, USA, and SPSS 17.0 (SPSS Inc.).
with cardiopulmonary bypass (CPB) and aortic cross-
clamping using cold crystalloid cardioplegia.10 SvO2 was mon- Results
itored on the venous line of the CPB circuit during and on
weaning from CPB. Before weaning from CPB, an epidural Patient characteristics
catheter cut 5 cm from its tip (Perifix-Katheter, B. Braun Mel- The mean (SD) age in the whole unselected cohort was 65 (9)
sungen AG, Germany) was introduced by the surgeon yr; 22% were female. The average Higgins score11 was 2.3
through the outflow tract of the right ventricle 15 cm into (2.5). Overall 30 day mortality was 1.9% and 5 yr survival
the pulmonary artery for monitoring of pulmonary artery 88.9%. The mean (SD) SvO2 on arrival to ICU was 66.3
pressure and intermittent blood sampling. An epidural (7.0)%. The distribution of SvO2 values is given in 5% intervals
needle was used for puncture of the right ventricular wall in Supplementary Figure S1, with the majority in the range
345
BJA Holm et al.
between 60% and 75%. Figure 1 displays 30 day mortality

Table 1 Perioperative characteristics in all patients with SvO2 related to these SvO2 intervals.
≥60% and SvO2 ,60%. Data presented as mean (SD) or %. BMI,
There were no complications recorded that were related to
body mass index; CVI, cerebrovascular injury; COPD, chronic
obstructive pulmonary disease; LVEF, left ventricular ejection
catheterization of the pulmonary artery.
fraction; CPB, cardiopulmonary bypass; GIK, glucose – insulin – The best cut-off for 30 day mortality related to heart
potassium; ICU, intensive care unit; IABP, intra-aortic balloon failure based on ROC analysis was SvO2 60.1% with an AUC
pump; LVAD, left ventricular assist device; MOF, multiorgan failure. of 0.74, a sensitivity of 59.3%, and a specificity of 82.4%
*P,0.01 was taken as statistically significant to reflect the (Fig. 2). The negative predictive value was 99.5%.
multiple endpoints and exploratory nature of the analyses On the basis of the ROC analysis, the patients were divided
Characteristics SvO2 SvO2 P-value
into two groups with higher (≥60%) and lower (,60%) SvO2 .
≥60% <60% Two thousand three hundred and nine patients (84%) had
(n52309) (n5446) SvO2 ≥60% and 446 patients (16%) had SvO2 ,60%.
Age (yr) 65 (9) 68 (9) ,0.0001* Patients with SvO2 ,60% had significantly higher Higgins’
Female gender 20.0% 31.4% ,0.0001* score, higher age, higher proportion of female gender, dia-
BMI (kg m22) 26.7 (3.7) 26.9 (3.6) 0.35 betes, hypertension, recent myocardial infarction, and poor
Diabetes mellitus 18.5% 24.7% 0.003* left ventricular function (Table 1).
Hypertension 39.3% 46.4% 0.005*
Preoperative CVI 6.2% 8.3% 0.10 Short-term outcome
COPD 5.9% 4.1% 0.13
Thirty day mortality was 1.0% (n¼24) in patients with SvO2
NYHA class III/IV 76.4% 80.6% 0.06
≥60% and 5.4% (n¼24) in patients with SvO2 ,60%
Myocardial infarction ,1 week 4.8% 9.4% 0.002*
(P,0.0001), yielding a relative risk of 5.2 (95% confidence
LVEF ≤0.30 4.4% 10.4% 0.0002*
interval 3.0–9.0%) for the patients with SvO2 ,60% com-
Higgins’ score 2.1 (2.4) 3.2 (3.0) ,0.0001
pared with those with SvO2 ≥60% (P,0.0001).
Urgent surgery 45.8% 52.5% 0.01
Thirty day mortality related to heart failure was 0.6%
Emergency surgery 4.7% 10.6% ,0.0001*
(n¼13) in patients with SvO2 ≥60% and 3.1% (n¼14) in
Redo procedure 2.7% 6.7% ,0.0001*
patients with SvO2 ,60% (P,0.0001), yielding a relative
Unstable angina 46.7% 57.7% ,0.0001*
risk of 5.6 (95% confidence interval 2.6– 11.8%) for the
Aortic cross-clamp time (min) 43 (18) 48 (20) ,0.0001*
patients with SvO2 ,60% compared with those with SvO2
CPB time (min) 80 (28) 94 (38) ,0.0001*
≥60% (P,0.0001).
Number of peripheral 3.6 (1.2) 3.8 (1.1) 0.0008*
Postoperative morbidity was also significantly higher in
anastomoses
patients with SvO2 ,60% including higher incidence of peri-
Left internal mammary artery 94.5% 93.9% 0.67
graft
operative myocardial infarction, renal failure, stroke, and re-
Inotropic drugs started 4.1% 7.2% 0.005*
operation for bleeding. ICU stay and ventilator treatment
intraoperatively was prolonged in patients with SvO2 ,60%. Patients with
GIK started intraoperatively 9.0% 15.9% ,0.0001* SvO2 ,60% received inotropic drugs more frequently both
Inotropic drugs, GIK, or both 11.3% 18.4% ,0.0001* intraoperatively and in the ICU.
started intraoperatively
Inotropic drugs started in ICU 5.6% 18.6% ,0.0001* Long-term follow-up
Mechanical assist (IABP/LVAD) 1.1% 4.9% ,0.0001*
Five yr survival was 90.5% (SvO2 ≥60%) and 81.4% (SvO2
SvO2 ICU (%) 68.5 (5.0) 55.0 (4.3) ,0.0001*
,60%), respectively (P,0.0001).
Reoperation for bleeding 2.5% 5.6% 0.0005*
,24 h Survival up to 12 yr related to different levels of SvO2
Perioperative stroke 1.4% 3.4% 0.004* according to Kaplan–Meier is presented in Figure 3.
Perioperative myocardial 5.0% 13.0% ,0.0001*
infarction
Subgroup analyses
Renal failure 1.9% 5.3% ,0.0001*
Postoperative dialysis 0.3% 0.9% 0.04 Patients without intraoperative treatment for heart
MOF 0.5% 1.1% 0.14 failure
Time in ICU (days) 1.5 (2.0) 2.3 (3.2) ,0.0001* From the total cohort of 2755 patients, 2411 were
Time on ventilator (h) 13 (40) 27 (67) ,0.0001* admitted to ICU without treatment for heart failure intra-
Mortality 30 days (total) 1.0% 5.4% ,0.0001* operatively. In this subgroup, the mean age was 65 (9.4)
Mortality 30 days (heart 0.6% 3.1% ,0.0001* yr and 21% were female. The mean (SD) SvO2 on arrival to
failure)
ICU was 66.5 (6.8)%. The average Higgins score was
5 yr survival 90.5% 81.4% ,0.0001*
1.9 (2.1). Overall 30 day mortality was 0.9% and 5 yr
survival 91.0%.
346
10
30 day mortality (%)

6
2
0
<50 50–54.9 55–59.9 60–64.9 65–69.9 70–74.9 >74.9
SvO2 group
Fig 1 Incidence of postoperative 30 day mortality related to different levels of SvO2 on admission to ICU.
Short-term outcome
1.0 Thirty day mortality was 0.5% (n¼10) in patients with SvO2
≥60% and 3.2% (n¼12) in patients with SvO2 ,60%
(P,0.0001), yielding a relative risk of 6.8 (95% confidence
0.8 interval 3.0–15.6%) for the patients with SvO2 ,60% com-
pared with those with SvO2 ≥60% (P,0.0001).
Thirty day mortality related to heart failure was 0.1%
0.6
(n¼3) in patients with SvO2 ≥60% and 1.4% (n¼5) in patients
Sensitivity
with SvO2 ,60% (P¼0.0002), yielding a relative risk of 9.5

(95% confidence interval 2.3 –39.4) for the patients with
0.4
SvO2 ,60% compared with those with SvO2 ≥60% (P¼0.002).
Thirty day mortality related to SvO2 interval is given in Sup-
0.2 AUC=0.74 (95% CI 0.64–0.84), P<0.0001 plementary Figure S2.
Postoperative morbidity was also significantly higher in
patients with SvO2 ,60% including higher incidence of peri-
0.0 operative myocardial infarction, renal failure, and reoperation
0.0 0.2 0.4 0.6 0.8 1.0 for bleeding. ICU stay and ventilator treatment were pro-
1-specificity longed in patients with SvO2 ,60% (Supplementary Table S1).
Fig 2 ROC to evaluate the prognostic performance of SvO2 on

Long-term follow-up
admission to ICU with regard to mortality related to heart Five yr survival was 92.1% (SvO2 ≥60%) and 84.5% (SvO2
failure within 30 days of surgery. AUC, area under the curve; CI, ,60%), respectively (P,0.0001).
confidence interval. The best cut-off for mortality related to
heart failure was SvO2 60.1%, with a sensitivity of 59.3% and a
specificity of 82.4%. The negative predictive value was 99.5%. Patients with intraoperative treatment for heart
failure
From the total cohort of 2755 patients, 344 patients were
Two thousand and forty-nine patients (85%) had SvO2 admitted to ICU with treatment for intraoperative heart
≥60% and 362 patients (15%) had SvO2 ,60%. Patients failure. These patients were treated with either inotropic
with SvO2 ,60% had significantly higher Higgins’ score, support, glucose –insulin –potassium, or both. In this sub-
higher age, higher proportion of females, diabetes, hyperten- group, the mean age was 67 (9) yr and 28% were female.
sion, recent myocardial infarction, and poor left ventricular SvO2 on arrival to ICU averaged 64.9 (8.3)%. The average
function (Supplementary Table S1). Higgins score was 4.7 (3.5). Overall 30 day mortality was
347
BJA Holm et al.
Cumulative survival related to SvO2 level

(Kaplan-Meier)
Complete + Censored
1.0
0.9
Cumulative proportion surviving
0.8
SvO2 >74.9%
0.7 SvO2 70-74.9%
SvO2 65-69.9%
SvO2 60-64.9
0.6 SvO2 55-59.9
SvO2 50-54.9
SvO2 <50%
0.5
0 2 4 6 8 10 12
Years
Fig 3 Cumulative long-term survival according to Kaplan –Meier related to SvO2 level on admission to ICU.
7.6% (n¼26), yielding a relative risk of 8.3 (95% confidence Discussion

interval 4.7–14.4%) for these patients compared with those
The major finding of this study was that SvO2 ,60% on
admitted to ICU without intraoperative treatment for heart
admission to ICU in a large unselected cohort of CABG
failure (P,0.0001). Five year survival was 74.1%.
patients was associated with worse short- and long-term
Two hundred and sixty patients (76%) had SvO2 ≥60% and
outcome.
84 patients (24%) had SvO2 ,60%. Preoperative character-
In the early postoperative period after cardiac surgery,
istics are given in Supplementary Table S2.
the heart is in a vulnerable state recovering from ischae-
Short-term outcome mia. Inotropic agents should be used judiciously particu-
larly after CABG since these drugs can aggravate the
Thirty day mortality was 5.4% (n¼14) in patients with SvO2 consequences of ischaemia and it has been demonstrated
≥60% and 14.3% (n¼12) in patients with SvO2 ,60% that ischaemia and evolving myocardial infarction account
(P¼0.007), yielding a relative risk of 2.7 (95% confidence for a large proportion of patients with postoperative heart
interval 1.3–5.5%) for the patients with SvO2 ,60% com- failure.4 10 12 13 In this situation, a goal-orientated strategy
pared with those with SvO2 ≥60% (P¼0.009). of SvO2 levels might lead to an overuse of inotropic support
Thirty day mortality related to heart failure was 3.8% and increased cardiac workload, which could be harmful to
(n¼10) in patients with SvO2 ≥60% and 10.7% (n¼9) in the heart. On the other hand, a more conservative
patients with SvO2 ,60% (P¼0.02), yielding a relative risk of approach with acceptance of lower SvO2 might lead to
2.8 (95% confidence interval 1.2–6.6%) for the patients with inadequate systemic perfusion jeopardizing perfusion of
SvO2 ,60% compared with those with SvO2 ≥60% (P¼0.02). vital organs. On the basis of this reasoning, it would be
Thirty day mortality related to SvO2 interval is presented in desirable with appropriate guidelines to assess the ade-
Supplementary Figure S2. quacy of circulation in individual patients. In the absence
Postoperative morbidity was also significantly higher in of studies providing generally accepted criteria for insti-
patients with SvO2 ,60% with a higher incidence of peri- tution of inotropic treatment, observational data such as
operative myocardial infarction and stroke. ICU stay and ven- ours may provide some guidance. The high negative predic-
tilator treatment was prolonged in patients with SvO2 ,60%. tive value found in our study suggests that it may be
reasonably safe to withhold inotropic treatment if SvO2
Long-term follow-up
exceeds 60% if other clinical, haemodynamic, and labora-
Five yr survival was 77.7% (SvO2 ≥60%) and 63.1% (SvO2 tory data do not suggest otherwise.
,60%), respectively (P¼0.008).
348
Although used for the assessment of haemodynamic preoperative risk factors for heart failure, intraoperative
state and even for goal-directed therapy, there are few events leading to heart failure, and recovery occurring
data in the literature regarding adequate SvO2 values before admission to ICU.
during the first postoperative hours after cardiac surgery.14 The limitations of this study are its retrospective nature
In a highly selected small cohort of patients treated accord- and that only SvO2 data obtained on admission to ICU were
ing to a metabolic strategy, there was markedly increased available in our database. The latter prevented interpret-
morbidity and mortality if patients SvO2 on admission to ations regarding therapeutic interventions to increase SvO2 .
ICU after CABG was below 55%.8 In a relatively small The data come from a cohort undergoing CABG more than
cohort of patients undergoing surgery for aortic stenosis, 10 yr ago, although this did provide the opportunity to
ROC analysis suggested a cut-off SvO2 of 53.7%.15 The study long-term outcome. Furthermore, the fundamentals
present study is considerably larger and includes all patients of circulatory physiology have remained unchanged during
undergoing CABG in southeast Sweden during a 5 yr period. this time, although shifts in patient profile and clinical man-
Basically, our results confirm the previous observations but agement have occurred.
suggest that the level of SvO2 that should lead to increased In conclusion, we found that in a large unselected cohort
attention after CABG is 60% rather than 55%. of CABG patients, the measurement of SvO2 on admission to
Outcome after cardiac surgery is to a large extent ICU estimates a haemodynamic state that predicts short-
determined by the preoperative status of the patient. and long-term outcome after CABG. This association
However, the outcome is also influenced by events during remained valid regardless of whether patients were admitted
surgery and anaesthesia and the patient’s prognosis on to ICU with or without treatment for intraoperative heart
arrival at ICU may differ markedly from the preoperative failure.
evaluation. Early re-evaluation on admission to ICU is
desirable for a proactive management plan, identification
Supplementary material
of patients who may benefit from further diagnostic and
therapeutic measures, and estimation of the need for Supplementary material is available at British Journal of
ICU resources and for better prediction of the prognosis Anaesthesia online.
for individual patients.16 Our results suggest that SvO2
data can be used not only to identify patients in need of Conflict of interest
more meticulous surveillance but also those who can be
None declared.
passed on according to fast-track protocols. With the
simple and inexpensive method to measure SvO2 used in
our practice, SvO2 data can be safely obtained in virtually Funding
all patients, despite a restrictive use of pulmonary artery This study was supported by grants from Östergötlands Läns
thermodilution catheters. It remains to be documented if Landsting and Linköping University Hospital.
information of similar value can be obtained by sampling
from central venous catheters.17 18
It can be argued that in most patients, a poor haemo- References
dynamic state is evident on admission to ICU without SvO2 1 O’Connor GT, Birkmeyer JD, Dacey LJ, et al. Results of a regional
measurements. In this study, we found that SvO2 provided study of modes of death associated with coronary artery
information of prognostic value even in patients admitted bypass grafting. Northern New England Cardiovascular Disease
to ICU without known heart failure. Thus, patients with Study Group. Ann Thorac Surg 1998; 66: 1323– 8
2 Vanky FB, Hakanson E, Svedjeholm R. Long-term consequences of
SvO2 below 60% had significantly higher postoperative mor-
postoperative heart failure after surgery for aortic stenosis
tality and morbidity. On the other hand, if patients were
compared with coronary surgery. Ann Thorac Surg 2007; 83:
admitted to ICU without intraoperative treatment for heart 2036– 43
failure and SvO2 ≥60%, the risk of death within 30 days 3 Gillies M, Bellomo R, Doolan L, Buxton B. Bench-to-bedside review:
related to heart failure was 0.1%. Furthermore, these Inotropic drug therapy after adult cardiac surgery—a systematic
patients comprising approximately three-quarters of the literature review. Crit Care 2005; 9: 266–79
CABG cohort had a 92.1% 5 yr survival. 4 Fellahi JL, Parienti JJ, Hanouz JL, Plaud B, Riou B, Ouattara A. Peri-
In patients operated for aortic stenosis, SvO2 below 55% operative use of dobutamine in cardiac surgery and adverse
was associated not only with increased postoperative mor- cardiac outcome: propensity-adjusted analyses. Anesthesiology
2008; 108: 979– 87
tality but also markedly impaired long-term survival.15 The
5 Murdoch SD, Cohen AT, Bellamy MC. Pulmonary artery catheteri-
present study confirms the long-term prognostic value of
zation and mortality in critically ill patients. Br J Anaesth 2000;
SvO2 on admission to ICU. In addition to a significantly
85: 611– 5
reduced long-term survival in the subgroup with SvO2 below
6 Ranucci M. Which cardiac surgical patients can benefit from
60%, we found a relationship between different levels of placement of a pulmonary artery catheter? Crit Care 2006;
SvO2 on admission to ICU and long-term survival (Fig. 3). 10: S6
This implies that SvO2 on admission to ICU reflects the 7 Hakanson E, Svedjeholm R, Vanhanen I. Physiologic aspects in
haemodynamic state adequately taking into account postoperative cardiac patients. Ann Thorac Surg 1995; 59: S12– 4
349
BJA Holm et al.
8 Svedjeholm R, Hakanson E, Szabo Z. Routine SvO2 measurement discontinue cardiopulmonary bypass. J Thorac Cardiovasc Surg
after CABG surgery with a surgically introduced pulmonary 1981; 82: 18–25
artery catheter. Eur J Cardiothorac Surg 1999; 16: 450–7 14 Polonen P, Ruokonen E, Hippelainen M, Poyhonen M, Takala J. A
9 Svedjeholm R, Vidlund M, Vanhanen I, Hakanson E. A prospective, randomized study of goal-oriented hemodynamic
metabolic protective strategy could improve long-term survival therapy in cardiac surgical patients. Anesth Analg 2000; 90:
in patients with LV-dysfunction undergoing CABG. Scand 2010; 1052– 9
44: 45–58 15 Holm J, Hakanson RE, Vanky F, Svedjeholm R. Mixed venous
10 Vanky F, Hakanson E, Maros T, Svedjeholm R. Different character- oxygen saturation is a prognostic marker after surgery for
istics of postoperative heart failure after surgery for aortic steno- aortic stenosis. Acta Anaesthesiol Scand 2010; 54: 589–95
sis and coronary disease. Scand Cardiovasc J 2004; 38: 152–8 16 Higgins TL, Estafanous FG, Loop FD, et al. ICU admission score for
11 Higgins TL, Estafanous FG, Loop FD, Beck GJ, Blum JM, predicting morbidity and mortality risk after coronary artery
Paranandi L. Stratification of morbidity and mortality outcome bypass grafting. Ann Thorac Surg 1997; 64: 1050–8
by preoperative risk factors in coronary artery bypass 17 Lorentzen AG, Lindskov C, Sloth E, Jakobsen CJ. Central venous
patients. A clinical severity score. J Am Med Assoc 1992; 267: oxygen saturation cannot replace mixed venous saturation in
2344– 8 patients undergoing cardiac surgery. J Cardiothorac Vasc Anesth
12 Maroko PR, Kjekshus JK, Sobel BE, et al. Factors influencing infarct 2008; 22: 853– 7
size following experimental coronary artery occlusions. Circula- 18 Lequeux PY, Bouckaert Y, Sekkat H, et al. Continuous
tion 1971; 43: 67– 82 mixed venous and central venous oxygen saturation in
13 Lazar HL, Buckberg GD, Foglia RP, Manganaro AJ, Maloney JV Jr. cardiac surgery with cardiopulmonary bypass. Eur J Anaesthesiol
Detrimental effects of premature use of inotropic drugs to 2010; 27: 295– 9
350
Interactive CardioVascular and Thoracic Surgery 15 (2012) 28–32 ORIGINAL ARTICLE - ADULT CARDIAC
doi:10.1093/icvts/ivr130 Advance Access publication 11 April 2012
Age and sex differences in perioperative myocardial infarction after

cardiac surgery
Casimiro Javierrea,*, Antoni Ricartb, Rafael Manezb, Elisabet Farrerob, Maria L. Carriob,
David Rodriguez-Castrob, Herminia Torradob and Josep L. Venturab
a
Department of Physiological Sciences II, School of Medicine, University of Barcelona, IDIBELL, L’Hospitalet, Barcelona, Spain
b
Post Cardiac Surgery Intensive Care Unit, Bellvitge University Hospital, IDIBELL, L’Hospitalet, Barcelona, Spain
* Corresponding author. Unitat de Fisiologia, Departament de Ciències Fisiològiques II, Facultat de Medicina, Campus de Bellvitge, Universitat de Barcelona
08907 L’Hospitalet de Llobregat, Barcelona, Spain. Tel: +34-93-3357011; e-mail: cjavierre@ub.edu (C. Javierre)
Received 18 August 2011; received in revised form 23 October 2011; accepted 3 November 2011
Abstract
We investigate age and sex differences in acute myocardial infarction (AMI) after cardiac surgery in a prospective study of 2038
consecutive patients undergoing cardiac surgery with cardiopulmonary bypass. An age of ≥70 years implied changes in the type of AMI
from the ST-segment elevation myocardial infarction (STEMI) to non-ST-segment elevation myocardial infarction (non-STEMI). Men
were more likely than women to suffer from AMI after cardiac surgery (11.8% vs. 5.6%), as a result of the higher frequency of STEMI (6%
of men vs. 1.8% of women; P < 0.001) in both age groups. A troponin-I (Tn-I) peak was significantly higher in patients ≥70 years old. In-
hospital mortality was higher in patients ≥70 (7.3%) than in those <70 years old (3.3%), because of the increased mortality observed in
men with non-AMI (2.1% vs. 6.3%) and women with STEMI (0% vs. 28.6%) and non-STEMI (0% vs. 36.8%, P < 0.05). Old age was asso-
ciated with a higher frequency of non-STEMI, Tn-I peak, mortality and length of stay in the intensive care unit (ICU). Regardless of age,
men more often suffer from AMI (particularly STEMI). AMI in women had a notable impact on excess mortality and ICU stay observed
in patients ≥70 years of age. Clinical and Tn-I peak differences are expected in relation to age and gender after AMI post-cardiac
surgery.
Keywords: Age differences • Sex differences • Cardiac surgery • Acute myocardial infarction • STEMI/non-STEMI • Outcome
INTRODUCTION infarction (AMI) after cardiac surgery. The increasingly necessary

consensus to define perioperative myocardial infarction is still far
In acute coronary syndromes, although women suffer myocardial away, but establishing the incidence and characteristics of AMI
infarctions at older ages than men, mortality and morbidity rates after cardiac surgery is mandatory in order to improve the accur-
are persistently higher in women [1]. acy of its definition, if necessary adjusting it for gender and age.
A recent histological study of coronary arteries from men and
women who died of acute coronary disease found similarities
between the sexes, although active, inflammatory atherosclerosis MATERIALS AND METHODS
developed earlier in men than in women. This difference grad-
ually faded with age, and disappeared altogether around age 70 This prospective observational study included 2434 consecutive
[2]. In women, some studies have found better outcomes in cor- patients undergoing elective cardiac surgery with cardiopulmon-
onary artery bypass grafting (CABG) after adjusting for different ary bypass (CPB; specifically, either valve surgery, CABG or both)
risk factors [3]. Perioperative myocardial infarction is often silent, between February 2004 and April 2009 in a tertiary level univer-
with transient ECG changes, and may occur as a result of an sity hospital. General characteristics are shown in Table 1.
acute coronary thrombosis involving an unstable or vulnerable Throughout the period there were no changes in the surgical
plaque (Type 1) or myocardial oxygen supply/demand imbal- procedures, anaesthetic or intensive care unit (ICU) manage-
ance (Type 2), which appears to be the most common cause of ment. Written informed consent was collected prior to surgery
this complication after cardiac surgery [4]. Our group previously from all patients. The study was approved by the Ethics
reported no clinically relevant sex-based differences in Committee.
troponin-I (Tn-I) levels after cardiac surgery, without periopera- Gender, age, the Parsonnet score, APACHE-II and -III scores, in-
tive STEMI, in patients undergoing CABG or valve surgery [5]. cidence of AMI (STEMI and non-STEMI), Tn-I curve and peak of
This prompted us to investigate whether a patient’s age and sex this curve, in-hospital mortality and length of ICU stay were all
may impact the incidence and outcome of acute myocardial recorded. Patients were classified into three groups: non-AMI,
© The Author 2012. Published by Oxford University Press on behalf of the European Association for Cardio-Thoracic Surgery. All rights reserved.
C. Javierre et al. / Interactive CardioVascular and Thoracic Surgery 29
Table 1: General characteristics of the patients
Age Sex Mean Statistical

group group differences
(years)
a
Age (years) <70 Women 59.6 ± 8.8
Men 58.1 ± 9.5
≥70 Women 74.7 ± 3.3
Men 74.7 ± 3.4
ORIGINAL ARTICLES
Body mass index <70 Women 28.1 ± 5.1
Men 27.9 ± 4.1
≥70 Women 28.5 ± 4.2
Men 27.3 ± 3.5
Parsonnet score <70 Women 12.6 ± 7.3
Men 9.4 ± 7.1
≥70 Women 13.6 ± 6.4
Men 12.1 ± 7.1
APACHE-II <70 Women 11.3 ± 3.8
Men 10.6 ± 4.3
≥70 Women 13.9 ± 4.2
Men 13.7 ± 4.2
APACHE-III <70 Women 45.6 ± 16.1 Figure 1: Patients’ distribution between AMI groups. Percentage of patients in
Men 43.0 ± 16.3 the STEMI group and non-STEMI group with respect to age and sex
≥70 Women 56.3 ± 18.1 shows significantly higher values in STEMI (P < 0.001) and non-STEMI groups
Men 56.4 ± 16.5 (P < 0.001) in men. In the group ≥70 years old, both sexes were more likely to
Cross-clamp time <70 Women 69.4 ± 28.3 suffer from non-STEMI (P < 0.05.)
(min) Men 69.7 ± 26.1
≥70 Women 72.8 ± 27.8
Men 72.3 ± 24.9 groups (non-AMI, STEMI and non-STEMI) for the quantitative
Cardiopulmonary <70 Women 103.6 ± 37.6
bypass time Men 108.9 ± 35.0
samples with a normal distribution. For non-parametric data, the
(min) ≥70 Women 111.8 ± 41.4 χ 2 test was used to analyse differences in sex, age and/or clinical
Men 113.6 ± 34.3 groups. Data are reported as means ± SD, with P-values <0.05
considered significant.
Incidence (%)
a
Hypertension <70 Men 51.90
Women 56
≥70 Men 76.3 RESULTS
Women 70.8
b
Diabetes <70 Men 10.3
Women 5.6 Acute myocardial infarction incidence
≥70 Men 10.1
Women 8.3 After applying the exclusion criteria to the initial sample of 2434
b
Dyslipaemia <70 Men 44.9 consecutive patients, the study population included 2038
Women 50.1
≥70 Men 48.4 b patients: 1276 men and 762 women. Of these, 1271 (649 men
Women 56 and 622 women) underwent valve surgery, 614 (496 men and
Chronic renal failure <70 Men 3 118 women) CABG and 153 underwent (114 men and 39
Women 3.8 women) both types of surgery. Age was <70 years in 1164
≥70 Men 4.8 b
Women 8.3
patients [798 (68.6%) men and 366 (31.4%) women] and ≥70
years in 874 [478 (54.7%) men and 396 (45.3%)]. Sex
a
Statistically significant differences between age groups (P < 0.05).
differences between the two age groups were statistically
b
Statistically significant differences between sex groups (P < 0.05). significant (P < 0.001).
AMI incidence was significantly lower in the valve surgery
group (valve group = 5.4%; CABG group = 14.6%, CABG and valve
STEMI according to classical ECG criteria [6] and non-STEMI, group = 19.8%; P < 0.001), although the distribution of STEMI and
defined as a Tn-I peak >20 ng/ml [7, 8] and a late peak in Tn-I [9]. non-STEMI was similar for the three types of surgery (valve
All blood samples were obtained using a central venous cath- surgery: STEMI 2.3% and non-STEMI 3.1%; CABG: STEMI 7.1%
eter. Serum Tn-I samples were obtained immediately after and non-STEMI 7.6%; both types of surgery: STEMI 10.5% and
surgery upon ICU admission, and again 6, 12, 24 and 48 h later. non-STEMI 9.3%). Age ≥70 years old was not associated with a
The samples were measured with a Dimension RxL analyser higher incidence of AMI after cardiac surgery when compared
(Dade Behring, Newark, DE, USA). with <70 years, but age did imply a change in the predominant
type of AMI. Patients ≥70 years old were more frequently
non-STEMI (6.3%) than STEMI (3.5%), in contrast to patients <70
Statistical methods years who demonstrated more STEMI (5.1%) than non-STEMI
(4.0%) (P < 0.05; Fig. 1). Overall, men were more likely than
ANOVA with post hoc Bonferroni correction was applied in order women to suffer from AMI after cardiac surgery (11.8% vs. 5.6%),
to evaluate any differences in sex and/or age and/or clinical as a result of the higher frequency of STEMI (6% of men vs. 1.8%
30 C. Javierre et al. / Interactive CardioVascular and Thoracic Surgery
of women; P < 0.0001). This difference was found in both age (F = 4.7, P < 0.05) and in women with STEMI and non-STEMI
groups: <70 (6.7% of men vs. 1.9% of women; P < 0.001) and ≥70 (F = 3.8, P < 0.05; Table 4).
years old (5.5% of men vs. 1.8% of women; P < 0.005; Table 2). As regards the Parsonnet score, there were no significant sex
differences in the whole sample or in patients with AMI.
However, a significant difference was observed between patients
aged <70 years (the Parsonnet score 10.4 ± 7.4) and those ≥70
Troponin-I peak years old (12.7 ± 6.8; P < 0.001).
The Tn-I peak was significantly lower in the non-AMI group

(10.4 ± 23.7 µg/l) than in the STEMI (62.8 ± 117.5 µg/l) and
non-STEMI groups (63.8 ± 107.1 µg/l; P < 0.001). At 48 h post- DISCUSSION
surgery, Tn-I was 4.4 ± 5.6 µg/l in the non-AMI group, 24.6 ±
28.6 µg/l in the STEMI group and 25.3 ± 34.6 µg/l in the In this study, we showed that both age and sex may influence
non-STEMI group (P < 0.001). The Tn-I peak also occurred sig- the incidence, mortality and morbidity of AMI after cardiac
nificantly earlier after surgery in the non-AMI group (10.9 ± 9.0 surgery. Thus (i) men are more likely to suffer from AMI after
h) than in the STEMI (17.0 ± 9.6 h) and non-STEMI groups (16.5 cardiac surgery, especially STEMI; (ii) both men and women are
± 8.7 h; P < 0.001). Tn-I peak was significantly higher in patients more likely to suffer from non-STEMI when aged ≥70 years; (iii)
≥70 than in those aged <70 years old (F = 19.95, P < 0.001). the Tn-I peak increases in the AMI group when patients are ≥70
These differences resulted from the increase observed in men years old; (iv) women with STEMI have a lower Tn-I peak than
with STEMI and non-STEMI (F = 5.67, P < 0.005; Fig. 2). men; (v) in-hospital mortality in patients with AMI increases only
in women aged ≥70 years old; and (vi) in the group ≥70 years
old, ICU stay increases in the non-AMI group in both men and
women, while the STEMI and non-STEMI groups show only a
Mortality, length of stay in intensive care unit longer stay in women.
and other data We observed sex differences between patients <70 and ≥70
years old receiving cardiac surgery. This is because men suffer
The in-hospital mortality rate was significantly higher in patients coronary syndromes younger than women, and therefore
≥70 years (7.3%) than in those <70 years old (3.3%) (P < 0.001) undergo surgery at an earlier age [10, 11]. The differences in the
because of the increase in mortality observed for men with present study are mostly owing to CABG, performed in 496 men
non-AMI (2.1% vs. 6.3%, P < 0.001) and women with STEMI (0% and 118 women.
vs. 28.6%, P < 0.05) and non-STEMI (0% vs. 36.8%, P < 0.05; There were no differences in the frequency of AMI after
Table 3). ICU stay was also significantly longer in patients ≥70 cardiac surgery between patients <70 and ≥70 years old.
years old than in those aged <70 (F = 10.0, P < 0.001). Prolonged However, the predominant type of AMI switched from STEMI in
ICU stay was observed for both sexes in patients with non-AMI patients aged <70 to non-STEMI in those ≥70 years old. The
relatively frequent occurrence of perioperative non-STEMI,
Table 2: Patients’ distribution between groups
Age (years) Sex AMI group n (%)
<70 Women (n = 366) Non-AMI 349 (95.4)

STEMI 7 (1.9)
Non-STEMI 10 (2.7)
Men (n = 798) Non-AMI 708 (88.7)
STEMI 53 (6.7)a
Non-STEMI 37 (4.6)a,b
Total (n = 1164) Non-AMI 1057 (90.8)
STEMI 60 (5.1)
Non-STEMI 47 (4.0)
≥70 Women (n = 396) Non-AMI 370 (93.5)
STEMI 7 (1.8)
Non-STEMI 19 (4.7)
Men (n = 478) Non-AMI 418 (87.5)
STEMI 27 (5.5)
Non-STEMI 33 (7.0)b
Total (n = 874) Non-AMI 788 (90.2)
STEMI 34 (3.5)
Non-STEMI 52 (6.3)
Number and percentage of patients in the non-AMI group, STEMI

group and non-STEMI group with respect to age and sex. Figure 2: Tn-I peak. Non-AMI group, STEMI group and non-STEMI group.
a
Significantly higher values in STEMI (P < 0.001) and non-STEMI Mean values and standard deviation for the Tn-I peak of the three groups.
groups (P < 0.001) in men. Lower in the non-AMI group when compared with the STEMI and
b
In the group ≥70 years old, both sexes were more likely to suffer non-STEMI groups (P < 0.001). Higher in patients ≥70 when compared with
from non-STEMI (P < 0.05). those aged <70 years (P < 0.05) due to the increase in men with STEMI and
non-STEMI.
C. Javierre et al. / Interactive CardioVascular and Thoracic Surgery 31
did not have any substantial age-related differences in the STEMI

Table 3: In-hospital mortality incidence, in contrast to the findings for non-surgery acute coron-
ary syndrome. This suggests that in the case of AMI after cardiac
In-hospital mortality Exitus (%) Exitus (%) surgery, the surgical procedures, or the reactions to them, are an
important cause of STEMI. The lower prevalence of AMI in
<70 years women is suggestive of a greater resistance to this condition, as is
Women 12/366 3.3 Non-AMI 12/349 (3.5) generally the case with acute coronary syndromes [10, 11].
STEMI 0/7 (0)a
Non-STEMI 0/10 (0)b
We did not use the universal definition of myocardial infarc-
Men 27/796 3.4 Non-AMI 15/708 (2.1) tion [13], because it did not exist when we begun the protocol
STEMI 8/53 (15.4) (2004) and also because it would be difficult to individualize the
ORIGINAL ARTICLES
Non-STEMI 4/37/11.4) ‘arbitrary convention’ of increases in the biomarker of more than
Total 39/1164 3.3 Non-AMI 27/1057 (3.2) five times the 99th percentile of the normal reference range, as
STEMI 8/60 (13.6)
Non-STEMI 4/47 (8.9) this reference will vary depending on the surgery characteristics
≥70 years (CABG, valve or valves, CABG plus valve, with or without CPB,
Women 30/396 7.6 Non-AMI 21/370 (5.9) and so on). The similarity in the time of the Tn-I peaks between
STEMI 2/7 (28.6)a the STEMI and non-STEMI groups, and their difference with
Non-STEMI 7/19 (36.8)b
Men 34/478 7.1 Non-AMI 26/418 (6.3)
regard to the non-AMI group stresses the accuracy of the late
STEMI 4/27 (16.7) Tn-I peak for detecting perioperative AMI.
Non-STEMI 4/33 (11.4) The Tn-I concentration and its evolution can indicate AMI in
Total 64/874 7.3 Non-AMI 47/788 (6.0) both STEMI and non-STEMI groups. The Tn-I elevation criteria
STEMI 7/34 (20.6) obtained from magnetic resonance imaging [9] seems to be an
Non-STEMI 10/52 (19.2)
important step forward in terms of identifying a condition that is
often difficult to diagnose. New imaging studies have reported
In-hospital mortality in the different groups with respect to age and
sex. that in the event of a perioperative AMI, the Tn-I peak occurs
a
Statistically significant increase in women ≥70 years old in the STEMI later and the Tn-I concentration at 48 h remains high. These two
group (P < 0.05). data are the most reliable signs of myocardial necrosis [9].
b
Statistically significant increase in women ≥70 years old in the The increase in Tn-I peak with age observed in the present
non-STEMI group (P < 0.05).
study could be owing to a greater sensitivity to aggressions. The
lower Tn-I peak concentrations in women with STEMI may have
some mid- and long-term advantages that merit further
investigation.
Table 4: ICU stay In general, it seems that women are less likely to suffer AMI
after cardiac surgery, but that they tolerate this condition worse
than men; furthermore, the potentially negative repercussions of
Age (years) Sex Group ICU stay (h)
age affects only some of the variables measured. At all events,
<70 Women Non-AMI (110.1 ± 126.8) 106.8 ± 122.2 women with non-perioperative STEMI also had higher adjusted
STEMI 221.7 ± 277.2 mortality rates than men [1]. Compared with men, women
Non-STEMI 147.2 ± 103.9 present some physiological differences in relation to cardiac is-
Men Non-AMI (107.8 ± 126.2) 100.6 ± 117.6
chaemia, such as higher myocardial oxygen consumption and
STEMI 164.0 ± 160.4
Non-STEMI 165.6 ± 187.3 lower myocardial glucose extraction fraction and utilization; they
Total Non-AMI (108.5 ± 126.3) 102.7 ± 119.1a also show greater sympathetic activation that lasts until reso-
STEMI 170.8 ± 175.9 lution at 9 months following uncomplicated AMI [14]. These dif-
Non-STEMI 161.6 ± 171.8 ferences may explain some of the sex-based differences in AMI
≥70 Women Non-AMI (150.9 ± 182.5) 140.7 ± 165.1
STEMI 356.7 ± 309.7 observed in this study.
Non-STEMI 269.3 ± 322.9 The Parsonnet scores’ similarity between the sexes suggests
Men Non-AMI (137.7 ± 157.9) 132.0 ± 158.6 that the differences found are not due to any pre-surgery char-
STEMI 169.0 ± 106.4 acteristics. This Parsonnet scores’ similarity between patients
Non-STEMI 183.1 ± 172.3
with and without non-STEMI and the lower Parsonnet scores in
Total Non-AMI (143.6 ± 169.5) 136.0 ± 161.6a
STEMI 211.4 ± 185.0 patients who go on to develop AMI illustrate the lack of sensitiv-
Non-STEMI 213.4 ± 237.0 ity of this score for detecting a propensity to perioperative AMI.
By contrast, the finding of higher APACHE-II and -III scores in
Non-AMI group, STEMI group and non-STEMI group. Mean values patients with AMI shows that these scores are sensitive enough
and SD for the ICU stay of the three groups. to detect important physiological changes resulting from AMI.
a
Significantly lower values in the non-AMI group with respect to the The longer ICU stay in women may be a sex-related difference
STEMI (P < 0.001) and non-STEMI (P < 0.001) groups.
that is not correlated with the variables studied here. As regards
the longer stay and mortality among patients with AMI, we do
not believe that this merits further comment.
which is very often subendocardial [9], in both men and women, In summary, patients with age ≥70 years old do not have a
suggests that subendocardial perfusion deteriorates with age. higher incidence of AMI after cardiac surgery, though advanced
Despite the increase in non-STEMI observed in women ≥70 age is associated with a higher frequency of non-STEMI, Tn-I
years old, the incidence of AMI was lower in women than men, as peak, mortality and length of ICU stay when compared with
is the case in non-surgery-related AMI [12]. In addition, women patients <70 years. Interestingly, after cardiac surgery, men suffer
32 C. Javierre et al. / Interactive CardioVascular and Thoracic Surgery
from AMI ( particularly STEMI) more often than women, regard- coronary revascularization: a report from the bypass angioplasty revascu-
less of age. However, AMI (STEMI and non-STEMI) in women larization investigation (BARI). Circulation 1998;98:1279–85.
[4] Landesberg G, Beattie WS, Mosseri M, Jaffe AS, Alpert JS. Perioperative
had a notable impact on excess mortality and ICU stay observed myocardial infarction. Circulation 2009;119:2936–44.
in patients ≥70 years. Older men are more likely to suffer from [5] Ricart A, Farrero E, Ventura JL, Javierre C, Carrio L, Rodríguez D et al. Are
non-STEMI and when they suffer STEMI or non-STEMI, they there sex-based differences in serum troponin I after cardiac surgery?
present a higher increase in the Tn-I peak. Crit Care Med 2009;37:2210–15.
[6] Alpert JS, Thygesen K, Antman E, Bassand JP. Myocardial infarction rede-
fined: a consensus document of the joint European Society of
Cardiology/American College of Cardiology Committee for the redefin-
CONCLUSION ition of myocardial infarction. J Am Coll Cardiol 2000;36:959–69.
[7] Benoit MO, Paris M, Silleran J, Fiemeyer A, Moatti N. Cardiac troponin I:
Clinical and Tn-I peak differences are expected in relation with its contribution to the diagnosis of perioperative myocardial infarction
age and gender after AMI post-cardiac surgery. and various complications of cardiac surgery. Crit Care Med 2001;29:
1880–6.
[8] Bojar RM. Manual of perioperative care in adult cardiac surgery, 4th ed.
Oxford: Blackwell, 2005.
ACKNOWLEDGEMENTS [9] Selvanayagam JB, Pigott D, Balacumaraswami L, Petersen SE, Neubauer S,
Taggart DP. Relationship of irreversible myocardial injury to troponin I
and creatine kinase-MB elevation after coronary artery bypass surgery:
We thank J. Valero for his analyses and the nursing team at our
insights from cardiovascular magnetic resonance imaging. J Am Coll
ICU. Cardiol 2005;45:629–31.
[10] Barrett-Connor E. Sex differences in coronary heart disease. Why are
Conflict of interest: none declared. women so superior? The 1995 Ancel keys lecture. Circulation 1997;95:
252–64.
[11] Wexler LF. Studies of acute coronary syndromes in women: lessons for
everyone. N Engl J Med 1999;341:275–6.
REFERENCES [12] Vaccarino V, Parsons L, Every NR, Barron HV, Krumholz HM. Sex-based
differences in early mortality after myocardial infarction. National
[1] Jneid H, Fonarow GC, Cannon CP, Hernandez AF, Palacios IF, Maree AO Registry of myocardial infarction to participants. N Engl J Med 1999;341:
et al. Sex differences in medical care and early death after acute myocar- 217–25.
dial infarction. Circulation 2008;118:2803–10. [13] Thygesen K, Alpert JS, White HD, Jaffe AS, Apple FS, Galvani M et al.
[2] Frink RJ. Gender gap, inflammation and acute coronary disease: are Universal definition of myocardial infarction. Circulation 2007;116:
women resistant to atheroma growth? Observations at autopsy. 2634–53.
J Invasive Cardiol 2009;21:270–7. [14] Hogarth AJ, Graham LN, Mary DA, Greenwood JP. Gender differences in
[3] Jacobs AK, Kelsey SF, Brooks MM, Faxon DP, Chaitman BR, Bittner V sympathetic neural activation following uncomplicated acute myocardial
et al. Better outcome for women compared with men undergoing infarction. Eur Heart J 2009;30:1764–70.
Interactive CardioVascular and Thoracic Surgery 14 (2012) 709–713 INSTITUTIONAL REPORT
doi:10.1093/icvts/ivs050 Advance Access publication 24 February 2012
Reoperation for bleeding in cardiac surgery

Katrine Lawaetz Kristensen*, Line Juul Rauer, Poul Erik Mortensen and Bo Juel Kjeldsen
Department of Cardio-, Thoracic- and Vascular Surgery, Odense University Hospital, Odense, Denmark
CARDIAC GENERAL
* Corresponding author. Katrine Lawaetz Kristensen, Dannebrogsgade 21 2.tv, 9000 Aalborg, Denmark. Tel: +45-20940296; e-mail: katrinelawaetz@gmail.com
(K.L. Kristensen).
Received 18 August 2011; received in revised form 20 December 2011; accepted 2 January 2012
Abstract
At Odense University Hospital (OUH), 5–9% of all unselected cardiac surgical patients undergo reoperation due to excessive bleeding.
The reoperated patients have an approximately three times greater mortality than non-reoperated. To reduce the rate of reoperations
and mortality due to postoperative bleeding, we aim to identify risk factors that predict reoperation. A total of 1452 consecutive
patients undergoing cardiac surgery using extracorporeal circulation (ECC) between November 2005 and December 2008 at OUH were
analysed. Statistical tests were used to identify risk factors for reoperation. We performed a case-note review on propensity-matched
patients to assess the outcome of reoperation for bleeding regarding morbidity and mortality. In total, 101 patients (7.0%) underwent
surgical re-exploration due to excessive postoperative bleeding. Significant risk factors for reoperation for bleeding after cardiac surgery
was low ejection fraction, high EuroSCORE, procedures other than isolated CABG, elongated time on ECC, low body mass index,
diabetes mellitus and preoperatively elevated s-creatinine. Reoperated patients significantly had a greater increase in postoperative
s-creatinine and higher mortality. Surviving reoperated patients significantly had a lower EuroSCORE and a shorter time on ECC
compared with non-survivors. The average time to re-exploration was 155 min longer for non-survivors when compared with survivors.
Keywords: Cardiac surgery • Reoperation • Postoperative bleeding • Risk factors
INTRODUCTION 1452 patients found, 101 patients (7.0%) underwent at least one
reoperation within 24 h due to excessive postoperative bleeding.
Risk of reoperation due to bleeding after cardiac surgery is Pre-, peri- and postoperative characteristics were collected from
shown to be 2.2–4.2% [1–6]. Risk factors are high age, low body WDHReg complemented by case records. Information included
mass index (BMI) or body surface area (BSA), time on extracor- the patient’s age, type of operation, complications, bleeding and
poreal circulation (ECC), five or more anastomosis or possible causes of death. We qualified risk factors for reoperation
non-elective operation [2–4]. Reoperated patients have a two to by comparing the results of 101 reoperated patients with the
six times greater mortality [1, 3–5] and a greater morbidity remaining 1351 patients.
regarding renal and pulmonary function, sepsis and arrhythmia Using recorded operative details, we classified the bleeding as
[1, 4, 5]. Postoperative bleeding can be due to surgical or coagu- coagulopathic or surgical. Blood or haematoma without ongoing
lopathic factors [1]. Patient-related factors are also indicative for bleeding and/or diffuse oozing was defined as coagulopathic
reoperation due to bleeding. bleeding. Specific bleeding requiring clips or suture was defined
At Odense University Hospital (OUH), 5–9% of all unselected as surgical bleeding.
cardiac surgical patients undergo reoperation due to excessive
bleeding. The reoperated patients had three times greater mor-
tality and, to lower the rate of reoperation due to bleeding and
improve the mortality, we have tried to identify risk factors in Propensity-matched group
patients as well as in procedures.
To assess the outcome of reoperation for bleeding regarding
morbidity and mortality, we performed a case-note review on
propensity-matched patients. Patients requiring re-exploration
MATERIALS AND METHODS for bleeding were propensity matched with a unique patient not
re-explored using the following criteria: priority of the operation,
Patient population and data age, BMI, medicine (acetylsalicylic acid, clopidogrel, coumadine
or heparin) within the last 5 days, EuroSCORE (European System
Patient data were collected consecutively from The Western for Cardiac Operative Risk Evaluation) and sex. When possible
Denmark Heart Registry [7] (WDHReg) during 18 November we also matched patients with regard to ejection fraction (EF),
2005 until 31 December 2008 at OUH. During this period, we diabetes mellitus (DM) and the type of procedure. The men-
included all patients undergoing cardiac surgery using ECC. Of tioned criteria are the result of a literature study [2, 5]. Matching
710 K.L. Kristensen et al. / Interactive CardioVascular and Thoracic Surgery
the patients using these criteria, we minimized their influence as

confounders when analysing the consequences of reoperations.
Mortality was defined as death within 30 days after the first
operation. Complications such as perioperative myocardial
infarction, transient cerebral ischaemic attack (TIA), stroke and
sternal infection had to be diagnosed by a doctor within the
same hospital admission as the index surgery. The morbidity
definitions were according to the definitions of the protocol at
WDHReg [7]. Stroke is defined as a neurological deficit of cere-
brovascular cause that persists beyond 24 h or is interrupted by
death within 24 h. TIAs share the same aetiology as strokes, but
the symptoms resolve within 24 h. Sternal infections often
appeared after the discharge, so we searched the case records Figure 1: Causes of death among the 16 reoperated and 6 matched patients.
after discharge.
The decision to perform resternotomy was made by the
surgeon on call and based on conventional guidelines: (i) drain-
age of >500 ml in the first hour, total drainage of >800 ml in the
first 2 h, 900 ml in the first 3 h, >1000 ml in the first 4 h or 1200 ml
in the first 5 h; (ii) sudden massive bleeding or cardiac
tamponade.
Differences between the two groups were explored using an in-
dependent t-test for continuous variables and Fisher’s exact test
and Pearson’s χ2 test for categorical variables. Data were given as
mean values, standard deviations of mean values and in
numbers and percentages. In all cases, a P-value <0.05 was con-
sidered as significant. Figure 2: Types of bleeding found at reoperation. Forty-two patients had a
coagulopathic bleeding, 57 patients had a surgical bleeding and 2 patients
Statistical analysis was performed with SPSS (SPSS Inc., PASW had an unknown type of bleeding as it was not described further in the oper-
Statistics 18). ation description.
RESULTS
Causes of death were divided into circulatory, multi-organ
Table 1: Risk factors for reoperation
failure, bowel ischaemia and unknown (see Fig. 1).
Variables Reoperated Not reoperated P-value
(n = 101), (n = 1351),
The total population mean ± SD or mean ± SD or
number (%) number (%)
Of the 1452 patients undergoing on-pump cardiac surgery
Age (year) 67.0 ± 11.6 66.1 ± 10.6 0.414
during the study period, 101 patients (7.0%) underwent a surgi- Sex (male) 73 (72.3) 984 (72.8) 0.908
cal re-exploration for postoperative bleeding (types of bleeding, Priority (acute) 15 (14.9) 125a (9.3) 0.079
Fig. 2). EuroSCORE 12.2 ± 13.1 8.3 ± 11.7 0.001
Mortality in patients re-explored for surgical bleeding and EF 48.5 ± 12.9 51.3 ± 11.8 0.022
Procedure: isolated 25 (24.8) 564 (41.7) 0.001
coagulopathic bleeding was 19 and 10.5%, respectively. CABG
Patients in the re-explored group had a higher mortality when Time on ECC (min) 158.3 ± 83.9 126.7 ± 57.6 0.000
compared with patients who were not re-explored (15.8 versus BMI (kg/m2) 25.4 ± 4.0 27.1 ± 4.4 0.000
5.7%, P < 0.001). Predictors for reoperation due to bleeding are DM 25 (25.8) 199a (14.7) 0.010
Hypertension 51 (50.5) 596a (44.4) 0.255
listed in Table 1.
Previous AMI 31 (30.7) 395a (29.3) 0.821
Previous PCI 16 (15.8) 176a (13.1) 0.447
Medicine within 5 days 23 (22.8) 302 (22.4) 0.902
Creatinine >134 µmol/l 17 (16.8) 105a (7.8) 0.004
Matched group at admission
The group of reoperated patients and propensity-matched group a

Number (n) different from 1351 due to missing data. Results are
were homogeneous with respect to preoperative and periopera- calculated with another ‘n’. Priority, DM, creatinine and epidural,
tive variables. However, preoperative s-creatinine was significantly n = 1350, hypertension, n = 1341; previous AMI, n = 1346; previous
higher in the re-explored group when compared with the PCI, n = 1340.
control group (Table 2).
K.L. Kristensen et al. / Interactive CardioVascular and Thoracic Surgery 711
Table 2: Comparison of case and match group Table 4: Comparing s-creatinine between reoperated
and match
Variables Reoperated (n = 101), Match (n = 101), P-value
mean ± SD or mean ± SD or Variables Reoperated Match P-value
number (%) number (%) (n = 82), (n = 97),
mean ± SD mean ± SD
Priority (acute) 15 (14.9) 15 (14.9) 1.000
Age (year) 67.0 ± 11.6 67.0 ± 11.6 0.990 S-creatinine admission 96.1 ± 16.8 94.3 ± 16.4 0.476
CARDIAC GENERAL
BMI (kg/m2) 25.4 ± 4.0 25.6 ± 4.2 0.711 (µmol/l)
EuroSCORE 12.2 ± 13.1 13.2 ± 16.1 0.647 S-creatinine max 160.4 ± 89.4 126.1 ± 47.3 0.001
Medicine 23 (22.8) 23 (22.8) 1.000 S-creatinine discharge/ 107.5 ± 39.2 103.5 ± 32.8 0.458
within 5 death
days Average increase 64.3 ± 84.2 31.8 ± 44.2 0.001
Sex (male) 73 (72.3) 72 (71.3) 1.000 Difference 11.5 ± 33.1 9.2 ± 28.0 0.626
EF 48.5 ± 12.9 49.1 ± 12.4 0.710 discharge-admission
DM 25 (24.8) 17 (16.8) 0.225
CABG 59 (58.4) 58 (57.4) 1.000
Aortic valve 55 (54.5) 42 (41.6) 0.091 Patients with creatinine above 134 at admission were excluded from
Mitral valve 11 (10.9) 17 (16.8) 0.309 this comparison. Furthermore, two patients are excluded due to
Other 33 (32.7) 22 (21.8) 0.114 missing data.
procedure
S-creatinine 109.8 ± 39.4 97.6 ± 24.2 0.008
(µmol/l)
Table 5: Comparison of reoperated survivors and

non-survivors
Variables Non-survivors Survivors P-value

Table 3: Comparing the outcome between reoperated (n = 16), (n = 85),
and match group mean ± SD or mean ± SD or
Variables Reoperated Match P-value
Sex (male) 12 (75.0) 61 (71.8) 1.000
(n = 101), (n = 101),
Age (year) 70.8 ± 5.8 66.3 ± 12.3 0.151
Mean ± SD or Mean ± SD or
EuroSCORE 19.9 ± 20.0 10.8 ± 10.9 0.010
BMI (kg/m2) 26.7 ± 5.1 25.1 ± 3.9 0.166
EF 43.8 ± 12.1 49.3 ± 13.0 0.114
TIAa 2 (2.0) 0 (0.0) 0.129
DM 5 (31.3) 20 (23.5) 0.535
Strokea 0 (0.0) 3 (3.0)
Priority (acute) 3 (18.8) 12 (14.1) 0.702
Lacunar infarcta 3 (3.0) 1 (1.0)
Medicine <5 days 4 (20.0) 19 (22.4) 0.756
Sternal infection 0 (0.0) 1 (1.0) 1.000
ECC-time (min) 216.7 ± 92.7 147.3 ± 77.9 0.002
AMI
S-creatinine admission 127.4 ± 42.5 106.5 ± 38.2 0.052
Yes 4 (4.0) 4 (4.0) 0.929
(µmol/l)
Uncertain 4 (4.9) 3 (3.0)
S-creatinine maxa 226.6 ± 84.1 172.4 ± 105.4 0.081
Mortality 16 (15.8) 6 (5.9) 0.040
S-creatinine increasea 97.0 ± 66.8 68.6 ± 92.3 0.290
Drainage during the 2905.3 ± 2332.9 829.9 ± 626.3 0.000
Type of bleeding 0.256
admission at intensive
(surgical)
care unit (ml)b
Isolated CABG 3 (18.8) 22 (25.9) 1.000
Drainageb 2049.3 ± 1135.1 1783.5 ± 1049.6 0.376
a
Number different from 202; reoperated 99, match 100. Time to reoperation 561.4 ± 331.1 406.4 ± 275.4 0.061
b
Number different from 202, 91 in both groups. (min)c
a
97 patients; 13 non-survivors and 84 survivors.
b
95 patients; 15 non-survivors and 80 survivors, drainage until
reoperation.
c
Patients undergoing re-exploration had higher mediastinal 99 patients; 14 non-survivors and 85 survivors.
drainage in the intensive care unit (2905.3 versus 829.9 ml).
Furthermore, the mortality rate was higher in the re-explored
group (15.8 versus 5.9%) (Table 3).
To analyse the impact of re-exploration on the renal function, Surviving patients in the re-explored group had a lower
we excluded patients with preoperatively elevated s-creatinine EuroSCORE than non-survivors (P = 0.010). Additionally, a sig-
(s-creatinine > 134 µmol/l at admission). This reduced the group nificant difference in the re-explored group is a longer ECC
of re-explored patients from 101 to 82 patients and the control time. On average, the surviving patients were re-explored
group was reduced to 97. The increase in postoperative 155 min earlier than the others (561.4 versus 406.4 min)
s-creatinine is greater for the re-explored when compared with (Table 5).
the control group (P = 0.001). The maximum value of s-creatinine One surgeon had a greater risk for re-exploration compared
is also different in the two groups (160.4 versus 126.1 µmol/l) with the other surgeons (surgeon 12, P = 0.003) (Fig. 3), but no
(Table 4). difference in mortality was seen among the surgeons.
712 K.L. Kristensen et al. / Interactive CardioVascular and Thoracic Surgery
We differentiated the haemorrhage in coagulopathic and

surgical to assess whether it had an influence on mortality or
not. While 56.4% of the patients had surgical bleeding, the
remaining 41.6% were coagulopathic bleeding. Mortality was
19.0% for patients with surgical bleeding, and 10.5% for
patients with coagulopathic bleeding; this difference is not
significant (P = 0.256). Hall et al. [1] differentiated the haemor-
rhage between surgical and coagulopathic to determine
the differences in patients’ outcome. Of the 2263 patients
undergoing cardiac surgery, 3.6% were reoperated due to
bleeding. While 66% had a surgical bleeding, the remaining
34% had coagulopathic bleeding. The mortality for reoperated
group was 8.7 and 12.5%, respectively, and for non-
Figure 3: Relative risk for reoperation with respect to the surgeon. The sur- reexplored it was 2%. Our study did not find the same ten-
geons’ identification numbers are listed to the right. The relative risk is shown dency as Hall et al.’s study regarding mortality as our patients
in the figure. with surgical bleedings had a greater mortality than the
coagulopathic.
Morbidity could be examined using the propensity-matched
DISCUSSION score. We did not find a significant difference in morbidity
between the re-explored group compared with the propensity-
Of 1452 patients undergoing cardiac surgery, 101 patients (7.0%) matched group regarding AMI, sternal infection, stroke and the
required at least one reoperation within 24 h due to excessive level of creatinine after discharge. This is similar to the findings
bleeding. This rate is greater than previously reported rates of from Karthik et al. [2]. It should be stressed that these post-
2.2–4.2% [1–6] and also greater than the target rate of the surgi- operative complications are infrequent and a large number of
cal ward of 5.0%. It has not been possible for the surgeons to patients are needed to show a difference.
stay within the recommended rates over the examined 3-year Focusing on renal function we excluded patients in both the
period, but the rate is no greater than the national average rate reoperated group and the propensity-matched group with pre-
for Denmark in the given period estimated by WDHReg [7]. operative elevated creatinine (>134 µmol/l). Having only patients
It has been shown that reoperation is associated with an with normal pre-operative creatinine in both groups, we evalu-
increased mortality [1, 3–6]. Our study confirmed this by showing ated the level of postoperative rise of s-creatinine, and patients
an increase in mortality by three times (2.7–2.8) for re-explored reoperated for bleeding had a greater increase in s-creatinine
patients compared with both all non-re-explored and matched both average and maximum which was significant. Surprisingly,
control group. there was no difference in the values at discharge/death
We identified seven significant pre- and intraoperative vari- between the two groups.
ables that were predictive for re-exploration due to bleeding: Among the patients undergoing reoperation the prolonged
low BMI; low EF; high EuroSCORE; preoperatively increased cre- ECC time was associated with increased mortality. We did not
atinine; DM; procedures other than coronary artery bypass graft find any significant difference regarding the time to
surgery (CABG) and prolonged time on ECC. These variables are re-exploration between the reoperated survivors and
intertwined in various ways. Other studies have found age and non-survivors, but one could consider that it might make a dif-
priority associated with the risk of reoperation, but our study did ference if the patients underwent earlier re-exploration. The
not show this association [2–4]. We did not find any significant average time to re-exploration in the two groups was 561 and
association between a medical history with hypertension or pre- 406 min, respectively. This is in conflict with the guidelines
vious AMI/PCI (respectively, acute myocardial infarction and per- regarding postoperative bleeding and reoperation outlining the
cutaneous coronary intervention) and risk for reoperation. volume of bleeding up to 5 h (300 min) after the first operation.
Interestingly we found that patients with DM have a significantly
greater risk for reoperation due to bleeding which is not seen in
other studies [2–4]. Thus the risk for reoperation seems to
depend on the type of preoperative morbidity the patients CONCLUSIONS
present themselves with.
Karthik et al. [2] studied 2898 patients undergoing CABG with Our aim was to find factors with significant influence on the
a reoperation rate of 3.1% to identify risk factors for reoperation. risk for reoperation due to haemorrhage after cardiac surgery
Variables such as low BMI, non-elective surgery, increasing and evaluate the consequence of the reoperation. Like several
number of grafts and increasing age was identified as having a other studies, we found a higher mortality for reoperated
significantly greater risk for reoperation for bleeding. Moulton patients. We also found that low EF, high EuroSCORE, proce-
et al. [4] studied 6015 patients undergoing cardiac surgery with a dures other than CABG, the time on ECC, low BMI and
reoperation rate of 4.2%. High age, preoperative renal insuffi- preoperatively elevated s-creatinine had a significant influence
ciency, procedures other than CABG and prolonged time on on the risk for reoperation. These variables are intertwined in
ECC were at greater risk of needing reoperation for bleeding. various ways. We found a significant association between DM
Dacey et al. [3] studied 8586 patients with a reoperation rate of and the risk for reoperation.
3.6%. In this substantial study, high rates of re-exploration for By comparing the reoperated patients with a propensity-
haemorrhage were observed in patients with increasing age, matched group, we found that the reoperated patients had a sig-
smaller BSA, prolonged time on ECC and number of grafts. nificantly greater increase in postoperative s-creatinine.
K.L. Kristensen et al. / Interactive CardioVascular and Thoracic Surgery 713
By analysing the reoperated patients we found that the group eComment. Postoperative bleeding in cardiac surgery: the issue is not resolved
of surviving patients had a lower EuroSCORE, a shorter time on yet
the ECC and a shorter time to re-exploration.

Author: Ovidio A. Garcia-Villarreal
The haemorrhage was differentiated in coagulopathic and sur- Department of Cardiac Surgery, Hospital of Cardiovascular Disease No. 34, IMSS,
gical. In total, 56.4% of the patients had surgical bleeding. Monterrey, Mexico
Patients with high EuroSCORE, low EF, low BMI, DM, pre- doi:10.1093/icvts/ivs126
operative s-creatinine >134 µmol/l and procedures other © The Author 2012. Published by Oxford University Press on behalf of the
European Association for Cardio-Thoracic Surgery. All rights reserved.
than CABG should have a very carefully planned operation.
CARDIAC GENERAL
Preoperatively, discontinuation of pertinent medication and
I read with great interest the article by Kristensen et al. [1]. They showed an inci-
screening coagulation in blood samples could reduce coagu- dence of reexploration for bleeding after cardiac surgery of 7%. Also, it was found
lopathic bleeding. Initiatives such as checklists, action cards, that low ejection fraction, high EuroSCORE, procedures other than isolated coron-
guidelines and regular audits can help reduce surgical ary artery bypass graft surgery (CABG), prolonged time on extracorporeal circula-
causes for reoperation due to bleeding. It is mandatory to tion, low body mass, and others were significant risk factors for reoperation for
bleeding. Indeed, operative mortality (15.8%) increased by three times for reex-
strictly follow guidelines regarding reoperation for post- plored patients.
operative bleeding and thereby possibly reduce the amount It is noteworthy that not only the reoperation for excessive haemorrhage per se
of time and blood spent before performing a necessary had a negative impact on operative mortality and morbidity. Christensen et al. [2]
reoperation. demonstrated that postoperative haemorrhage exceeding 200 ml/h in any single
hour or part thereof, or 2 ml/kg/h for 2 consecutive hours in the first 6 hours after
surgery, or > 495 ml in the first 24 hours was associated with a higher 30-day mor-
tality and other major postoperative complications. In fact, death at 30-day after
Conflict of interest: none declared. surgery increased from 5.5% to 22.4% in the postoperative haemorrhage group.
When postoperative haemorrhage was present, reexploration for bleeding occurred
in 50% of the cases. Postoperative haemorrhage was also associated with ICU stays
> 72 h, and mechanical ventilation > 24 h. Possible explanations for these circum-
REFERENCES stances include more blood transfusions, a more hypovolemic status, and systemic
hypotension resulting in secondary organ failure, such as prolonged respiratory
[1] Hall TS, Brevetti GR, Skoultchi AJ, Sines JC, Gregory P, Spotnitz AJ. support, the need for renal replacement therapy, and a higher incidence of system-
Reexploration for hemorrhage following open heart surgery differenti- ic inflammatory response syndrome. Vivacqua et al. [3] have demonstrated that
ation on the causes of bleeding and the impact on patient outcomes. both greater blood transfusion and reoperation for excessive haemorrhage are in-
Ann Thorac Cardiovasc Surg 2001;7:352–7. dependently associated with an elevated risk of mortality and major morbidity.
[2] Karthik S, Grayson AD, McCarron EE, Pullan DM, Desmond MJ. Traditionally, the decision to perform resternotomy is based on conventional
Reexploration for bleeding after coronary artery bypass surgery: risk factors, guidelines: drainage of > 500 ml in the first hour, > 800 ml in the first 2 h, 900 ml in
outcomes, and the effect of time delay. Ann Thorac Surg 2004;78:527–34. the first 3 h, 1000 ml in the first 4 h, 1200 ml in the first 5 h, sudden massive bleed-
[3] Dacey LJ, Munoz JJ, Baribeau YR, Johnson ER, Lahey SJ, Leavitt BJ et al. ing, or cardiac tamponade. However, the important point here is that postoperative
Reexploration for hemorrhage following coronary artery bypass grafting: bleeding, according data mentioned in [2], may carry a higher risk of major compli-
incidence and risk factors. Northern New England Cardiovascular Disease cations and early mortality, even when the patient does not require reoperation
Study Group. Arch Surg 1998;133:442–7. for bleeding.
[4] Moulton MJ, Creswell LL, Mackey ME, Cox JL, Rosenbloom M.
Reexploration for bleeding is a risk factor for adverse outcomes after References
cardiac operations. J Thorac Cardiovasc Surg 1996;111:1037–46.
[5] Ranucci M, Bozzetti G, Ditta A, Cotza M, Carboni G, Ballotta A. Surgical [1] Kristensen KL, Rauer LJ, Mortensen PE, Kjeldsen BJ. Reoperation for bleeding
reexploration after cardiac operations: why a worse outcome?. Ann in cardiac surgery. Interact CardioVasc Thorac Surg 2012;14:709–713.
Thorac Surg 2008;86:1557–62. [2] Christensen MC, Dziewior F, Kempel A, von Heymann C. Increased chest
[6] Gwozdziewicz M, Olsak P, Lonsky V. Re-operations for bleeding in cardiac tube drainage is independently associated with adverse outcome aftercardiac
surgery: treatment strategy. Biomed Pap Med Fac Univ Palacky Olomouc surgery. J Cardiothorac Vasc Anesth 2012;26:46–51.
Czech Repub 2008;152:159–62. [3] Vivacqua A, Koch CG, Yousuf AM, Nowicki ER, Houghtaling PL, Blackstone EH,
[7] The Danish Heart Register. www.dhreg.dk (8 February 2010, date last Sabik JF 3rd. Morbidity of bleeding after cardiac surgery: Is it blood transfu-
accessed). sion, reoperation for bleeding, or both? Ann Thorac Surg 2011:91;1780–90.
Cardiovascular Surgery
Frail Patients Are at Increased Risk for Mortality and

Prolonged Institutional Care After Cardiac Surgery
Dana H. Lee, BSc; Karen J. Buth, MSc; Billie-Jean Martin, MD;
Alexandra M. Yip, MSc, OT(C); Gregory M. Hirsch, MD, FRCSC
Background—Frailty is an emerging concept in medicine yet to be explored as a risk factor in cardiac surgery. Where
elderly patients are increasingly referred for cardiac surgery, the prevalence of a frail group among these is also on the
rise. We assessed frailty as a risk factor for adverse outcomes after cardiac surgery.
Methods and Results—Functional measures of frailty and clinical data were collected prospectively for all cardiac surgery
patients at a single center. Frailty was defined as any impairment in activities of daily living (Katz index), ambulation, or a
documented history of dementia. Of 3826 patients, 157 (4.1%) were frail. Frail patients were older, were more likely to be
female, and had risk factors for adverse surgical outcomes. By logistic regression, frailty was an independent predictor of
in-hospital mortality (odds ratio 1.8, 95% CI 1.1 to 3.0), as well as institutional discharge (odds ratio 6.3, 95% CI 4.2 to 9.4).
Frailty was an independent predictor of reduced midterm survival (hazard ratio 1.5, 95% CI 1.1 to 2.2).
Conclusions—Frailty is a risk for postoperative complications and an independent predictor of in-hospital mortality,
institutional discharge, and reduced midterm survival. Frailty screening improves risk assessment in cardiac surgery
patients and may identify a subgroup of patients who may benefit from innovative processes of care. (Circulation.
2010;121:973-978.)
Key Words: frail elderly 䡲 outcomes research 䡲 cardiac surgery
A dvances in modern medicine have led to people living

longer and healthier lives. The fastest growing demo-
graphic in North America is that of octogenarians. As a result,
tion, and mortality.12–14 Frailty has been less thoroughly inves-
tigated as a risk factor for patients undergoing procedural
interventions.15 It has not been investigated as a risk factor for
the number of people of advanced age presenting for cardiac cardiac surgical intervention. We proposed that preoperative
surgery is increasing.1 However, elderly patients generally pres- assessment of patients for frailty would result in more
ent with a larger burden of disease and higher potential periop- accurate risk assessment for cardiac surgery and may
erative morbidity and mortality despite advanced pharmacother- identify a challenging subset of patients who would benefit
apy and aggressive surgical management. In surgical from novel processes of care.
populations, age is an independent risk factor for postoperative In an effort to determine the degree of frailty in our patient
complications, including mortality and major adverse clinical population, in June 2004, we initiated data collection con-
events.2–5 Although age has been demonstrated to be a risk factor cerning the Katz index of activities of daily living (ADL), an
for both mortality and major morbidity in cardiac surgical internationally validated measure of dependency in elderly
outcomes, it has also been amply demonstrated in several studies patients.16 In addition, we collected data on independence in
that excellent outcomes can be achieved in selected populations ambulation and the presence of dementia. The goal of this
of elderly patients.6 –9 study was to determine the impact of these measures of frailty
on mortality and the need for prolonged postoperative insti-
Clinical Perspective on p 978 tutional care.
Chronological age does not always reflect biological age, and
elderly people have a range of biological status that varies from Methods
robust to frail.10,11 Frailty is an emerging concept in clinical
Patient Population
medicine, most extensively investigated in community-dwelling We identified all patients undergoing cardiac surgery at the Queen
geriatric populations, in which it has been demonstrated that frail Elizabeth II Health Sciences Centre in Halifax, Nova Scotia, Canada,
patients are predisposed to falls, hospitalization, institutionaliza- between June 2004 and December 2007. The Queen Elizabeth II Health
Received December 5, 2008; accepted December 14, 2009.

From the Division of Cardiac Surgery, Department of Surgery, Dalhousie University, Halifax, Nova Scotia, Canada.
Earlier versions of this article were presented at the Canadian Cardiovascular Congress 2008, Toronto, Ontario, Canada, October 25–29, 2008, and at
the American Heart Association Scientific Sessions 2008, New Orleans, Louisiana, November 8 –12, 2008.
Correspondence to Gregory M. Hirsch, MD, FRCSC, Queen Elizabeth II Health Sciences Centre, 1796 Summer St, Room 2006, Halifax, NS B3H 3A7,
Canada. E-mail ghirsch@dal.ca
© 2010 American Heart Association, Inc.
Circulation is available at http://circ.ahajournals.org DOI: 10.1161/CIRCULATIONAHA.108.841437
973
974 Circulation March 2, 2010
Sciences Centre is the sole cardiac surgical center serving the entire Outcomes
province of Nova Scotia as well as parts of the surrounding provinces. Primary outcomes in this study were in-hospital mortality, midterm
For patients undergoing more than 1 cardiac surgical procedure during all-cause mortality (including in-hospital and after discharge deaths),
the study period, only the index procedure was considered. and discharge to an institution (community hospital, rehabilitation or
restorative care facility, or skilled nursing facility, among patients
Ethics discharged alive). Secondary in-hospital outcomes included transfu-
The study protocol was approved by the local institutional research sion, low cardiac output syndrome, sepsis, pneumonia, permanent
ethics board. stroke, delirium, prolonged ventilation (ⱖ24 hours), postoperative
RF (serum creatinine ⬎176 ␮mol/L, not present preoperatively), and
prolonged postoperative stay (length of stay, ⱖ10 days).
Data Collection and Variable Selection
The Maritime Heart Center Cardiac Surgery Registry is a detailed
clinical database that has prospectively captured pre-, intra-, and
Statistical Analysis
Preoperative characteristics and postoperative in-hospital outcomes
postoperative information on all of the cardiac surgery patients at the
in frail versus non-frail patients were compared univariately using ␹2
Queen Elizabeth II Health Sciences Centre since March 1995, with
tests or Fisher exact tests for categorical variables and t tests or
ongoing auditing to ensure data accuracy. Collection of preoperative Wilcoxon rank sum tests for continuous variables. To satisfy the
measures of frailty within the registry began in June 2004. Trained assumption of linearity inherent in logistic regression and propor-
nurse practitioners or clinical associates administered questions to tional hazards modeling, age and BMI were transformed using a
patients and family members concerning the Katz index of ADL restricted cubic spline function with 3 knots placed at the 10th, 50th,
(independence in feeding, bathing, dressing, transferring, toileting, and 90th percentiles of each distribution.17
and urinary continence), as well as independence in ambulation (no Clinically relevant preoperative variables with ␹2 P⬍0.20 were
walking aid or assist required). Additionally, clear evidence of a included in the fully adjusted multivariable models. Colinearity was
previous diagnosis of dementia by a specialist physician was sought assessed through correlation matrices as Pearson rⱖ0.3; only 1
from the patient’s record. Patients with any deficiency in the Katz variable for each correlated pair was retained based on clinical
index of ADL, in ambulation, or with a previous diagnosis of importance.
dementia were defined as frail for the purposes of this study. The association of frailty with in-hospital mortality and institu-
Other preoperative characteristics included age at time of surgery, tional discharge was examined by logistic regression models fully
sex, diabetes, chronic obstructive pulmonary disease (COPD, defined as adjusted for relevant prognostic variables. Predictive accuracy of
pharmacological therapy for the treatment of chronic pulmonary com- each model was assessed by the receiver operating characteristic
promise, or forced expiratory volume in 1 second ⬍75% of predicted curve.18 A bootstrap procedure was performed on 200 subsamples,
value), congestive heart failure (CHF, defined as at least 3 of presence and the 95% CI of the receiver operating characteristic was obtained
of dyspnea, rales thought to represent pulmonary congestion, peripheral from the 2.5th and 97.5th percentiles of the bootstrap distribution.
edema, cardiomegaly on chest x-ray, or chest x-ray compatible with The impact of frailty on all-cause mortality was examined by a
interstitial edema), preoperative renal failure (RF, defined as serum fully adjusted Cox proportional hazards model. To assess the
creatinine ⬎176 ␮mol/L), cerebrovascular disease (CVD, defined as proportional hazards assumption, a time-dependent covariate was
any transient ischemic attack, reversible ischemic neurological deficit, created for each predictor as a function of survival time, and tests of
proportionality were applied. For variables that did not satisfy the
cerebrovascular accident or stroke, history of cerebrovascular surgery,
assumption of proportional hazards, time-dependent covariates were
or any carotid disease, including asymptomatic carotid disease), periph-
included in the models. Statistical analysis was performed using SAS
eral vascular disease (defined as history of aneurysm and/or occlusive software version 9.2 (SAS, Cary, North Carolina).
vascular disease with or without previous extracardiac vascular surgery), The authors had full access to the data and take full responsibility
left ventricular ejection fraction ⬍40%, urgency of surgery (elective for its integrity. All of the authors have read and agree to the article
[stable at home], in-house [requiring hospitalization until the time of as written.
surgery], urgent [requiring surgery within 24 hours to minimize further
clinical deterioration], or emergent [no delay in surgery]), body mass
index (BMI, kg/m2), complexity of procedure (isolated coronary artery
Results
bypass grafting versus other cardiac surgery), and reoperation (repeat Baseline Characteristics
cardiac surgery). Between June 2004 and December 2007, a total of 3826 patients
For analysis of midterm survival, clinical data from the Maritime had cardiac surgery at the Queen Elizabeth II Health Sciences
Heart Center Cardiac Surgery Registry were linked to the adminis-
Centre. Of these 3826 patients, 64 (1.7%) demonstrated a
trative Vital Statistics database for the province. This database is
created by the Department of Health and housed by the Population deficiency in the Katz index of ADL, 124 patients (3.2%) had
Health Research Unit in the Department of Community Health and some degree of dependence in ambulation, and 22 patients
Epidemiology at Dalhousie University. Links with administrative (0.6%) carried a previous diagnosis of dementia. The frail group
data were available for all of the cardiac surgery patients in this study comprised 157 (4.1%) patients having at least 1 of these deficits.
who resided in Nova Scotia at the time of surgery and were eligible Among the frail patients, 106 (67.5%) had a deficit in only 1 of
for provincial medical insurance.
these categories, 49 (31.2%) in 2 categories, and 2 (1.3%) in all
3 categories, In-hospital deaths in each frail category were 15
Process of Surgical Care
The great majority of cases were performed on pump (⬍1.4% of
(14.2%), 7 (14.3%), and 1 (50%), respectively.
coronary artery bypass grafting cases during the study period were The baseline preoperative characteristics are presented in
off-pump coronary artery bypass grafting). In the majority of cases, Table 1. Frail patients were older than nonfrail patients with a
cardiopulmonary bypass with modest hypothermia (32°C) was used median age of 71 years (interquartile range 61 to 78) versus 66
with a crystalloid prime. Cannulation was routinely performed by years (interquartile range 57 to 74; P⫽0.0001), although the age
distal ascending aorta/proximal arch and right atrium. Synthesis range was similar in both groups. Frail patients were more likely
oxygenators were used. Antegrade, cold blood, 4:1 ratio cardioplegia
to be female and presented with greater comorbidity burden,
was used in cases without significant aortic insufficiency or in the
case of aortic valve surgery by direct osteal delivery. Retrograde including higher rates of diabetes, COPD, CHF, RF, and CVD.
cardioplegia was added as clinically indicated (poor left ventricular Furthermore, frail patients presented with greater acuity and
function, aortic insufficiency). underwent more complex operative procedures than the nonfrail
Lee et al Frailty and Increased Risk After Cardiac Surgery 975
Table 1. Preoperative Patient Characteristics Table 2. Unadjusted In-Hospital Outcomes

Nonfrail Frail Nonfrail Frail
Variable, n (%) (n⫽3669) (n⫽157) P Variable, n (%) (n⫽3669) (n⫽157) P
Median age, y (IQR) 66 (57–74) 71 (61–78) 0.0001 Mortality 164 (4.5) 23 (14.7) 0.0001
Age range 15–94 18–88 ... Discharge location (patients
Median BMI, kg/m2 (IQR) 28 (25–32) 27 (24–32) 0.30 discharged alive)
Female sex 937 (25.5) 61 (38.9) 0.0002 Home 3189 (91.0) 69 (51.5) 0.0001
Diabetes 1212 (33.0) 65 (41.4) 0.03 Institution 316 (9.0) 65 (48.5) ...
COPD 563 (15.3) 43 (27.4) 0.0001 Blood transfusion 1239 (33.8) 97 (61.8) 0.0001
CHF 746 (20.3) 83 (52.9) 0.0001 Low cardiac output syndrome 373 (10.2) 34 (21.7) 0.0001
RF 212 (5.8) 19 (12.1) 0.0011 Sepsis 120 (3.3) 18 (11.5) 0.0001
PVD 604 (16.5) 30 (19.1) 0.38 Pneumonia 266 (7.3) 32 (20.4) 0.0001
CVD 510 (13.9) 38 (24.2) 0.0003 Permanent stroke 70 (1.9) 5 (3.2) 0.23
Urgency of surgery Delirium 335 (9.1) 23 (14.7) 0.020
Urgent/emergent 394 (10.7) 31 (19.8) 0.0001 Postoperative RF 361 (9.8) 36 (22.9) 0.0001
Inhouse 1514 (41.3) 87 (55.4) ... Prolonged ventilation 584 (15.9) 57 (36.3) 0.0001
Elective 1761 (48.0) 39 (24.8) ... Prolonged LOS 1075 (29.3) 87 (55.4) 0.0001
Procedure
Isolated CABG 2348 (64.0) 73 (46.5) 0.0001 [14.7%]; P⬍0.0001, Table 2). Other preoperative factors
Other 1321 (36.0) 84 (53.5) ... associated univariately with increased mortality included age,
Reoperation (vs first operation) 293 (8.0) 12 (7.6) 0.88 female sex, CHF, COPD, RF, CVD, peripheral vascular
disease, BMI, urgency of surgery, complex procedure, and
reoperation (all P⬍0.01, data not shown). By logistic regres-
patients (all P⬍0.05, Table 1). However, the 2 groups did not
differ significantly in other characteristics, as shown in Table 1. sion analysis, frailty was identified as an independent risk
Age and BMI were used as continuous variables in factor for in-hospital mortality (odds ratio [OR] 1.8, 95% CI
analysis. To satisfy the assumption of linearity in logistic 1.1 to 3.0; Table 3).
regression and proportional hazards modeling, a restricted
cubic spline transformation was applied to age, with knots at Institutional Discharge
49, 66, and 80 years, corresponding to the 10th, 50th, and Frailty was predictive of discharge to an institution for
90th percentiles of the distribution. Similarly, a restricted prolonged care in univariate analyses (nonfrail n⫽316
cubic spline transformation was applied to BMI, with knots at [9.0%], frail n⫽65 [48.5%]; P⬍0.0001; Table 2). Other
22.7, 28.1, and 35.7 kg/m2, corresponding to the 10th, 50th, preoperative factors having a univariate association with
and 90th percentiles of the distribution.
A total of 3254 cases (85%) were linked with the Vital Table 3. Risk-Adjusted Impact of Frailty on
Statistics administrative database maintained by the province, In-Hospital Mortality
with follow-up through March 2008. Median follow-up time Preoperative Characteristics OR 95% CI P
was 1.8 years (interquartile range, 0.9 to 2.8 years). Cases that Frail 1.8 1.1–3.0 0.03
did not link with the Vital Statistics database were 555
Age* 1.6 1.4–1.9 0.0001
patients not residing in Nova Scotia at the time of surgery and
17 who resided in the province but were ineligible for Female sex 1.2 0.8–1.6 0.36
provincial medical insurance. Compared with cases that COPD 1.3 0.9–1.8 0.22
linked, cases that did not link were similar with regard to RF 2.3 1.5–3.5 0.0002
frailty, age, sex, comorbidities, and urgency of surgery. CHF 2.2 1.5–3.0 0.0001
However, the cases that did not link were more likely to be PVD 1.5 1.1–2.2 0.03
complex cases or require reoperation. This reflects the refer- Urgency of surgery
ral pattern for out-of-province patients in our practice.
Urgent/emergent 5.1 3.3–8.1 0.0001
Unadjusted In-Hospital Outcomes In-house 1.6 1.1–2.4 0.03
Mortality, discharge to an institution, transfusion, low cardiac Elective 1.0 — —
output syndrome, sepsis, pneumonia, delirium, prolonged Procedure (other vs isolated CABG) 1.8 1.3–2.5 0.0008
ventilation, postoperative RF, and prolonged length of stay Reoperation (vs first operation) 1.7 1.1–2.7 0.03
were more prevalent among frail than nonfrail patients (all
*A restricted cubic spline transformation was applied to age, with knots at
P⬍0.05, Table 2). 49, 66, and 82 years; OR (odds ratio) represents the increase in risk for each
unit change in transformed age.
In-Hospital Mortality In addition to the variables listed in Table 3, the model was also adjusted for
In unadjusted analysis, frailty was associated with increased hypertension, CVD, and a restricted cubic spline transformation of BMI.
in-hospital mortality (nonfrail n⫽164 [4.5%], frail n⫽23 ROC 81%, 95% CI 78%– 85%.
Table 4. Risk-Adjusted Impact of Frailty on Prolonged Table 5. Risk-Adjusted Impact of Frailty on

Institutional Care Mid-Term Mortality
Preoperative Characteristics OR 95% CI P Preoperative Characteristics HR 95% CI P
Frail 6.3 4.2–9.4 0.0001 Frail 1.5 1.1–2.2 0.01
Age* 2.0 1.8–2.3 0.0001 Age* 1.5 1.4–1.7 0.0001
Female sex 1.6 1.2–2.0 0.0003 Female sex 1.2 0.9–1.5 0.19
Diabetes 1.4 1.1–1.8 0.01 Diabetes 1.3 1.1–1.7 0.02
COPD 1.5 1.1–2.0 0.01 COPD 2.5 1.9–3.4 0.0001
RF 1.6 1.1–2.4 0.02 RF 1.9 1.5–2.6 0.0001
CHF 1.4 1.1–1.8 0.02 CHF 1.8 1.4–2.2 0.0001
PVD 1.9 1.4–2.5 0.0001 PVD 1.8 1.4–2.3 0.0001
Urgency of surgery Urgency of surgery
Urgent/emergent 4.5 3.0–6.5 0.0001 Urgent/emergent 1.8 1.2–2.7 0.0012
In-house 2.6 2.0–3.5 0.0001 In-house 1.4 1.1–2.0 0.67
Elective 1.0 — — Elective 1.0 — —
Procedure (other vs isolated CABG) 2.0 1.5–2.6 0.0001 Procedure (other vs isolated CABG) 1.6 1.3–2.0 0.0001
*A restricted cubic spline transformation was applied to age, with knots at Reoperation 1.3 1.0–1.8 0.09
49, 66, and 82 years; OR represents the increase in risk for each unit change *A restricted cubic spline transformation was applied to age, with knots at
in transformed age. 49, 66, and 82 years; HR (hazard ratio) represents the increase in risk for each
In addition to the variables listed in Table 4, the model was also adjusted for unit change in transformed age.
hypertension, CVD, and a restricted cubic spline transformation of BMI. In addition to the variables listed in Table 5, the model was also adjusted for
ROC 83%, 95% CI 79%– 84%. CVD, a restricted cubic spline transformation of BMI, and the time-dependent
covariates associated with COPD and urgency of surgery.
prolonged institutional care were age, female sex, diabetes,
COPD, CHF, RF, CVD, peripheral vascular disease, BMI, outcomes. Furthermore, the effect of frailty was independent
urgency of surgery, and complex procedure (all P⬍0.02, data of age. Several aspects of this study merit comment.
not shown). In logistic regression, frailty was an independent There is no generally agreed-on definition of frailty be-
predictor of institutional discharge (OR 6.3, 95% CI 4.2 to 9.4; cause it is an emerging concept and has been operationally
Table 4). defined with a variety of scales. It is agreed that frailty is a
biological state or syndrome of decreased resistance to
Midterm Survival stressors that results from deterioration in multiple physio-
By univariate analysis, frailty was associated with reduced logical systems.12,19,20 Frailty includes a constellation of
midterm survival (nonfrail n⫽330 deaths [10.6%], frail n⫽41 clinical attributes, including loss of skeletal muscle mass, low
deaths [29.5%]; P⬍0.0001). Other preoperative factors asso- activity levels, and poor endurance.12 It has been demon-
ciated univariately with reduced midterm survival were age, strated that frailty and the onset of dependence in ADLs are
female sex, diabetes, COPD, CHF, RF, CVD, peripheral strongly associated.13 This informed our decision to use the
vascular disease, BMI, urgency of surgery, complex proce- Katz index of ADL16 as part of our frailty measure, a widely
dure, and reoperation (all P⬍0.01, data not shown). After accepted measure of overall dependency in elderly people.15
adjusting for other relevant risk factors by Cox proportional We also included information concerning independence in
hazards modeling, frailty was identified as an independent ambulation and a history of dementia, both measures having
risk factor for reduced midterm survival (hazard ratio 1.5, been associated with adverse clinical outcomes in a number
95% CI 1.1 to 2.2, Table 5). Adjusted survival at 2 years was of studies.21–23
84% (95% CI 79% to 91%) in frail and 89% (95% CI 85% to The majority of frailty studies have used a geriatric patient
93%) in nonfrail patients (Figure). cohort, whether community dwelling or institutionalized,
rather than patients undergoing surgical interventions. These
Frailty–Age Interaction studies have demonstrated that frailty is predictive of adverse
The interaction of frailty and age was examined and was not a health outcomes, including falls, hospitalization, institution-
significant predictor of any of the 3 primary outcomes. There- alization, and mortality.12–14 A study by Dasgupta et al
fore, this interaction term was not included in the final models. evaluated the outcomes of patients undergoing noncardiac
surgical procedures and concluded that frailty was indepen-
Discussion dently associated with postoperative complications, increased
Ours is the first study, to our knowledge, to demonstrate length of hospitalization, and inability to be discharged
frailty as a risk factor for adverse clinical outcomes after home.15 However, no studies take measures of frailty into
cardiac surgery, including mortality and prolonged institu- account in patients undergoing cardiac surgery. We hypoth-
tional care. Analyzing a consecutive cohort of all of the esized that frailty is associated with an increased risk of
patients undergoing cardiac surgery in a single institution, we mortality and prolonged institutional care in patients under-
found that frailty markedly increased the risk for these going cardiac surgery.
Lee et al Frailty and Increased Risk After Cardiac Surgery 977
Figure. Survival curves for frail and non-

frail patients after cardiac surgical inter-
vention, adjusted by Cox proportional
hazards modeling. Frailty is an indepen-
dent risk factor for reduced midterm sur-
vival (see Table 5).
There were significant differences in the preoperative These data will help to inform frail patients and their clinicians
characteristics between the frail and nonfrail populations. about the advisability of surgical intervention. Our findings raise
Overall, the frail patients were sicker (as demonstrated by an the question of whether it is justifiable to perform cardiac
increased burden of disease), older, and, interestingly, more surgical operations on frail patients in the first instance. Our own
likely to undergo complex procedures. This latter observation data indicate that even in our center, where cases are vetted in a
may reflect a referral bias in that clinicians withhold referring peer-reviewed, multidisciplinary “cath conference,” patients
frail elderly patients for coronary artery bypass grafting who qualify as frail are being operated on. Thus, it is important
unless they have an insurmountable burden of disease pre- for cardiac caregivers to engage frail patients in fully informed
cluding medical management. consent. Recent studies indicate that the majority of surgical
The impact of frailty on mortality is perhaps expected in consent discussions fail to meet criteria for fully informed
that by definition frail patients have diminished physiological consent.25,26 It may be that decision aids designed to convey
reserve and capacity to maintain homeostasis. Furthermore, at risks and benefits of cardiac surgical intervention in frail pa-
baseline the frail patients had a decreased ability to mobilize tients, particularly with regard to increased rates of prolonged
and ambulate, predisposing them to postoperative pneumo- institutional care, hospital mortality, and reduced long-term
nia, reintubation, and urinary tract infections related to survival, would afford a more fully informed, shared decision-
prolonged catheterization. All of these factors increase the making process. It is equally important that the providers of
cardiac surgical care work to mitigate the risks faced by frail
risk of a protracted recovery and prolonged postoperative
patients through modification of processes of care. Possibilities
institutional care.
include programs to increase mobility in frail patients in a safe,
In this study, we have demonstrated the utility of easily
supervised manner, and to address nutritional deficiencies prior
applied measures of frailty in predicting both mortality and
to intervention.
prolonged institutional care. Importantly, we considered even
Several limitations should be noted. First, this is a single-
1 impairment in any of these measures as evidence of frailty.
center, retrospective study. The relatively recent implemen-
Despite this low threshold, our operational definition of
tation of measures of frailty in our database limits our power
frailty had a marked impact on the outcomes of interest. Thus, to discriminate more fully about other major adverse out-
our definition of frailty may provide a relatively insensitive comes. Furthermore, prolonged institutional care, although a
measure of frailty, unable to distinguish more subtle risks for reasonable surrogate for potential failure to return to a fully
these adverse outcomes. Other more comprehensive measures independent life, is not conclusive in this regard. Information
of frailty, although more time-consuming and thus costly to concerning the late functional status of patients, beyond
administer, may be more appropriate for measuring preoper- discharge from the secondary institution, would more fully
ative frailty. Ideally, in addition to the measures we used, inform us about the potential for eventual recovery of
frailty assessment would include an assessment of nutrition, independence in these patients.
depression, social support, and cognitive impairment. Other In summary, we have demonstrated that preoperative
authors have used an instrument that measures these param- functional assessment of frailty increases our ability to
eters and demonstrated a similar relationship between frailty predict patients at greater risk for mortality, prolonged
and outcome in noncardiac surgery patients.14,24 Thus, more institutional care, and reduced midterm survival. It identifies
sensitive measures of preoperative frailty may allow us to a subset of patients who need to be more fully informed about
better discern more subtle risk for adverse outcomes. risks, thus improving informed, shared decision making about
Analysis of longitudinal data demonstrated that frail pa- the advisability of surgical intervention. It may also identify
tients experienced increased in-hospital mortality, increased a subset of patients who could benefit from altered processes
rates of institutional discharge, and reduced midterm survival. of care designed to offset the burden of frailty.
Sources of Funding 11. Mitnitski AB, Graham JE, Mogilner AJ, Rockwood K. Frailty, fitness and
late-life mortality in relation to chronological and biological age. BMC
This study was supported in part by a Canadian Institutes for Health
Geriatr. 2002;2:1.
Research Team Grant to the Canadian Cardiovascular Outcomes
12. Fried LP, Tangen CM, Walston J, Newman AB, Hirsch C, Gottdiener J,
Research Team. Seeman T, Tracy R, Kop WJ, Burke G, McBurnie MA, Cardiovascular
Health Study Collaborative Research Group. Frailty in older adults:
Disclosures evidence for a phenotype. J Gerontol A Biol Sci Med Sci. 2001;56:
None. M146 –M156.
13. Boyd CM, Xue QL, Simpson CF, Guralnik JM, Fried LP. Frailty, hos-
pitalization, and progression of disability in a cohort of disabled older
References women. Am J Med. 2005;118:1225–1231.
1. Rosamond W, Flegal K, Furie K, Go A, Greenlund K, Haase N, Hailpern 14. Rockwood K, Stadnyk K, MacKnight C, McDowell I, Hébert R, Hogan
SM, Ho M, Howard V, Kissela B, Kittner S, Lloyd-Jones D, McDermott DB. A brief clinical instrument to classify frailty in elderly people.
M, Meigs J, Moy C, Nichol G, O’Donnell C, Roger V, Sorlie P, Stein- Lancet. 1999;353:205–206.
berger J, Thom T, Wilson M, Hong Y, Am Heart Association Statistics 15. Dasgupta M, Rolfson DB, Stolee P, Borrie MJ, Speechley M. Frailty is
Committee and Stroke Statistics Subcommittee. Heart disease and stroke associated with postoperative complications in older adults with medical
statistics: 2008 update: a report from the Am Heart Association Statistics problems. Arch Gerontol Geriatr. 2009;48:78 – 83.
Committee and Stroke Statistics Subcommittee. Circulation. 2008;117: 16. Katz S, Ford AB, Moskowitz RQ, Jackson BA, Jaffe MW. Studies of
e25– e146. illness in the aged: the index of ADL: a standardized measure of bio-
2. Marcantonio ER, Goldman L, Mangione CM, Ludwig LE, Muraca B, logical and psychological function. JAMA. 1963;185:914 –919.
Haslauer CM, Donaldson MC, Whittemore AD, Sugarbaker DJ, Poss R. 17. Durrleman R, Simon R. Flexible regression models with cubic splines.
A clinical prediction rule for delirium after elective noncardiac surgery. Stat Med. 1989;8:551–561.
JAMA. 1994;271:134 –139. 18. Efron B. Nonparametric estimates of standard error: the jackknife, the
3. Arozullah AM, Khuri SF, Henderson WG, Daley J. Participants in the bootstrap and other methods. Biometrika. 1981;68:589 –599.
National Veterans Affairs Surgical Quality Improvement Program: devel- 19. Campbell AJ, Buchner DM. Unstable disability and the fluctuations of
opment and validation of a multifactorial risk index for predicting post- frailty. Age Ageing. 1997;26:315–318.
operative pneumonia after major noncardiac surgery. Ann Intern Med. 20. Hamerman D. Toward an understanding of frailty. Ann Intern Med.
2001;135:847– 857. 1999;130:945–950.
4. Alibhai SM, Leach M, Tomlinson G, Krahn MD, Fleshner N, Holowaty 21. De Rooij SEJA, Govers AC, Korevaar JC, Giesbers AW, Levi ML, de
E, Naglie G. 30-day mortality and major complications after radical Jonge E. Cognitive, functional, and quality-of-life outcomes of patients
prostatectomy: influence of age and comorbidity. J Natl Cancer Inst. aged 80 and older who survived at least 1 year after planned or unplanned
2005;97:1525–1532. surgery or medical intensive care treatment. J Am Geriatr Soc. 2008;56:
5. Polanczyk CA, Marcantonio E, Goldman L, Rohde LEP, Orav J, 816 – 822.
Mangione CM, Lee TH. Impact of age on perioperative complications 22. Lang PO, Heitz D, Hédelin G, Dramé M, Jovenin N, Ankri J, Somme D,
and length of stay in patients undergoing noncardiac surgery. Ann Intern Novella JL, Gauvain JB, Couturier P, Voisin T, De Wazière B, Gonthier
Med. 2001;134:637– 643. R, Jeandel C, Jolly D, Saint-Jean O, Blanchard F. Early markers of
6. Fruitman DS, MacDougall CE, Ross DB. Cardiac surgery in octoge- prolonged hospital stays in older people: a prospective, multicenter study
narians: can elderly patients benefit? Quality of life after cardiac surgery. of 908 inpatients in French acute hospitals. J Am Geriatr Soc. 2006;54:
Ann Thorac Surg. 1999;68:2129 –2135. 1031–1039.
7. Baskett R, Buth K, Ghali W, Norris C, Maas T, Maitland A, Ross D, 23. Purser JL, Kuchibhatla MN, Fillenbaum GG, Harding T, Peterson ED,
Forgie R, Hirsch G. Outcomes in octogenarians undergoing coronary Alexander KP. Identifying frailty in hospitalized older adults with sig-
artery bypass grafting. CMAJ. 2005;172:1183–1186. nificant coronary artery disease. J Am Geriatr Soc. 2006;54:1674 –1681.
8. Filsoufi F, Rahmanian PB, Castillo JG, Chikwe J, Silvay G, Adams DH. 24. Rolfson DB, Majumdar SR, Tsuyuki RT, Tahir A, Rockwood K. Validity
Results and predictors of early and late outcomes of coronary artery and reliability of the Edmonton Frail Scale. Age Ageing. 2006;35:
bypass graft surgery in octogenarians. J Cardiothorac Vasc Anesth. 2007; 526 –529.
21:784 –792. 25. Braddock CH III, Edwards KA, Hasenberg NM, Laidley TL, Levinson
9. Huber CH, Goeber V, Berdat P, Carrel T, Eckstein F. Benefits of cardiac W. Informed decision making in outpatient practice: time to get back to
surgery in octogenarians: a postoperative quality of life assessment. Eur basics. JAMA. 1999;282:2313–2320.
J Cardiothorac Surg. 2007;31:1099 –1105. 26. Brezis M, Israel S, Weinstein-Birenshtock A, Pogoda P, Sharon A,
10. Wilson JF. Frailty, and its dangerous effects, might be preventable. Ann Tauber R. Quality of informed consent for invasive procedures. Int J Qual
Intern Med. 2004;141:489 – 492. Health Care. 2008;20:352–357.
CLINICAL PERSPECTIVE
Frailty is an emerging concept in clinical medicine, most extensively investigated in community dwelling geriatric
populations where it has been demonstrated that frail patients are predisposed to falls, hospitalization, institutionalization,
and mortality. Frailty has been less thoroughly investigated as a risk factor for patients undergoing procedural
interventions. Its role as a risk factor for cardiac surgical intervention has not been investigated previously. We have
prospectively examined our cardiac patient population for frailty as measured by The Katz Index of Activities of Daily
Living, an internationally validated measure of dependency in elderly patients, as well as for deficits in independent
ambulation and for documented history of dementia. Patients having any defect in these measures were defined as frail.
We demonstrated that frailty was an independent predictor of in-hospital mortality, reduced medium term survival, as well
as discharge to institutional care rather than home. Although age was also a predictor of these outcomes, after adjusting
for age, frailty remained an independent predictor of these outcomes. Our data have implications for frail patients who have
cardiac disease amenable to surgical repair. As a result of our work, both patients and surgeons can be better informed about
potential adverse outcomes before arriving at a decision to go ahead with cardiac surgical intervention. Additionally, frail
patients would potentially benefit from altered approaches to care that could mitigate the risks these patients face, for
example, in improving mobilization and nutrition prior to a planned intervention.
Interactive CardioVascular and Thoracic Surgery Advance Access published December 12, 2012
Interactive CardioVascular and Thoracic Surgery (2012) 1–7 ORIGINAL ARTICLE
doi:10.1093/icvts/ivs501
CARDIAC GENERAL
Short-term independent mortality risk factors in patients
with cirrhosis undergoing cardiac surgery
Juan Carlos Lopez-Delgadoa,*, Francisco Estevea, Casimiro Javierreb, Xose Pereza, Herminia Torradoa,
Maria L. Carrioa, David Rodríguez-Castroa, Elisabet Farreroa and Josep Lluís Venturaa
a
Department of Intensive Care, Hospital Universitari de Bellvitge, L’Hospitalet de Llobregat, Barcelona, Spain
b
Department of Physiological Sciences, Universitat de Barcelona, IDIBELL, Barcelona, Spain
* Corresponding author. Department of Intensive Care, Hospital Universitari de Bellvitge, IDIBELL (Institut d’Investigació Biomèdica Bellvitge; Biomedical
Investigation Institute of Bellvitge), C/Feixa Llarga s/n. 08907 L’Hospitalet de Llobregat, Barcelona, Spain. Tel: +34-65-0506985; fax: +34-93-3319412;
e-mail: juancarloslopezde@hotmail.com ( J.C. Lopez-Delgado).
Received 17 June 2012; received in revised form 13 September 2012; accepted 18 September 2012
Abstract
OBJECTIVES: Cirrhosis represents a serious risk in patients undergoing cardiac surgery. Several preoperative factors identify cirrhotic
patients as high risk for cardiac surgery; however, a patient’s preoperative status may be modified by surgical intervention and, as yet,
no independent postoperative mortality risk factors have been identified in this setting. The objective of this study was to identify pre-
operative and postoperative mortality risk factors and the scores that are the best predictors of short-term risk.
METHODS: Fifty-eight consecutive cirrhotic patients requiring cardiac surgery between January 2004 and January 2009 were prospect-
ively studied at our institution. Forty-two (72%) patients were operated on for valve replacement, 9 (16%) for a CABG and 7 (12%) for
both (CABG and valve replacement). Thirty-four (58%) patients were classified as Child-Turcotte-Pugh class A, 21 (36%) as class B and 3
(5%) as class C. We evaluated the variables that are usually measured on admission and during the first 24 h of the postoperative
period together with potential operative predictors of outcome, such as cardiac surgery scores (Parsonnet, EuroSCORE), liver scores
(Child-Turcotte-Pugh, model for end-stage liver disease, United Kingdom end-stage liver disease score) and ICU scores (acute
physiology and chronic health evaluation II and III, simplified acute physiology score II and III, sequential organ failure assessment).
RESULTS: Seven patients (12%) died in-hospital, of whom 5 were Child-Turcotte-Pugh class B and 2 class C. Comparing survivors vs
non-survivors, univariate analysis revealed that variables associated with short-term outcome were international normalized ratio
(1.5 ± 0.24 vs 2.2 ± 0.11, P < 0.0001), presurgery platelet count (171 ± 87 vs 113 ± 52 l nl−1, P = 0.031), presurgery haemoglobin count
(11.8 ± 1.8 vs 10.2 ± 1.4 g dl−1, P = 0.021), total need for erythrocyte concentrates (2 ± 3.4 vs 8.5 ± 8 units, P < 0.0001), PaO2/FiO2 at 12 h
after ICU admission (327 ± 84 vs 257 ± 78, P = 0.04), initial central venous pressure (11 ± 3 vs 16 ± 4 mmHg, P = 0.02) and arterial blood
lactate concentration 24 h after admission (1.8 ± 0.5 vs 2.5 ± 1.3 mmol l−1, P = 0.019). Multivariate analysis identified initial central
venous pressure as the only independent factor associated with short-term outcome (P = 0.027). The receiver operating characteristic
curve showed that the model for end-stage Liver disease score had a better predictive value for short-term outcome than other scores
(AUC: 90.5 ± 4.4%; sensitivity: 85.7%; specificity: 83.7%), although simplified acute physiology score III was acceptable.
CONCLUSIONS: We conclude that central venous pressure could be a valuable predictor of short-term outcome in patients with cirrho-
sis undergoing cardiac surgery. The model for end-stage liver disease score is the best predictor of cirrhotic patients who are at high
risk for cardiac surgery. Sequential organ failure assessment and simplified acute physiology score III are also valuable predictors.
Keywords: Liver cirrhosis • Cardiac surgery • Short-term outcome • Mortality scores
INTRODUCTION system for cardiac operative risk evaluation (EuroSCORE), and

the cardiopulmonary bypass (CPB) time have all been identified
Liver cirrhosis (LC) is a major preoperative risk factor in general as potential predictors of mortality after cardiac surgery in those
surgery, especially in cardiac surgery, and the outcome is strong- patients [3]. However, evidence comes mainly from several small
ly related to the severity of liver disease in those patients [1]. studies; due to the lack of evidence from larger pools of data,
While in patients without advanced cirrhosis, cardiac surgery postoperative risk factors remain unidentified.
can be done safely, the risk of mortality is higher in patients At the same time, the option of liver transplantation as a treat-
with Child-Turcotte-Pugh (CTP) class B and C or with a model ment for patients with LC has produced an increase in survival
for end-stage liver disease score (MELD) >13 [1, 2]. Preoperative rate and the evaluation of concomitant cardiac diseases, which
total plasma bilirubin, cholinesterase concentrations, the European increase post-liver transplantation complications, is crucial for
2 J.C. Lopez-Delgado et al. / Interactive CardioVascular and Thoracic Surgery
preoperative risk assessment [4]. Thus, cardiac surgery is increas- of the CTP group was carried out with the Kaplan–Meier estima-
ing in those patients awaiting liver transplantation. tor for comparison with previous studies. Receiver operating
Consequently, identifying independent cardiac surgery post- characteristic (ROC) curve analyses were applied to determine
operative risk factors for these patients is an area of interest if optimal cut-off values of the different scores for short-term
we want to optimize post-surgical management and improve outcome and to further evaluate the predictive power between
outcome, especially post-surgical short-term outcome. In this them, considering the differences of the areas under the empir-
study, we also wanted to evaluate different score systems to ical ROC curves. A P-value of 0.05 was considered statistically
identify the best predictors of mortality. significant in all cases.
MATERIALS AND METHODS RESULTS

This study is a prospective single-centre observational study per- Forty-one patients (70.7%) were operated on for valve replace-
formed between January 2004 and January 2009. Data were ment, 10 (17.2%) for CABG, 6 of them off-pump, and 7 (12.1%)
included from 58 patients of 2825 (2.05%) consecutive patients were both CABG and valve replacement. Only 3 patients under-
with LC who underwent cardiac surgery in our hospital. The went urgent surgery for CABG and there were no mortalities. All
study was approved by the Institutional Ethics Committee. All of valve replacement operations were isolated: 34 (70.83%) were
the patients had previously granted permission for their medical mitral valve and 14 (29.17%) aortic. None of the patients had
records to be used for research purposes. previously undergone cardiac surgery.
LC was confirmed either by a liver biopsy or by clinical, la- Aetiologies for LC were predominantly infective hepatitis in
boratory and radiographical findings showing impaired hepatic 37.9% (hepatitis C, 31% (n = 18); hepatitis B, 6.9% (n = 4)),
function and portal hypertension. The CTP classification score alcohol-induced in 34.48% (n = 20) and both hepatitis C and
was calculated for each patient (CTP A: 7 points; CTP B: 8–10 alcohol-induced in 13.8% (n = 8). The other were cryptogenic
points; CTP C >11 points); 58.6% (n = 34) were classified as class cirrhosis/others (13.8% (n = 8)) and in 10 patients, it was
A, 36.2% (n = 21) as class B and 5.2% (n = 3) as class C. because of hepatocellular carcinoma.
We evaluated demographical data and comorbidities, treat- The preoperative characteristics of the patients, including
ment before surgery, bedside variables currently measured treatment before surgery, presented differences between groups
during the first 24 h of postoperative clinical care and complica- in platelet and haemoglobin counts (see Table 1). Three patients
tions/mortality during their admission. We calculated different were admitted previously at the cardiology department for acute
prognosis scores for each patient: cardiac surgery scores myocardial infarction and underwent urgent cardiac surgery
(Parsonnet and EuroSCORE), liver scores (CTP, MELD and during the same admission. None of them died and their post-
United Kingdom end-stage liver disease (UKELD)), ICU scores operative course did not differ from the other patients. Six
(sequential organ failure assessment (SOFA), acute physiology patients (10.3%) were treated with aspirin before going into
and chronic health evaluation (APACHE II and III) and simplified theatre. None of them died and there was no significant increase
acute physiology score (SAPS II and III). Finally, survival of the in terms of postoperative bleeding or the requirement for blood
different CTP groups was shown to allow a comparison with pre- products. Despite there being a considerable prevalence of pre-
vious studies. operative risk factors in these patients in terms of LC complica-
Cardiac surgical procedures were performed in all patients tions due to end-stage liver disease, there was no significant
using median sternotomy, standard cardiopulmonary bypass difference between survivors and non-survivors.
(CPB) with moderate hypothermia (34°C) and antegrade cardio- There were no differences in intraoperative data, such as CPB
plegia. A mean aortic pressure of >60 mmHg was maintained time and aortic cross-clamping (ACC), between groups (see
during surgery. For revascularization, we used the internal thor- Table 2). Differences in postoperative data were observed for ar-
acic artery (or bilateral if possible) and saphenous vein grafts. terial oxygen pressure of O2 and the fraction of inspired oxygen
Bypass graft flow was assessed for each graft by Doppler transit ratio (PaO2/FiO2), which was higher in survivors, while central
time flowmetry. Protamine was administered to reverse heparin venous pressure (CVP) on admission and 24 h after admission
according to standard practice. For CABG surgery, aspirin was and arterial lactate (AL) 24 h after admission were all lower in
routinely administered within the first 6 h after surgery following survivors. With regard to postoperative morbidities, patients who
local protocol. died required a large amount of erythrocyte concentrates during
admission, but there were no differences in terms of post-
surgical bleeding. They also required a longer period on mech-
Statistical analysis anical ventilation, and had a greater need for renal replacement
therapies (RRT) and an increased the need for vasopressors.
Statistical analysis was carried out using PASW statistics 13.0 The median ICU stay was 9 ± 10 days, with a difference
(SPSS, Inc., Chicago, IL, USA). Data are expressed as mean ± between groups (7.7 ± 1 in the survival group vs 13 ± 5 in the
standard deviation. We analyzed differences in data between non-survival group, P = 0.002). However, the median hospital
survivors and non-survivors. For the comparisons between the stay was 34 ± 20 days, and there were no differences between
two groups, the Mann–Whitney U-test was used or, when appro- groups (21 ± 3 vs 14.8 ± 5.6 days).
priate (after applying the one-sample Kolmogorov–Smirnov test), Mortality was 12.1% (n = 7); 5 patients were CTP class B and 2
the two-sample t-test was used. The χ² test was used to evaluate class C. The class C died of multi-systemic organ failure (MSOF),
categorical prognostic factors. A multivariate analysis was carried and the class B MSOF (3 patients) and septic shock (2).
out using Cox regression model to show independent risk mor- Short-term survival evaluated by Kaplan–Meier in Fig. 1 showed
tality factors for short-term outcome. Finally, the survival analysis differences between CTP class groups (log-rank test, P = 0.035).
J.C. Lopez-Delgado et al. / Interactive CardioVascular and Thoracic Surgery 3
Table 1: Demographics and baseline data
CARDIAC GENERAL
All patients Survivors Non-survivors P
(n = 58) (n = 51) (n = 7)
Sex (male) 69% (40) 70.6% (36) 57.1% (4) 0.66

Age (years) 64.9 ± 11.6 64.6 ± 9.6 66.9 ± 10.3 0.92
Body mass index (kg m−2) 27 ± 4.2 27.6 ± 4.6 26.6 ± 4.2 0.54
Hypertension 56.9% (33) 54.9% (28) 71.4% (5) 0.68
Diabetes mellitus 32.8% (19) 33.3% (17) 28.6% (2) 0.99
Dyslipidaemia 34.5% (20) 33.3% (17) 42.9% (3) 0.68
Chronic renal insufficiency 8.6% (5) 7.8% (4) 14.3% (1) 0.12
Renal failure (on dialysis) 19% (11) 19.60% (10) 14.3% (1) 0.60
Creatinine before surgery (mmol l−1) 114.4 ± 100.8 106.4 ± 93.7 170.3 ± 136.3 0.15
Previous stroke 12.1% (7) 12.1% (7) 0% 0.23
Chronic obstructive pulmonary disease 17.2% (10) 17.6% (9) 14.3% (1) 0.85
Active smokers 19% (11) 19.6% (10) 14.3% (1) 0.64
Active alcohol consumption 3.4% (2) 3.9% (2) 0% 0.84
Previous atrial fibrillation 31% (18) 33.3% (17) 14.3% (1) 0.78
Previous myocardial infarction 12.1% (7) 11.8% (6) 14.3% (1) 0.53
NYHA class III-IV 34.5% (20) 35.3% (18) 28.6% (2) 0.58
On B-blockers 39.7% (23) 41.2% (21) 28.6% (2) 0.69
On statins 25.9% (15) 25.50% (13) 28.6% (2) 0.92
Ascites (moderate to severe) 69% (40) 70.6% (36) 57.1% (4) 0.45
Oesophageal varices 31% (18) 25.5% (13) 71.4% (5) 0.26
Variceal bleeding 17.2% (10) 17.6% (9) 14.3% (1) 0.14
Encephalopathy 34.5% (20) 33.3%(17) 42.9% (3) 0.32
Hypertrophic cardiomyopathy 31% (18) 31.4% (16) 28.6% (2) 0.68
Dilated cardiomyopathy 27.6% (16) 27.5% (14) 28.6% (2) 0.91
Left ventricular ejection fraction (%) 60.3 ± 11.2 59.3 ± 11.7 62.6 ± 10.1 0.71
Pulmonary arterial pressure (mmHg) 48.7 ± 15.4 48.6 ± 15.6 49.4 ± 14.7 0.58
Haemoglobin before surgery (g dl−1) 11.67 ± 1.82 11.8 ± 1.8 10.2 ± 1.05 0.02
Platelet count before surgery (1 nl−1) 164 ± 85 171 ± 87 113 ± 52 0.03
International normalized ratio before surgery 1.5 ± 0.83 1.45 ± 0.15 1.85 ± 0.76 0.18
NYHA: New York Heart Association classification; Results are expressed as mean ± standard deviation or percentage.
Some scores revealed significant differences between groups: In view of the complexity of the procedure, the postoperative
only SAPS II and III and SOFA showed a significant predictive morbidity and mortality rates reported in the literature are con-
power similar to that of UKELD and CTP. However, the other siderably higher for cirrhotic patients undergoing cardiac
ICU scores and cardiac surgery scores were not as useful surgery. [1]. The mortality risk in CTP class B patients is around
(Table 3). In order to compare differences between potential 32.2% and increases to 66.6% in CTP class C patients [2]; even
preoperative (liver and cardiac surgery scores) and postoperative when there is a minimal degree of impaired liver function in
(ICU scores) predictions, predictors of outcome for short-term combination with elective surgery, the incidence of complica-
survival were analysed using the ROC curve. The MELD score tions significantly increases [5]. Careful patient selection is critical
was the most predictive for in-hospital mortality. The optimal to improve surgical outcome in patients with cirrhosis [6];
cut-off level for the MELD score was 18.5, with a sensitivity of however, there is a lack of factors that can be used to identify
85.7% and a specificity of 83.7% (Fig. 2). the mortality risk in those patients, especially after surgery. The
To evaluate preoperative and postoperative predictors of lower incidence of comorbidities, the low number of urgent pro-
death for all patients, a multivariate analysis was conducted (See cedures and the low mortality rate found highlight the import-
Table 4). We included those univariate factors that showed sig- ance of our aim to select and prepare those patients for surgery
nificant differences between groups in a Cox regression model. carefully. Despite the differences in haemoglobin and platelets,
After risk adjustment, the multivariate analysis revealed initial the groups of survivors and non-survivors were comparable in
CVP as the only independent factor associated with short-term almost all presurgery risk factors except the grade of liver
outcome. disease. The major need for erythrocyte concentrates and RRT
needs in non-survivors can be explained by initial presurgical
lower haemoglobin, post-surgical INR differences and larger ICU
DISCUSSION admission and presence of MSOF as a cause of mortality, re-
spectively. In any case, the risk of mortality increases with the
The most important finding of the current study was that in deterioration of liver function [1–6].
terms of predicting short-term mortality, both the CVP and the In this scenario, INR progressively worsens during cirrhosis, also
SAPS III and SOFA postoperative scores proved effective. We also reflecting the current status of end-stage liver disease [7]. The re-
confirm that the MELD score is the most effective predictor for plenishment of vitamin K-dependent factors beyond a normal
the short-term outcome of these patients and that the CTP is a INR has not proven its efficacy; however, individualized heparin
valuable score. and protamine dosing, antifibrinolytic drug administration,
Table 2: Intraoperative and postoperative data

(n = 58) (n = 51) (n = 7)
Intraoperative data
Isolated CABG 15.5% (9) 15.7% (8) 14.3% (1) 0.95
Isolated valve surgery 72.4% (42) 72.50% (37) 71.4% (5) 0.97
CABG + valve surgery 12.1% (7) 11.76% (6) 14.3% (1) 0.78
Fluid balance during surgery (ml) 1325 ± 850 1250 ± 980 1350 ± 785 0.58
Aortic cross-clamping time (min) 72 ± 44 74 ± 41 69 ± 50 0.85
Cardiopulmonary bypass time (min) 107 ± 37 106 ± 48 108 ± 53 0.35
Postoperative data and major postoperative complications
Ventilation time (days) 5.3 ± 10.2 3.16 ± 7.7 21 ± 12 0.01
PaO2/FiO2 on admission 287 ± 95 293 ± 93 245 ± 110 0.28
PaO2/FiO2 12 h after admission 318 ± 86 327 ± 84 257 ± 78 0.04
PaO2/FiO2 24 h after admission 307 ± 75 315 ± 70 253 ± 96 0.23
MAP on admission (mmHg) 83 ± 15 85 ± 15 74 ± 18 0.72
MAP 24 h after admission (mmHg) 80 ± 10 80 ± 9 75 ± 11 0.51
CVP on admission (mmHg) 12 ± 3.6 11.4 ± 3 16.5 ± 4.4 0.02
CVP 24 h after admission (mmHg) 12.5 ± 3.6 12 ± 2.8 16.3 ± 6 0.002
Need of vasoactive drugs (h) 165 ± 197 112 ± 109 490 ± 304 0.016
Low cardiac output syndrome 31% (18) 34% (17) 14.3% (1) 0.25
Perioperative myocardial infarction 7.1% (4) 6.1% (3) 14.3% (1) 0.18
Arterial lactate on admission (mmol l−1) 2.6 ± 1.4 2.45 ± 1.3 3.6 ± 1.5 0.22
Arterial lactate 24 h after admission (mmol l−1) 1.9 ± 0.7 1.8 ± 0.5 2.5 ± 1.3 0.02
Creatinine 24 h after surgery (mmol l−1) 129 ± 108 118 ± 101 207 ± 138 0.15
Urine output first 24 h (ml) 1860 ± 650 1920 ± 570 1444 ± 1066 0.28
Need for renal replacement therapy 8.9% (5) 2% (1) 57.1% (4) <0.0001
Albumin (g l−1) 27 ± 4 27.9 ± 4 27.8 ± 4.5 0.97
International normalized ratio on admission 1.8 ± 0.32 1.5 ± 0.24 2.2 ± 0.11 <0.0001
Drainage loss first 12 h (ml) 464 ± 308 446 ± 299 595 ± 369 0.34
Major bleeding 1.7% (1) 2% (1) 0% 0.85
Re-exploration 19% (11) 21.6% (11) 0% 0.15
Erythrocyte concentrates (units) 3 ± 4.6 2 ± 30.4 8.5 ± 8 <0.0001
CABG: coronary artery bypass graft; PaO2/FiO2: arterial partial pressure of O2 and fraction of inspired oxygen ratio; MAP: mean arterial pressure; CVP:
central venous pressure; Results are expressed as mean ± standard deviation or percentage.
Table 3: Evaluation scores for risk assessment

(n = 58) (n = 51) (n = 7)
SAPS II 25.2 ± 10.4 24 ± 9.4 33.7 ± 14 0.02

SAPS III 45.9 ± 10.8 44.7 ± 10.4 54.7 ± 10.4 0.045
APACHE II 13.9 ± 4.4 13.5 ± 4.1 16.8 ± 6 0.19
APACHE III 56.6 ± 18 55.2 ± 17.7 66.7 ± 19 0.17
SOFA 5.41 ± 2.72 6.6 ± 2.7 9.4 ± 1.8 0.005
EuroSCORE 6.48 ± 3 6.2 ± 2.9 8.8 ± 3.7 0.12
Parsonnet 9.43 ± 6.42 9.2 ± 6.4 11.4 ± 6.8 0.43
score
MELD 16 ± 5.4 15 ± 4.57 23 ± 5.4 0.005
UKELD 49.8 ± 4 49.6 ± 4 52.6 ± 3.3 0.044
CTP class A 58.6% 66.7% 0% <0.0001
(n = 34) (n = 34)
CTP class B 36.2% 31.4% 71.4% (n = 5) <0.0001
(n = 21) (n = 16)
CTP class C 5.2% (n = 3) 2% (n = 1) 28.6% (n = 2) 0.045
Figure 1: Short-term survival rate according to Child-Turcotte-Pugh score. SAPS: simplified acute physiology score; APACHE: acute physiology
and chronic health evaluation; SOFA: sequential organ failure
minimization of blood loss and dilution, and minimal CPB time assessment; EuroSCORE: European system for cardiac operative risk
evaluation; MELD: model for end-stage liver disease score; UKELD:
could still potentially help achieve surgical homeostasis [8]. All United Kingdom end-stage liver disease; CTP: Child-Turcotte-Pugh.
these efforts are reflected in our results, in that drainage loss was Results are expressed as mean ± standard deviation or percentage.
similar between the groups despite postoperative INR differences.
Arterial partial pressure of O2 and fraction of inspired oxygen

ratio (PaO2/FiO2) is a new marker for outcome in some types of
cardiac surgery [11]. Hypoxaemia depicted by low PaO2/FiO2 is
CARDIAC GENERAL
common after CPB, and is associated with different variables,
which are preoperative factors (age, obesity, chest X-ray with al-
veolar oedema 1 h after surgery, decreased baseline PaO2/FiO2,
previous myocardial infarction), operative factors (emergency
surgery, prolonged CPB) and postoperative factors (low cardiac
output syndrome (LCOS), renal failure, persistent hypothermia 2–
6 h after surgery, requirement for re-exploration). A lower PaO2/
FiO2 ratio correlated significantly with the time required to carry
out extubation and also to lung injury. However, in these
patients, it had minimal effect on the postoperative clinical
course [12]. Although PaO2/FiO2 12 h after admission was lower
in non-survivors, it did not have an independent significant
impact on the outcome of surgery.
Central venous pressure (CVP) is used almost universally to
guide fluid therapy in hospitalized patients. Some authors argue
that there is a very poor relationship between CVP and blood
volume as well as the inability to predict the haemodynamic
response to a fluid challenge, being a good indicator of blood
volume only at the extreme values [13]. Nevertheless, the condi-
tions that influence CVP are well known, and as such, CVP
remains a useful tool for evaluating haemodynamic status if it is
performed under controlled conditions. CVP has the great ad-
Figure 2: ROC curve for MELD. Comparison of AUC for MELD, SAPS III and vantage of being able to be measured at the patient’s bedside
SOFA scores. AUC: area under curve; ROC: receiver operating characteristic without the need of invasive methods [14]. Dynamic evaluation
curve; SAPS: simplified acute physiology score; SOFA: sequential organ failure
assessment; MELD: model for end-stage liver disease score; NS: non-
of CVP could be a reliable predictor of fluid responsiveness in
statistically significant. Results are expressed as mean ± standard deviation or patients under mechanical ventilation, similar to the variation of
percentage. arterial pulse pressure after cardiac surgery [15]. The proper use
of CVP requires a good understanding of the waveform because
higher values and CVP tracing are concordant with rhythm
disorders, tricuspid regurgitation, cardiac tamponade, cardiac
Table 4: Multivariate analysis-dependent variable deceased restriction and decreased thoracic compliance [16]. Limitations of
during admission CVP as a surrogate variable of preload are caused by the influ-
ence of intrathoracic and intra-abdominal pressures. However,
Hazards ratio (95% CI) P these limitations do not impair the importance of CVP as the
downstream pressure of the systemic venous system [15, 16].
Age 0.99 (0.94–1.036) 0.69
Platelets before surgery 0.96 (0.79–1.164) 0.68 We found CVP on admission to be the only independent factor
Haemoglobin before surgery 1.13 (0.65–1.97) 0.66 for short-term outcome in the multivariate analysis. We hy-
INR after surgery 0.65 (0.17–2.51) 0.53 pothesize that CVP could be a surrogate marker of underscored
CVP on admission 0.88 (0.78–0.98) 0.027 right ventricular failure, which can ultimately explain the higher
SOFA score 1.02 (0.86–1.195) 0.82
AL 24 h after admission 0.81 (0.60–1.094) 0.17
mortality, but we cannot confirm our suspicions [17]. However,
PaO2/FiO2 12 h 1.00 (0.99–1.004) 0.91 non-survivors did not receive larger amount of fluids in the
MELD score 0.96 (0.87–1.068) 0.48 operating theatre and did not have higher incidences of low
cardiac output syndrome, which could have biased the CVP
PaO2/FiO2: arterial partial pressure of O2 and fraction of inspired measurement.
oxygen ratio; AL: arterial lactate; INR: international normalized ratio; Although EuroSCORE is widely accepted in Europe as a valu-
CVP: central venous pressure; SOFA: sequential organ failure
able score in cardiac surgery, in some populations, it does not
assessment; MELD: model for end-stage liver disease score.
have acceptable discriminatory ability. The development of local
mortality risk scores corresponding to local epidemiological
characteristics may improve the prediction of outcome [18]. In
addition, it does not take into account surgical prognosis factors
Hyperlactataemia in the ICU, which is caused mainly by such as CPB time, and there is a lack of postoperative factors to
shock, is associated with increased mortality and is more fre- determine short-term mortality [19]. Furthermore, the Parsonnet
quent when respiratory and/or renal failures are/is present [9]. It score does not consider specific liver variables. However, some
predicts postoperative mortality after cardiac surgery with a authors suggest that it can be used to predict 3-month mortality,
maximum lactate threshold of ≥4.4 mmol l−1 in the first 10 h prolonged length of stay and specific postoperative complica-
after operation [10]. Arterial lactate tends to be higher in non- tions such as renal failure, sepsis and respiratory failure in the
survivors, though it could be a reflection of a presurgery poorer whole context of cardiac surgery [20]. Because mortality in cir-
liver function or an exacerbation of liver dysfunction in the rhotic patients undergoing cardiac surgery is associated with
setting of CPB. liver function, liver scores such as the MELD or CTP score are
associated with outcome [1–3]. Our results confirm that the have contributed with their efforts in the care of the patients
MELD score most reliably identifies cirrhotic patients at high risk reported in this manuscript.
for cardiac surgery, with better results than in previous studies
[1]. In our study, the MELD values are higher than in previous Conflict of interest: none declared.
studies, which is likely due to the high number of patients await-
ing liver transplantations. With regard to CTP class scores, mor-
tality was higher in postoperative cardiac surgery in patients with
a CTP score of class C [1–3, 6]. With a lack of a large data series REFERENCES
in previous research or a significant number of CTP class C
[1] Thielmann M, Mechmet A, Neuhäuser M, Wendt D, Tossios P, Canbay A
patients described in the literature, there is no basis for compari-
et al. Risk prediction and outcomes in patients with liver cirrhosis under-
son. The UKELD score can be used as a local score for end-stage going open-heart surgery. Eur J Cardiothorac Surg 2010;38:592–9.
liver disease, but unlike the MELD, it has never been evaluated [2] Modi A, Vohra HA, Barlow CW. Do patients with liver cirrhosis undergo-
in cardiac surgery. It evaluates sodium as well as INR, creatinine ing cardiac surgery have acceptable outcomes? Interact CardioVasc
and bilirubin, identifying cirrhotic patients with the poorest Thorac Surg 2010;11:630–4.
[3] An Y, Xiao YB, Zhong QJ. Open-heart surgery in patients with liver cir-
quality of life and the highest complication rates [21]. The results rhosis: indications, risk factors, and clinical outcomes. Eur Surg Res 2007;
for UKELD were statistically significant in the univariate analysis, 39:67–74.
though the ROC analysis raised doubts about its clinical rele- [4] Lima B, Nowicki ER, Miller CM, Hashimoto K, Smedira NG,
vance. ICU scores such as SOFA have been previously evaluated Gonzalez-Stawinski GV. Outcomes of liver transplantation and elective
cardiac surgical procedures. Ann Thorac Surg 2011;92:1580–4.
in cardiac surgery for the same purpose [22]. We also evaluated
[5] Bizouarn P, Ausseur A, Desseigne P, Le Teurnier Y, Nougarede B, Train M
other ICU scores such as SAPS and APACHE. SAPS scores pro- et al. Early and late outcome after elective cardiac surgery in patients
vided an estimate of the risk of death without having to specify with cirrhosis. Ann Thorac Surg 1999;67:1334–8.
a primary diagnosis, including liver failure and cardiac insuffi- [6] Filsoufi F, Salzberg SP, Rahmanian PB, Schiano TD, Elsiesy H, Squire A
ciency grade [23]. Furthermore, higher SAPS scores have been et al. Early and late outcome of cardiac surgery in patients with liver cir-
rhosis. Liver Transpl 2007;13:990–5.
associated with a poor quality of life, with the worst outcome [7] Hayashida N, Aoyagi S. Cardiac operations in cirrhotic patients. Ann
occurring both before and after general surgery [24]; additionally, Thorac Cardiovasc Surg 2004;10:140–7.
a higher mortality rate was found in elderly patients (>70 years) [8] Milas BL, Jobes DR, Gorman RC. Management of bleeding and coagulo-
who required dialysis after cardiac surgery [25]. In our series, pathy after heart surgery. Semin Thorac Cardiovasc Surg 2000;12:326–36.
[9] Juneja D, Singh O, Dang R. Admission hyperlactatemia: causes, inci-
SAPS III provided an acceptable level of sensitivity and specifi-
dence, and impact on outcome of patients admitted in a general
city, comparable with MELD results of other series [1]. APACHE medical intensive care unit. J Crit Care 2011;26:316–20.
scores were not found to be valuable tools. [10] Kogan A, Preisman S, Bar A, Sternik L, Lavee J, Malachy A et al. The
Our study presents certain limitations. The most important are impact of hyperlactatemia on postoperative outcome after adult cardiac
that it was a single-centre observational study. Results should be surgery. J Anesth 2012;26:174–8.
[11] Kurabayashi M, Okishige K, Azegami K, Ueshima D, Sugiyama K, Shimura
viewed cautiously due to the low number of patients and events. T et al. Reduction of the PaO2/FiO2 ratio in acute aortic dissection—rela-
However, we have shown a larger number of patients than any tionship between the extent of dissection and inflammation. Circ J 2010;
other study of this kind to date, and observed a low mortality 74:2066–73.
rate despite the level of end-stage liver disease. [12] Weiss YG, Merin G, Koganov E, Ribo A, Oppenheim-Eden A, Medalion B
et al. Postcardiopulmonary bypass hypoxemia: a prospective study on in-
We conclude that cardiac surgery can be performed safely in
cidence, risk factors, and clinical significance. J Cardiothorac Vasc Anesth
CTP class A and in some class B patients. Regarding CTP class C 2000;14:506–13.
patients, due to the higher mortality in these patients, we think [13] Marik PE, Baram M, Vahid B. Does central venous pressure predict fluid
that liver function should be optimized prior to cardiac surgery, responsiveness? A systematic review of the literature and the tale of
perhaps even performing liver transplantation. Indeed, synchron- seven mares. Chest 2008;134:172–8.
[14] Gelman S. Venous function and central venous pressure. Anesthesiology
ous surgery has modestly improved survival in some patients 2008;108:735–48.
with cirrhosis when cardiac surgery is needed [4]. We recom- [15] Westphal GA, Silva E, Caldeira Filho M, Roman Gonçalves AR,
mend proper preoperative selection of patients and apply Poli-de-Figueiredo LF. Variation in amplitude of central venous pressure
careful operative and postoperative management, especially in curve induced by respiration is a useful tool to reveal fluid responsive-
ness in postcardiac surgery patients. Shock 2006;26:140–5.
terms of fluid balance, in order to increase the short-term sur-
[16] Magder S. Central venous pressure monitoring. Curr Opin Crit Care
vival rate. A higher CVP at ICU admission may make physicians 2006;12:219–27.
aware of a patient’s prognosis, but its efficacy as a valuable pre- [17] Richter HP, Petersen C, Goetz AE, Reuter DA, Kubitz JC. Detection of
dictor of short-term outcome must be shown in future studies. right ventricular insufficiency and guidance of volume therapy are facili-
MELD score and postoperative ICU scores such as SAPS III and tated by simultaneous monitoring of static and functional preload para-
meters. J Cardiothorac Vasc Anesth 2011;25:1051–5.
SOFA can be used to predict short-term outcome in those [18] Sadeghi MM, Arasteh M, Gharipour M, Nilfroush P, Shamsolketabi H,
patients. In our opinion, in the setting of end-stage liver disease Etesampour A et al. Evaluation of accuracy of Euroscore risk model in
and cardiac surgery, postoperative evaluation is as important as prediction of perioperative mortality after coronary bypass graft surgery
preoperative evaluation in terms of predicting short-term in Isfahan. J Res Med Sci 2011;16:787–92.
[19] Roques F, Nashef SA, Michel P, Gauducheau E, de Vincentiis C, Baudet E
outcome.
et al. Risk factors and outcome in European cardiac surgery: analysis of
the EuroSCORE multinational database of 19030 patients. Eur J
Cardiothorac Surg 1999;15:816–22.
ACKNOWLEDGEMENTS [20] Toumpoulis IK, Anagnostopoulos CE, Swistel DG, DeRose JJ Jr. Does
EuroSCORE predict length of stay and specific postoperative complica-
tions after cardiac surgery? Eur J Cardiothorac Surg 2005;27:128–33.
We thank Toral and Ortiz of the Department of Cardiothoracic [21] Doerr F, Badreldin AM, Heldwein MB, Bossert T, Richter M, Lehmann T
Surgery who both supported and corrected this manuscript, es- et al. A comparative study of four intensive care outcome prediction
pecially regarding surgical issues and also the ICU nurses who models in cardiac surgery patients. J Cardiothorac Surg 2011;6:21.
[22] Neuberger J, Gimson A, Davies M, Akyol M, O’Grady J, Burroughs A [24] Abelha FJ, Santos CC, Barros H. Quality of life before surgical ICU admis-
et al. Selection of patients for liver transplantation and allocation of sion. BMC Surg 2007;7:23.
donated livers in the UK. Gut 2008;57:252–7. [25] Van Den Noortgate N, Mouton V, Lamot C, Van Nooten G, Dhondt A,
CARDIAC GENERAL
[23] Le Gall JR, Lemeshow S, Saulnier F. A new simplified Acute Physiology Vanholder R et al. Outcome in a post-cardiac surgery population with
Score (SAPS II) based on a European/North American multicenter study. acute renal failure requiring dialysis: does age make a difference?
JAMA 1993;270:2957–63. Nephrol Dial Transplant 2003;18:732–6.
Coronary artery bypass grafting in acute

myocardial infarction: analysis of predictors of in-
hospital mortality
Cirurgia de revascularização miocárdica na fase aguda do infarto: análise dos fatores preditores de
mortalidade intra-hospitalar
Omar Asdrúbal Vilca Mejía1, Luiz A Ferreira Lisboa2, Marcos Gradim Tiveron3, José Augusto Duncan
Santiago4, Rafael Angelo Tineli5, Luis Alberto Oliveira Dallan6, Fabio Biscegli Jatene7, Noedir Antonio
Groppo Stolf8
DOI: 10.5935/1678-9741.20120011 RBCCV 44205-1352
Abstract Univariate and multivariate analysis for the outcome of in-

Objective: Coronary artery bypass grafting (CABG) hospital mortality were performed.
during the acute phase of infarction (AMI) is associated Results: The mean time between AMI and CABG was 3.8
with increased operative risk. The aim of this study was to ± 3 days. The overall mortality was 19%. In the multivariate
determine predictors of in-hospital mortality in patients analysis: age > 65 years OR [16.5 (CI 1.8 to 152), P= 0.013];
undergoing CABG in AMI. CPB > 108 minutes [OR 40 (CI 2.7 to 578), P= 0.007],
Methods: During three years, all patients undergoing creatinine> 2 mg/dl [OR 35.5 (CI 1.7 to 740), P= 0.021] and
CABG in AMI were retrospectively analyzed of the systolic pulmonary pressure > 60 mmHg [OR 31 (CI 1.6 to
institutional database. Sixty variables per patient were 591), P= 0.022] were predictors of in-hospital mortality.
evaluated: 49 preoperative variables from the 2000 Conclusion: Conventional preoperative variables such
Bernstein-Parsonnet and EuroSCORE models, 4 as age > 65 years, creatinine > 2 mg/dl and systolic
preoperative variables not considered in these models (time pulmonary pressure > 60 mmHg were predictive of in-
between AMI and CABG, maximum CKMB, Troponin hospital mortality in patients underwent CABG in AMI.
maximum and ST-segment elevation) and 7 intraoperative
variables [(cardiopulmonary bypass (CPB), CPB time, type
of cardioplegia, endarterectomy, number of grafts, use of Descriptors: Risk factors. Myocardial infarction.
internal thoracic artery and complete revascularization]. Revascularização miocárdica. Coronary artery bypass.
1. Specialist in Aortic Surgery; Medical Residence in Cardiovascular Director of the Thoracic Surgery Service at InCor-HCFMUSP,
Surgery at Heart Institution, Clinics Hospital of the Faculty of São Paulo, SP, Brazil.
Medicine, University of São Paulo (InCor-HCFMUSP), São Paulo, 8. Titular Professor of the Cardiovascular Surgery Discipline at
SP, Brazil. FMUSP, Director of the Surgical Division at InCor-HCFMUSP,
2. Full Professor at Faculty of Medicine, University of São Paulo São Paulo, SP, Brazil.
(FMUSP); Assistant Physician at Surgical Coronaropathy Unit This study was carried out at Heart Institute, Clinics Hospital of the
at InCor-HCFMUSP, São Paulo, SP, Brazil. Faculty of Medicine, University of São Paulo, SP, Brazil.
3. Specialist in Cardiovascular Surgery at FMUSP; Preceptor
Physician of Cardiovascular Surgery Residence at FMUSP, São Correspondence address:
Paulo, SP, Brazil. Omar Asdrúbal Vilca Mejía
4. Specialist in General Surgery; Resident in Cardiovascular Surgery Av. Dr. Enéas Carvalho de Aguiar, 44 – Cerqueira César – São Paulo,
at InCor-HCFMUSP, São Paulo, SP, Brazil. SP, Brasil – CEP: 05403-000
5. Specialist in Cardiovascular Surgery at FMUSP; Collaborator E-mail: omarvilca@incor.usp.br
Physician at FMUSP, São Paulo, SP, Brazil.
6. Associate Professor at FMUSP; Head of the Surgical
Coronaropathy at InCor-HCFMUSP, São Paulo, SP, Brazil. Article received on May 23th, 2011
7. Titular Professor of the Thoracic Surgery Discipline at FMUSP; Article accepted on February 6th, 2012
66
Mejia OAV, et al. - Coronary artery bypass grafting in acute myocardial Rev Bras Cir Cardiovasc 2012;27(1):66-74
infarction: analysis of predictors of in-hospital mortality
por paciente foram avaliadas: 49 variáveis pré-operatórias

Abbreviations, Acronyms & Symbols
provenientes dos escores 2000 Bernstein-Parsonnet e
ACC American College of Cardiology EuroSCORE; 4 variáveis pré-operatórias não consideradas
AHA American Heart Association por esses escores (tempo entre o IAM e a CRM, valor máximo
IAB Intra-aortic balloon de CKMB, valor máximo de troponina e supradesnivelamento
CPB Cardiopulmonar bypass do segmento ST) e 7 variáveis intraoperatórias [uso de
CKMB creatine kinase MB isoenzyme circulação extracorpórea (CEC), tempo de CEC, tipo de
CABG coronary artery bypass graft surgery cardioplegia, endarterectomia, número de enxertos, uso da
EuroSCORE European System for Cardiac Operative Risk artéria torácica interna e revascularização completa]. Análise
Evaluation
univariada e multivariada para o desfecho mortalidade intra-
AMI acute myocardial infarction
NSTEMI NSTEMI acute myocardial infarction without hospitalar foram realizadas.
ST-segment elevation Resultados: O tempo médio entre o IAM e a CRM foi de
STEMI acute myocardial infarction with ST-segment 3,8 ± 3 dias. A mortalidade global foi 19%. Na análise
elevation multivariada: idade > 65 anos [OR 16,5 (IC 1,8-152),
IQ interquartile P=0,013]~˜ CEC >108 minutos [OR 40 (IC 2,7-578), P=0,007],
SPSS Statistical Package for Social Sciences creatinina > 2 mg/dl [OR 35,5 (IC 1,7-740), P=0,021] e pressão
pulmonar sistólica > 60 mmHg [OR 31(IC 1,6-591), P=0,022]
Resumo foram preditores de mortalidade intra-hospitalar.
Objetivo: A cirurgia de revascularização miocárdica Conclusão: Variáveis pré-operatórias clássicas como idade
(CRM) na fase aguda do infarto do miocárdio (IAM) está > 65 anos, creatinina > 2 mg/dl e pressão pulmonar sistólica
associada a aumento do risco operatório. O objetivo do estudo > 60 mmHg foram preditoras de mortalidade intra-hospitalar
foi determinar fatores preditores de mortalidade intra- nos pacientes operados de revascularização miocárdica na
hospitalar nos pacientes submetidos a CRM no IAM. fase aguda do infarto.
Métodos: Durante três anos, todos os pacientes submetidos
a CRM no IAM foram analisados retrospectivamente, Descritores: Fatores de risco. Infarto do miocárdio.
utilizando o banco de dados institucional. Sessenta variáveis Revascularização miocárdica. Ponte de artéria coronária.
INTRODUCTION The aim of this study was to elucidate what are the
predictors of in-hospital mortality in patients undergoing
The high postoperative morbidity and high mortality CABG in AMI. To have a better scientific basis, most of the
rates of patients undergoing surgical revascularization variables used in the study would have the same definitions
for acute myocardial infarction led from the old to of the EuroSCORE model [6] and 2000 Bernstein-Parsonnet
postpone the procedure. Even though the wait could [7], the same that have been validated for predicting in-
worsen and increase infarct size with borderline irrigation, hospital mortality in the Heart Institute of University of
resulting in greater myocardial damage and ventricular São Paulo (São Paulo, SP, Brazil) [8].
remodeling [1], few studies have analyzed clinical and
laboratory factors to determine that increased morbidity METHODS
and mortality. Recent studies can demonstrate positive
impact on survival with early revascularization [2], Sample Size
however this is still being delayed because of reports that Between 2008 and 2010, 62 consecutive patients
describe mortality by 31% [3]. underwent CABG during the acute phase of AMI with either
Therefore, there is no consensus about the risks and ST-segment elevation (NSTEMI) or without ST-segment
benefits of bypass surgery (CABG) for acute myocardial elevation (NSTEMI) in the Heart Institute, Clinics Hospital,
infarction (AMI), especially regarding the ideal time between Faculty of Medicine at University of São Paulo.
diagnosis and surgery [4]. Although CABG is safe when
performed electively, the effects caused by ischemic injury Criteria for inclusion and exclusion
associated with other factors such as gender, age and other We included all patients who underwent CABG during
clinical data have not been fully reported [5]. the acute phase of AMI in the period defined. CABG during
67
AMI has been shown in the following situations: (a) EuroSCORE and Parsonnet-Bernstein 2000, following the
percutaneous coronary intervention unsuccessful or definitions given by both scores [9.11].
technically infeasible, (b) patient with frank hemodynamic Data for categorical variables were expressed through
instability or (c) patients with refractory persistent their frequencies and percentages and continuous variables
symptoms to drug treatment and/or intra-aortic balloon as means and their standard deviation or median and
(IAB). The hospital records of patients were reviewed to interquartile range 25-75% (IQ25 - 75%) when the variable
determine pre-, intra- and postoperative data. Four patients is not normally distributed. Differences in categorical
were excluded from analysis due to missing data. variables were analyzed using the chi-square or Fisher
exact, when the expected values were less than five, the
Diagnostic and/or clinical criteria differences between continuous variables were analyzed
The MI was defined following the criteria of the World by independent Student t test or Mann-Whitney test when
Health Organization, which determines the presence of the variable is not normally distributed. The odds ratio
necessary diagnostic criteria in three areas: clinical, (odds ratio) was used for the weighting of the risk posed
electrocardiographic, and biochemical, within 7 days after by each categorized variable. In order to consider the
the event. Electrocardiographic abnormalities were difference of means, frequencies or the presence of
classified following the guidelines of the American Heart correlation between variables was used the statistical
Association (AHA)/American College of Cardiology (ACC) significance value less than or equal to 5% (p d” 0.05).
and NSTEMI, when the ST segment is a convex or straight Multivariate logistic regression (forward stepwise) of
curve e” 1 mm in two or more contiguous leads, or NSTEMI, pre- and intraoperative variables was performed to identify
when there is ST segment depression e” 1 mm, T wave predictors of in-hospital mortality. The modeling and
inversion or normal on ECG. Considering the fact that statistical tests were performed using the Statistical Package
patients are seen at varying times after the onset of ischemia, for Social Sciences (SPSS) version 13.0 (SPSS Inc., Chicago,
we analyzed the maximum value of the biomarkers of the IL, USA).
MB isoenzyme of creatine kinase (CK) and troponin T,
following the recommendations of the AHA / ACC to collect Ethics and consent
the samples in a sequence of 0, 3, 6 and 12 hours, followed This study was approved by the Ethics Committee for
by serial determinations of each period of 6-8 hours. The Analysis of Research Projects of the Clinics Hospital of
time interval was given by the emergency records and the University of São Paulo under the number 1575, which
calculated as the time between symptom onset and surgical exempted the need for the written informed consent because
intervention. The in-hospital mortality was defined as death it was a retrospective study. There was no conflict of
between the procedure and hospital discharge. interest of authors in this study.
Surgical Technique RESULTS

After median sternotomy, patients underwent surgery
with or without the use of cardiopulmonary bypass (CPB), The demographic and clinical characteristics of patients
the second option of the surgeon in charge. When operated are summarized in Table 1 (A, B and C). The rate of in-
using CPB, CABG was performed in normothermia or mild hospital mortality was 19% (11 of 58 patients). Of the deaths,
hypothermia and arterial cannulation was performed in the 80% were operated using CPB, of whom two patients had a
ascending aorta and venous in the right atrium. Myocardial diagnosis of VSD after AMI. Complete revascularization
protection was induced by blood or crystalloid cardioplegia was achieved in 57% (33 patients) with a mean of 2.8 ± 0.9
using the antegrade route. In patients operated without grafts per patient. The left internal thoracic artery was used
CPB it was used regional cardiac stabilizer. in all patients. Coronary thromboendarterectomy was
performed in one patient. The mean duration of CPB was
Statistical Analysis 104 ± 34 minutes. Postoperatively, 20 (34.5%) patients
Sixty variables per patient were analyzed: 49 preoperative required IABP. The stay in the intensive care unit was 10 ±
scores from the 2000 Bernstein-Parsonnet and EuroSCORE, 13 days and the duration of mechanical ventilation, 80 ±
4 preoperative variables not considered in the scores (time 149 hours. Hemodialysis was initiated in 5.2% of patients.
between the onset of AMI and CABG, maximum CKMB,
peak troponin and ST-segment elevation) and 7 Analysis of risk factors
intraoperative variables (CPB, CPB time, type of In the univariate analysis (Table 2), the risk factors related
cardioplegia, presence of carotid endarterectomy, grafts, to in-hospital mortality were severe congestive heart failure,
use of internal thoracic artery and the presence of complete advanced age, VSD after AMI, cardiogenic shock, CPB,
revascularization). All data were transformed into values of pulmonary hypertension and increased creatinine. The
68
Table 1. A. Preoperative characteristics included in the EuroSCORE and Parsonnet-Bernstein in 2000 for
deaths and survivors
Preoperative characteristics Deaths (11) Survivors (47) P
Age 69.7±13.3 63.7±13.3 0.176
Female gender 45% 27.7% 0.290
Congestive failure 90% 51.1% 0.019
Severe COPD 0% 0%
Diabetes 45.5% 42.6% 1.000
LCT > 50% 36.4% 29.8% 0.724
EF<30% 27.3% 17.0% 0.421
Arterial hypertension 90.9% 87.2% 1.000
Morbid obesity 9.1% 6.4% 1.000
Preoperative IBA 63.6% 76.6% 0.450
Reoperation - -
Treatment of aortic valve - -
Treatment of mitral valve - -
Valve treatment and CABG - -
Cardiogenic shock 27.3% 2% 0.019
Acute endocarditis - -
Treated endocarditis - -
LV aneurysmectomy - -
Treatment of tricuspid valve - -
Pacemaler dependant 9.1% 4.3% 0.474
AMI 48 h 100% 100%
Post-IVC AMI 18.2% 0% 0.033
Ventricular tachycardia 0% 2.1% 1.000
Asthma - -
Preoperative Oi 9.1% 6.4% 1.000
Pulmonary hypertension 36.4% 4.3% 0.009
Purpura thrombocytopenic 9.1% 0% 0.190
Cirrhosis 0% 0%
Dialysis dependant 9.1% 2.1% 0.346
Acute/chronic renal failure 18.2% 0% 0.033
Preoperative serum creatinine 1.49(IQ 1.22–2.34) 1.01 (IQ 0.87 – 1.26) 0.011
Presence of aortic aneurysm 0% 0%
Carotid artery disease > 70% 0% 0%
Peripheral vascular disease 0% 6.4% 1.000
Reaction to blood products 0% 2.1% 1.000
Neurologic dysfunction 0% 0%
Prior percutaneous intervention 45.5%) 19.1% 0.112
Severe smoking 9.1% 21.3% 0.671
Serum creatinine > 200 µmol/L 27.3% 2.1% 0.019
Preoperative inotropic support 36.4% 12.8% 0.083
Preoperative massage 0% 0%
Unstable angina 54.5% 80.9% 0.112
Recent myocardial infarction (<90 days) 11 (100%) 47 (100%) 1.000
Emergency surgery 36.4% 27.7% 0.715
Early aortic intervention 9.1% 4.3% 0.474
Surgery including thoracic aorta 0% 0%
Atrial fibrillation 0% 0%
Left atrial size 46 IQ(42–52) 41 IQ (39.75 – 44) 0.021
Dyslipidemia 72.7% 40.4% 0.091
Additive EuroSCORE 12.1 ± 3.2 8.3 ± 3.5 0.002
2000Bernstein-Parsonnet 39.6 ± 9.7 22.6 ± 9.9 <0.001
CABGwith/AMI: Myocardial revascularization in acute myocardial infarction; CABGwithtout/AMI: Myocardial

revascularization without acute myocardial infarction, COPD: Chronic obstructive pulmonary disease; LCT =
Trunk of left coronary artery; EF: ejection fraction; IAB = Intra-aortic balloon, post-IVC AMI = interventricular
communication after acute myocardial infarction; Oi: orotracheal intubation
69
Table 1. B. Preoperative characteristics not included in the EuroSCORE and Parsonnet-Bernstein in 2000.
Preoperative characteristics Deaths (11) Survivors (47) P
Dt: AMI-CABG 3 (IQ 1 – 9) 3 (IQ 2 – 5) 0.928
Maximum value of CKMB 9 (IQ 7.31 – 31.5) 19.6 (IQ 8.2 – 70) 0.194
Maximum value of troponin 27 (IQ 4.92 – 46.7) 8.2 (IQ 2 – 26) 0.208
ST 54.56% 23.4% 0.064
Dt = Time between the acute myocardial infarction and coronary artery bypass surgery, CK-MB = MB isoenzyme
of creatine kinase; SupraST = ST-segment elevation
Table 1. C. Intraoperative characteristics.

Intraoperative characteristics Deaths (11) Survivors (47) P
CPB use 91% 85.1% 1.000
CPB time 125.3 ± 45.1 100.2 ± 27.7 0.122
Cardioplegia time 81.8% 78.3% 1.000
Thromboendarterectomy 0% 2.2% 1.000
Number of grafts 3.0 (IQ 2.0 – 3.0) 3.0 (IQ 2.0 – 3.0) 0.759
Use of internal thoracic artery 100% 95.7% 1.000
Full revascularization (by injuried artery) 54.5% 57.4% 1.000
CPB = cardiopulmonary bypass
Table 2. Univariate analysis.

Variable OR CI P
Gender 0.46 0.12 1.77 0.290
Congestive failure 9.59 1.14 80.95 0.02
Age 4.3 1.01 18.34 0.05
Diabetes 1.13 0.3 4.21 1.000
Morbid obesity 1.47 0.14 15.61 1.000
Preoperative IAB 0.54 0.13 2.17 0.45
LCT 1.35 0.39 5.35 0.72
Pacemaker dependant 2.25 0.19 27.31 0.474
post-IVC AMI 6.22 3.42 11.32 0.033
Cardiogenic shock 17.25 1.59 187.22 0.019
Ventricular tachycardia 1.24 1.09 1.41 1.000
Vascular disease 1.25 1.1 1.43 1.000
Reaction to blood products 1.24 1.09 1.41 1.000
Dt< 4 days 1.131 0.26 4.91 1.000
Dialysis 4.60 0.27 79.92 0.346
CKMB 0.39 0.09 1.66 0.314
Troponin 2.05 0.23 18.36 1.000
ST 3.93 1.01 15.3 0.064
CPB 5.10 1.27 20.54 0.029
Prior PTCA 3.52 0.86 14.15 0.112
Inotropic support 3.91 0.87 17.46 0.083
Unstable angina 0.28 0.07 1.14 0.112
Postoperative AMI 1.39 0.31 6.22 0.696
Blood cardioplegia 1.25 0.23 6.739 1.000
Left mammary 0.80 0.71 0.92 1.000
DLP 3.93 0.92 16.74 0.091
Full revascularization 0.89 0.24 3.33 1.000
Ejection fraction < 30 1.83 0.4 8.44 0.421
Systolic pulmonary pressure >60 mmHg 12.86 1.98 83.83 0.009
Creatinine > 2 mg/dl 17.25 1.6 187.22 0.019
IAB = intra-aortic balloon; LPT = Trunk of left coronary artery; post-IVC AMI = interventricular communication
after acute myocardial infarction; Dt = time between AMI and CABG, CK-MB = MB isoenzyme of creatine kinase;
Supra ST = ST-segment elevation; CPB = cardiopulmonary bypass; prior PTCA = percutaneous intervention
prior; DLP = Dyslipidemia
70
Table 3. Predictor variables from the multivariate analysis.

Variable OR CI P
Age > 65 years 16.50 1.8 152 0.013
Creatinine > 2 mg/dl 35.45 1.7 740 0.021
Systolic pulmonary pressure >60 mmHg 30.98 1.62 591.05 0.022
CPB > 107 minutes 39.7 2.79 577.71 0.007
Constant -5.86 <0 001
CPB = cardiopulmonary bypass
mean EuroSCORE (12.1 ± 3.2) and the 2000 Bernstein- Regarding surgery as primary option, the poor results in
Parsonnet (40 ± 9.7) and deaths in relation to mean the 70s (> 20% mortality) postponed the indication for 30
EuroSCORE (8.2 ± 3.5) and the 2000 Bernstein- Parsonnet days after infarction [9]. However, actual results of the CABG
(23 ± 10) in survivors, showed statistical difference (P <0.05) can not be compared to those obtained in the 70s. Studies
for both scores. The incidence of death in patients who that compare current strategies for this invasive reperfusion
received stent prior to surgery was 37% (5/14) and who with conservative medical management group note that the
had severe lesions of the left main coronary artery, 20% (4/ invasive (surgical or percutaneous) presents a shorter
20). Among the intraoperative variables, the increase in the hospital stay, less re-hospitalization frequency, lower
time of CPB was the only significant variable in univariate incidence of nonfatal reinfarction and lower mortality [10,
analysis. Among the postoperative factors, the prolonged 11]. Otherwise, we will have a greater chance of reinfarction,
stay in intensive care unit, ventilation time and the need for due to lingering injuries, and expansion of the infarcted
hemodialysis were not significant predictors of in-hospital area, with consequent ventricular remodeling and greater
mortality in univariate analysis (Table 2). likelihood of aneurysm formation.
In multivariate analysis, preoperative variables, age> This new tendency responds to better understand of
65 years, creatinine> 2 mg/dL and systolic pulmonary the results and the selection of patients, progress of CPB
pressure> 60 mmHg were predictive of in-hospital mortality and myocardial protection, greater use of arterial grafts,
(Table 3). On the other hand, CPB> 108 minutes was not the and use of IABP and mechanical circulatory support,
only preoperative variable predictive of in-hospital mortality. improvement of postoperative care and the benefits of
Although the in-hospital mortality was higher among surgery without cardiopulmonary bypass.
patients with NSTEMI compared to NSTEMI, there was no Jatene et al. [12] published in 2001, one of the earliest
statistically significant difference. Ejection fraction <30 was studies in Brazil related to CABG in AMI, where 49 patients
not predictor of in-hospital mortality in patients with AMI. were divided into two groups: group I without complications
The time between symptom onset and surgery was divided and group II with complications (recurrent ischemia,
into three groups: < 4 days, 4 to 6 days and > 6 days. For congestive heart failure, cardiogenic shock, hypotension,
the sample period < 4 days had the worst prognosis, reinfarction, sustained ventricular tachycardia and
followed by the period> 6 days and improvement in patients ventricular fibrillation). Patients with mechanical
operated on between the 4th and 6th days, although not complications were excluded. Mortality occurred only in
statistically significant. Likewise, there was no statistical Group II (15%), with an average of 12 days from the onset
significance in relation to enzyme markers CKMB and of AMI and CRM. In our analysis, 10 years later and in a
troponin T. group that, besides the clinical characteristics of group II,
includes mechanical complications, there was a mortality
DISCUSSION of 19%, with an average of 72 hours between the onset of
AMI and CABG.
In recent years, the mortality benefit with early and late The analytic epidemiology reveals that approximately
invasive treatment strategies in patients with acute 10% of patients with MI had the same inclusion criteria as
coronary syndrome has been clearly demonstrated. In these patients in our sample and that when undergoing CABG
studies, the majority of procedures performed during the had a mortality rate of approximately 26% [13].
acute phase of AMI were the percutaneous intervention. With regard to time of intervention, we found that the
71
CABG between the fourth and sixth day had the lowest mortality, especially in STEMI. Furthermore, patients with
mortality but without statistical significance. Several NSTEMI require more than IBA and vasoactive drugs [9,10].
authors have been concerned with the optimal time of In this study, we found no statistical difference between
surgery. An analysis of 32,099 patients undergone surgery these groups.
after AMI in the state of New York (USA) between 1991 The preoperative clinical conditions determine
and 1996 showed that mortality decreased with increasing significantly different changes in the postoperative period
time interval between AMI and CABG. On the third day, the and the relationship of greater importance in predicting the
mortality curve showed an inflection, after which the levels risk. In general, there are two groups, patients with AMI
of mortality are similar to those of elective surgery. The without clinical complications and patients with AMI who
conclusion of this study was that, whenever possible present with complications of any kind, including
(considering the clinical conditions), surgery should be mechanical damage or heart failure. Clinically stable patients
postponed the first three days after AMI [14]. These trends present 1.4% mortality after CABG, compared with 12.5% in
have resulted in decreased length of stay, angina time, patients with a clinical complication preoperatively [12].
recurrent AMI, hospital infection and decreased mortality Some groups show that the left internal thoracic artery
[15]. In our analysis, a larger sample might have defined in was 50% less used to irrigate the territory of the left anterior
favor of what is currently accepted. descending artery surgery in less than 48 hours after AMI.
One important aspect relates to the inclusion of The same studies indicate that only those arteries that
appropriate variables for the type of population. In our caused ischemia should be revascularized, and the number
opinion, preoperative variables, such as type of infarction, of grafts reduced to a minimum. Thus, it is reduced surgical
time between AMI and CABG, time to angina, levels of enzyme time, and therefore the incidence of complications [9]. In
markers and critical preoperative state, are not specified by the sample, the internal thoracic artery was used in all
the scores used. The EuroSCORE adequately determines patients. We had a rate of complete revascularization in
the mortality rate for groups of medium and low risk but in 58% of patients, this using the concept of a diseased artery
high-risk group it is needed subgroups that have not yet revascularization [21], reaching 85% when used the
been established [6.16]. Thus, for example, there is evidence definition by the myocardial wall.
that the failure to percutaneous intervention during the AMI The presence of cardiogenic shock increased mortality
increases the risk of mortality, which is considered the model rate up to 59% in emergency CABG and, when the coronary
of the 2000 Bernstein-Parsonnet, but not by the EuroSCORE blood flow was not reversed, the mortality rate can reach
and the majority of existing risk models. 78% [22]. In our group, patients with AMI and cardiogenic
In our analysis, the mean EuroSCORE value was > 9 and shock who underwent surgery presented mortality rate that
the 2000 Bernstein-Parsonnet > 25, both describing a very reached 75%. Likewise, there are reports describing that in
high risk patients and showed a direct relationship between critically ill patients, the CABG without CPB decreases the
the score and mortality. The high surgical morbidity and rate of in-hospital mortality relative to CABG with CPB,
mortality that may accompany this group of patients is a mainly by decreasing the surgical time (incomplete
reflection of poor preoperative clinical condition [8], revascularization) [23]. Although CPB time exceeding 108
expressed by the high prevalence of specific predictive minutes was the only preoperative variable that was not
factors. Inside the variables, some of which had statistical predictive of mortality, it was not considered. The reason is
significance in our analysis are reported in the literature as due to the fact that in the on-pump group, there were
important. Age > 65 years had a significant relationship patients with post-AMI IVC and no analysis (at least
with mortality and this relationship was reported by univariate) comparing the mortality of patients undergone
Applebaum et al. [17] and Kaul et al. [18] in patients aged surgery with CPB with those patients on whom CPB was
over 70 years of age. The female gender has been suggested not performed.
by some authors, among them Kaul et al. [18], as a predictor In the sample there was no relation of mortality to the
of hospital mortality after CABG in AMI. Although in our levels of troponin and CK-MB, similar to what was found
analysis this correlation has not been established, other by Hagl et al. [24], reporting that the maximum values of
authors such as Applebaum et al. [17] and Naunheim et al. marker enzymes of myocardial necrosis had no impact on
[19] also found no association between female gender and survival.
early mortality. There are studies that conclude that antegrade/
Lee et al. [20] in a multicenter study involving 44,365 retrograde blood cardioplegia is superior to crystalloid
patients, found higher mortality in the group with NSTEMI cardioplegia in CABG after AMI [25]. In the sample,
compared to STEMI,only when they underwent surgery in although not statistically significant, 81% of deaths
the first week after AMI. Other authors note that the surgery undergone surgery using crystalloid cardioplegia,
within the first three days after infarction is a predictor of supporting this hypothesis.
72
Limitations of this study relate to its retrospective design database for predicting mortality after coronary artery bypass
and the fact that it was performed in a single center. Although grafting during acute myocardial infarction. Am J Cardiol.
most publications on the subject are based on small samples, 2002;90(1):1-4.
multicenter, randomized studies should answer about the
6. Nashef SA, Roques F, Michel P, Gauducheau E, Lemeshow S,
costs and benefits of CABG in AMI.
Salamon R. European system for cardiac operative risk
evaluation (EuroSCORE). Eur J Cardiothorac Surg.
CONCLUSION 1999;16(1):9-13.
Preoperative classic variables, such as age > 65 years, 7. Bernstein AD, Parsonnet V. Bedside estimation of risk as an
creatinine > 2 mg/dL and systolic pulmonary pressure > 60 aid for decision-making in cardiac surgery. Ann Thorac Surg.
mmHg, were important in predicting in-hospital mortality 2000;69(3):823-8.
of patients undergoing CABG in the acute phase of AMI.
8. Mejía OA, Lisboa LA, Puig LB, Dias RR, Dallan LA,
Pomerantzeff PM, et al. The 2000 Bernstein-Parsonnet score
and EuroSCORE are similar in predicting mortality at the Heart
Institute, USP. Rev Bras Cir Cardiovasc. 2011;26(1):1-6.
9. Braxton JH, Hammond GL, Letsou GV, Franco KL, Kopf GS,
Elefteriades JA, et al. Optimal timing of coronary artery bypass
graft surgery after acute myocardial infarction. Circulation.
1995;92(9 Suppl):II66-8.
10. Raghavan R, Benzaquen BS, Rudski L. Timing of bypass

surgery in stable patients after acute myocardial infarction.
Can J Cardiol. 2007;23(12):976-82.
11. Alter DA, Tu JV, Autsin PC, Naylor CD. Waiting times,
revascularization modality, and outcomes after acute myocardial
REFERENCES infarction at hospitals with and without on-site
revascularizations facilities in Canada. J Am Coll Cardiol.
1. Sintek CF, Pfeffer TA, Khonsari S. Surgical revascularization 2003;42(3):410-9.
after acute myocardial infarction. Does timing make a
difference? J Thorac Cardiovasc Surg. 1994;107(5):1317-21. 12. Jatene FB, Nicolau JC, Hueb AC, Atik FA, Barafiole LM,
Murta CB, et al. Fatores prognósticos da revascularização na
2. Van de Werf F, Ardissino D, Betriu A, Cokkinos DV, Falk E, fase aguda do infarto agudo do miocárdio. Rev Bras Cir
Fox KA, et al; Task Force on the Management of Acute Cardiovasc. 2001;16(3):195-202.
Myocardial Infarction of the European Society of Cardiology.
Management of acute myocardial infarction in patients 13. Yavuz S. Surgery as early revascularization after acute
presenting with ST-segment elevation. The task force on the myocardial infarction. Anadolu Kardiyol Derg. 2008;8(Suppl
Management of Acute Myocardial Infarction of the European 2):84-92.
Society of Cardiology. Eur Heart J. 2003;24(1):28-66.
14. Lee DC, Oz MC, Weinberg AD, Ting W. Appropriate timing
3. Eagle KA, Guyton RA, Davidoff R, Edwards FH, Ewy GA, of surgical intervention after transmural acute myocardial
Gardner TJ, et al; American College of Cardiology; American infarction. J Thorac Cardiovasc Surg. 2003;125(1):115-9.
Heart Association. ACC/AHA 2004 guideline update for
coronary artery bypass graft surgery: a report of the American 15. Mohr R, Moshkovitch Y, Shapira I, Amir G, Hod H, Gurevitch
College of Cardiology/American Heart Association Task Force J. Coronary artery bypass without cardiopulmonary bypass
on Practice Guidelines (Committee to Update the 1999 for patients with acute myocardial infarction. J Thorac
Guidelines for Coronary Artery Bypass Graft Surgery). Cardiovasc Surg. 1999;118(1):50-6.
Circulation. 2004;110(14):e340–437.
16. Nashef SA, Roques F, Hammill BG, Peterson ED, Michel P,
4. Deeik RK, Schmitt TM, Ihrig TG, Sugimoto JT. Appropriate Grover FL, et al; EurpSCORE Project Group. Validation of
timing of elective coronary artery bypass graft surgery European System for Cardiac Operative Risk Evaluation
following acute myocardial infarction. Am J Surg. (EuroSCORE) in North American cardiac surgery. Eur J
1998;176(6):581-5. Cardiothorac Surg. 2002;22(1):101-5.
5. Zaroff JG, diTommaso DG, Barron HV. A risk model derived 17. Applebaum R, House R, Rademaker A, Garibaldi A, Davis Z,
from the National Registry of Myocardial Infarction 2 Guillory J, et al. Coronary artery bypass grafting within thirty
73
days of acute myocardial infarction. Early and late results in 22. Hochman JS, Sleeper LA, Webb JG, Sanborn TA, White HD,
406 patients. J Thorac Cardiovasc Surg. 1991;102(5):745-52. Talley JD, et al. Early revascularization in acute myocardial
infarction complicated by cardiogenic shock. SHOCK
18. Kaul TK, Fields BL, Riggins SL, Dacumos GC, Wyatt DA, Investigators. Should We Emergently Revascularize Occluded
Jones CR. Coronary artery bypass grafting within 30 days of Coronaries for Cardiogenic Shock. N Engl J Med.
an acute myocardial infarction. Ann Thorac Surg. 1999;341(9):625-34.
1995;59(5):1169-76.
23. Rastan AJ, Eckenstein JI, Hentschel B, Funkat AK, Gummert
19. Naunheim KS, Kesler KA, Kanter KR, Fiore AC, McBride LR, JF, Doll N, et al. Emergency coronary artery bypass graft
Pennington DG, et al. Coronary artery bypass for recent infarction. surgery for acute coronary syndrome: beating heart versus
Predictors of mortality. Circulation. 1988;78(3 Pt 2):I122-8. conventional cardioplegic cardiac arrest strategies. Circulation.
2006;114(1 Suppl):I477- 85.
20. Lee DC, Oz MC, Weinberg AD, Lin SX, Ting W. Optimal
timing of revascularization: transmural versus nontransmural 24. Hagl C, Khaladj N, Peterss S, Martens A, Kutschka I, Goerler
acute myocardial infarction. Ann Thorac Surg. H, et al. Acute treatment of ST-segment-elevation myocardial
2001;71(4):1197-202. infarction: is there a role for the cardiac surgeon? Ann Thorac
Surg. 2009;88(6):1786-92.
21. Sergeant P, Blackstone E, Meyns B, Stockman B, Jashari R.
First cardiological or cardiosurgical reintervention for ischemic 25. Pichon H, Chocron S, Alwan K, Toubin G, Kaili D, Falcoz P,
heart disease after primary coronary artery bypass grafting. et al. Crystalloid versus cold blood cardioplegia and cardiac
Eur J Cardiothorac Surg. 1998;14(5):480-7. troponin I release. Circulation. 1997;96(1):316-20.
74
Revista Cubana de Cirugía
versión On-line ISSN 1561-2945
Rev Cubana Cir v.42 n.2 Ciudad de la Habana abr.-un. 2003
Instituto de Cardiología y Cirugía Cardiovascular

Ciudad de La Habana
Factores de riesgo y evaluación del riesgo de muerte

hospitalaria en la sustitución valvular mitral con
prótesis mecánica
Dr. Karel Morlans Hernández,1 Dr. José Santos Gracia,2 Dr. Fidel Manuel Cáceres
Lóriga, Dr. Horacio Pérez López4 y Dr. Asad Mirza Saadat5
Resumen
Se realizó un estudio prospectivo en 1301 pacientes sometidos a sustitución valvular

mitral, en el Instituto de Cardiología y Cirugía Cardiovascular entre enero de 1996 y
mayo de 2001. Los objetivos fueron conocer la mortalidad operatoria y establecer las
bases de un sistema de puntaje de riesgo de mortalidad. Se determinó el riesgo relativo
y las probabilidades de muerte. La mortalidad hospitalaria fue de 9,0 % (27 pacientes).
La causa de muerte más frecuente fue el fallo multiorgánico (14 pacientes,51,8 %). Los
factores de riesgo más importantes fueron: en el preoperatorio, la cardiomegalia severa
(RR 6,1) y la urgencia de la operación (RR 5,1); en el transoperatorio, el bajo gasto
cardíaco (RR 6,0) y en el posoperatorio la disfunción neurológica (RR ,33,6) y el fallo
multiorgánico (RR 26,9). La mortalidad operatoria es aceptable. El sistema de puntaje
de riesgo de muerte es factible y debe ser automatizado y validado.
DeCS: VALVULA MITRAL/cirugía; CIRUGIA TORACICA/mortalidad;

COMPLICACIONES POSTOPERATORIAS/mortalidad; MORTALIDAD
HOSPITALARIA; FACTORES DE RIESGO; IMPLANTACION DE PROTESIS DE
VALVULAS CARDIACAS/mortalidad; ISQUEMIA MIOCARDICA.
A pesar de los avances en las técnicas quirúrgicas y de preservación miocárdica, la

cirugía de sustitución valvular mitral tiene una mortalidad no despreciable. La
mortalidad quirúrgica depende de la calidad de los cuidados y procedimientos
brindados, del grado de la enfermedad previa del paciente, su comorbilidad y del azar.
En la cirugía cardiovascular, la estimación del riesgo quirúrgico basada en la mortalidad
es de utilidad para la toma de decisiones clínicas, control de calidad de la atención y la
gestión administrativa relacionadas con el procedimiento.
No existen estudios nacionales publicados sobre mortalidad y factores de riesgo para

ella en este tipo de cirugía cardíaca. Los objetivos fueron conocer la mortalidad
operatoria y establecer las bases de un sistema de puntaje de riesgo de mortalidad.
Métodos
Se estudiaron prospectivamente 301 pacientes a los cuales se les realizó cirugía de

sustitución valvular mitral en el Instituto de Cardiología y Cirugía Cardiovascular
(ICCCV) de La Habana desde enero de 1996 hasta mayo de 2001.
Se exploraron 475 variables de los períodos pre, trans y posoperatorio inmediato y se

siguió la supervivencia hasta el primer mes de la operación.
La cirugía se realizó bajo anestesia general balanceada, circulación extracorpórea (CEC)

y pinzamiento anóxico hipotérmino (PA) con cardioplejia cristaloidea fría anterógrada.
Las prótesis mitrales mecánicas utilizadas fueron de diferentes modelos.
Se realizó como prueba estadística el chi cuadrado y se determinó el riesgo relativo

(RR) de los factores de riesgo. Se tomó como nivel de significación estadística una p <
0,05. Las probabilidades de muerte (pp+ y pp-) se calcularon por análisis bayesiano.1
Resultados
La mortalidad hospitalaria fue de 9,0 % ( 27 pacientes). Cuatro pacientes fallecieron

durante el acto operatorio (14,8%), 3 casos (11,1%) en las primeras 24 h y 5 (18,5%)
entre las 24 y 72 h. Esto da una mortalidad temprana de 4,0 % (12 pacientes). La
mayoría de los decesos ocurrió después de las 72 h y hasta los primeros 30 días (15
pacientes / 5,0 %).
Las causas de muerte más frecuentes fueron el síndrome de fallo multiorgánico (FMO)
(14 pacientes/ 51,8 %), el bajo gasto cardíaco (BGC) (6 pacientes/ 22,2 %), el shok
hemorrágico y la disfunción neurológica (2 casos cada uno/ 14,8 %). Otras causas
fueron el infarto del miocardio agudo (IMA), el distrees respiratorio del adulto (SDRA)
y la disfunción renal con un caso por causa (11,1%).
Los mayores riesgos relativos y probabilidades del período preoperatorio fueron: la

cardiomegalia severa (RR 6,1 y pp+ 40,0) y la operación urgente (RR 5,1 pp+ 33,3), los
menores, la edad superior a 50 años (RR 2,7 y pp+ 15,2) y la hipertensión pulmonar
(RR 2,6 y pp+ 16,7) (tabla 1).
Tabla 1. Factores de riesgo preoperatorio de mortalidad
Factor de Vivos Muertos

R.R. P.P+ P.P- P= Total
riesgo No. % No. %
Cardiomegalia 6 2,2 4 17,4 6,1 40,0 6,5 0,0001 10
severa
Operación
8 2,9 4 17,4 5,1 33,3 6,6 0,0006 12
urgente
Endocarditis 7 2,5 3 13,0 4,4 30,0 6,9 0,002 10
Nefropatía 5 1,8 2 8,7 4,0 28,6 7,1 0,04 7
Edad > 50 años 56 20,1 10 43,5 2,7 15,2 5,5 0,009 66
HTP 30 10,8 6 26,1 2,6 16,7 6,4 0,02 36
Fuente: Dato primario.
Los factores de riesgo transoperatorio fueron el BGC (RR 6,0 pp+ 26,6 ) y tiempos de
CEC mayores de 90 min (RR 4,3 pp+ 21,7) (tabla 2).
Tabla 2. Factores de riesgo transoperatorios de mortalidad
Factor de Vivos Muertos

R.R. P.P+ P.P- P= Total
riesgo No. % No. %
BGC 30 10,8 9 47,4 6,0 26,6 4,7 0,0001 39
Bypass > 90
35 12,6 8 42,1 4,3 21,7 5,2 0,0008 43
min
Inotropos + 111 39,9 14 73,7 3,9 13,2 3,5 0,004 125
En el posoperatorio los factores de riesgo de mayor riesgo relativo y probabilidades de

muerte fueron la disfunción neurológica (RR 33,6 y pp+ 75,9), el síndrome de FMO
(RR 26,9 y pp+ 54,8) y la disfunción renal (RR 24,6 y pp+ 58,9) (tabla3).
Tabla 3. Factores de riesgo posoperatorios de mortalidad
Vivos Muertos
Factor de riesgo R.R. P.P+ P.P- P= Total
No. % No. %
Disfunción neurológica Mayor 5 1,8 13 68,4 33,6 75,9 2,6 0,0001 18
SFMO 14 5,0 14 73,7 26,9 54,8 2,2 0,0001 28
Disfunción renal 11 4,0 13 68,4 24,6 58,9 2,6 0,0001 24
Disfunción GI 6 2,2 9 47,4 16,9 64,5 4,3 0,0001 15
Paro cardiorrespiratorio 5 1,8 7 36,8 13,9 62,9 5,1 0,0001 12
Pancreatitis 2 0,7 4 21,1 12,9 70,8 6,2 0,0001 6
Disfunción pulmonar 9 3,2 11 57,9 19,0 59,7 3,5 0,0001 20
Sepsis 11 4,0 8 42,1 10,6 46,8 4,7 0,0001 19
IMAPO 4 1,4 3 15,8 7,8 47,6 6,6 0,0001 7
EAP 9 3,2 4 21,1 5,8 35,0 6,3 0,0002 13
Coagulopatía 12 4,3 4 21,1 4,7 28,7 6,4 0,002 16
Sangramiento > 100 mL/m2/h
40 14,4 8 42,1 3,8 19,5 5,3 0,002 48
en 5 1ras horas
Reintervención 17 6,1 6 31,6 5,5 29,9 5,7 0,0001 23
BGC 40 14,4 13 68,4 10,0 28,2 3,0 0,0001 53
Inotropos + 120 43,2 14 73,7 3,4 12,4 3,7 0,01 134
Fibrilación auricular 63 22,7 11 57,9 4,1 17,4 4,3 0,0006 74
Taquicardia ventricular 14 5,0 4 21,1 4,1 25,7 6,4 0,005 18
No se halló asociación entre la mortalidad y la etiología de la valvulopatía mitral,

antecedentes personales de enfermedades cerebrovascular, hepática, renal y vascular
periférica crónicas, la diabetes mellitus, el infarto miocárdico reciente, la insuficiencia
cardíaca congestiva, la CF IV según la NYHA, los diámetros aumentados del ventrículo
izquierdo y la fracción de eyección del ventrículo izquierdo baja. Tampoco se halló
asociación entre la muerte y los trastornos de la conducción avanzados durante el
período transoperatorio y el taponamiento y disfunción protésica posoperatorios.
Por otra parte, existen factores de riesgo que no exploramos en nuestra serie al no
presentar los ningún paciente como: preservación del aparato cordal mitral y tratamiento
inmunosu-presor prequirúrgico.
Discusión
La mortalidad temprana en nuestro estudio estuvo dentro del rango reportado por otros
estudios (3,7 a 12,4 %).2-5 La mortalidad hospitalaria disminuyó notablemente en
relación con el estudio anterior de nuestro centro, donde hubo el 20,7 % de fallecidos
(Navarrete J. Morbilidad cardioquirúrgica inmediata en pacientes valvulares operados
en el ICCCV entre 1988 y 1992 [Trabajo de terminación de Residencia], Ciudad de La
Habana, 1994). Nuestra cifra de mortalidad hospitalaria se ubicó dentro del rango
reportado por la mayoría de los estudios revisados (4,1 a 13,5 %).6-13
Las causas de muerte en nuestra serie fueron similares a las reportadas anatomopatoló-
gicamente por Fabri y otros,14 aunque difieren las frecuencias relativas, pues ese
estudio reportó por orden de frecuencia el shock hemorrágico (29,5 %), la FMO y la
bronconeumonía con 9,1 % cada uno.
Los factores de riesgo de mortalidad son similares a los señalados en otros

estudios,6,11,15-17 aunque pueden existir diferencias cuantitativas en algunos de ellos
como la edad, que la STS U:S: Cardiac National Database16 la reporta como riesgo si
es mayor de 65 años. La cardiomegalia severa no aparece reportada en los trabajos
revisados.
Los factores de riesgo posoperatorios para la muerte están dados por las complicaciones
que se presentan en dicha fase.
Otros informes6,8,9,11,15-18 señalan varios factores de riesgo mencionados
anteriormente en los resultados con los que no hallamos asociación en nuestros
pacientes.
El teorema de las probabilidades de Bayes, ha encontrado hasta el momento su

aplicación en medicina en el cálculo de probabilidades de la enfermedad cardíaca
isquémica.1,19,20 No hallamos ningún informe donde se aplique este método para el
cálculo de probabilidades de complicaciones en la cirugía cardíaca.
El método por lo dinámico de sus cambios, se asemeja al pensamiento lógico humano.

El manejo de los factores de riesgo como variables dicotómicas (sí o no) y de las
probabilidades en porcentajes, lo hacen más familiar al personal médico. Lo engorroso
del cálculos se solucionaría con su automatización mediante hojas de cálculo (Excel) o
base de datos (Access).
Antes de su aplicación rutinaria será necesario validarlo, de manera prospectiva, con los
datos obtenidos en este estudio, lo cual es el propósito de una próxima investigación.
Por lo tanto, los factores de riesgo preoperatorio para la mortalidad no son modificables,
pero los detectados en los períodos trans y posoperatorios sí lo son mediante su
prevención, diagnóstico y terapéutica precoces y enérgicos.
Summary
A prospective study was conducted among 1 301 patients who underwent valvular
mitral replacement at the Institute of Cardiology and Cardiovascular Surgery from
January, 1996, to May, 2001, aimed at knowing the operative mortality and to establish
the bases of a mortality risk scoring system. The relative risk and the death probabilities
were determined. The hospital mortality was 9.0 % (27 patients). The most frequent
cause of death was multiple organ failure (14 patients, 51.8 %) The most important risk
factors were: in the preoperative, severe cardiomegaly (RR 6.1) and emergency surgery
(RR 5.1); in the transoperative, low cardiac output (RR 6.0); and in the postoperative,
neurological dysfunction (RR 33.6) and multiple organ failure (RR 26.9). Operative
mortality was acceptable. The death risk scoring system is feasible and it should be
automated and validated.
DeCS: VALVULA MITRAL/cirugía; CIRUGIA TORACICA/mortalidad;

COMPLICACIONES POSTOPERATORIAS/mortalidad; MORTALIDAD
HOSPITALARIA; FACTORES DE RIESGO; IMPLANTACION DE PROTESIS DE
VALVULAS CARDIACAS/mortalidad; ISQUEMIA MIOCARDICA.
Referencias bibliográficas
1. Diamond GA, Forrester JS. Analysis of probability as an aid in the clinical

diagnosis of coronary artery disease. N Engl J Med 1979;300:1350.
2. Cesnjevar RA, Feyrer R, Walther F, Mahmoud FO, Lindemann Y, von der
Emde J. High-risk mitral valve replacement in severe pulmonary hypertension-
30 years experience. Eur J Cardiothorac Surg 1998;13(4):344-51.
3. STS U.S. Cardiac surgery database: 1997 mitral valve replacement patients:
incidence of complications summary. 2001; Disponible en:
[http://www.sts.org/doc/2990.]
4. Santini F, Gatti G, Casali G, Pessotto R, Mazzucco A. Mid-term results after
heart valve replacement with the CarboMedics bileaflet prosthesis: experience
with 785 implants. G Ital Cardiol 1999;29(7):790-5.
5. Jamieson WR, Fradet GJ, Miyagishima RT, Henderson C, Brownlee RT, Zhang
J, et al. CarboMedics mechanical prosthesis: performance at eight years. J Heart
Valve Dis 2000;9(5):678-87.
6. Braunwald E. Valvular heart disease. En: Braunwald E, ed. Heart disease. 5ed.
Philadelphia: W.B. Saunders, 1997.p.1007-76.
7. Brandäo CM, Pomerantzeff PM, Cunha CR, Morales JI, Puig LB, Grinberg M,
et al. Substituiçäo valvar com próteses mecânicas de duplo folheto. Rev Bras Cir
Cardiovasc 2000;15(3):227-33.
8. Hassouna A, Elmahalawy N. Valve replacement in rheumatic mitral
incompetence: total versus posterior chordal preservation. Cardiovasc Surg
1998;6(2):133-8.
9. Wu ZK, Sun PW, Zhang X, Zhong FT, Tong CW, Lu K. Superiority of mitral
valve replacement with preservation of subvalvular structure to conventional
replacement in severe rheumatic mitral valve disease: a modified technique and
results of one-year follow up. J Heart Valve Dis 2000;9/5):616-22.
incidence of complications summary. 2001; Disponible en:
[http://www.sts.org/doc/2990.]
11. Bouchard D, Pellerin M, Carrier M, Perrault LP, Pagé P, Hébert Y, et al. Results
following valve replacement for ischemic mitral. Can J Cardiol 2001;17(4):427-
31.
12. Remadi JP, Baron O, Roussel C, Bizouarn P, Habasch A, Despins P, et al.
Isolated mitral valve replacement with St. Jude medical prosthesis: long-term
results: a follow-up of 19 years. Circulation 2001;103(11):1542-5.
13. Delay D, Pellerin M, Carrier M, Marchand R, Auger P, Perrault LP, et al.
Immediate and long-term results of valve replacement for native and prosthetic
valve endocarditis. Ann Thorac Surg 2000;70(4):1219-23.
14. Fabri HA, Pomerantzeff PM, Monteiro AC, Gutiérrez P, Violante R, Auler JO,
et al. Análise do estudo anatomopatológico de pacientes submetidos a cirurgia
valvar no incór. Rev Bras Cir Cardiovasc 1992;7(4):243-9.
15. STS U.S. Cardiac surgery database: 1997 mitral valve replacement patients.
Univariate analysis: preoperative risk variables. 2001; Disponible en:
[http://www.ctsnet.org/doc/3061 y 65.]
univariate analysis: intraoperative variables. 2001: Disponible en:
[http://www.sts,org/doc/3062.]
univariate analysis: postoperative complications. 2001: Disponible en:
[http://www.ctsnet.org/doc/3063 y 3064.]
18. Gill IS, Masters RG, Pipe AL, Walley VM, Keon WJ. Determinants of hospital
survival following reoperative single valve replacement. Can J Cardiol
1999;15(11):1207-10.
19. Morlans J, Morlans K. Aplicación del análisis bayesiano al diagnóstico de la
cardiopatía isquémica. Rev Cubana Cardiol Cir Cardiovasc 1993;7:68-71.
20. Morlans K, Morlans J, Morlans A. Optimización de las indicaciones
diagnósticas de la cardiopatía isquémica: aplicación del análisis bayesiano
computadorizado. Rev Cubana Cardiol Cir Cardiovasc 1994;8:50-5.
Recibido: 22 de noviembre de 2001. Aprobado: 8 de diciembre de 2001.

Dr. Karel Morlans Hernández. Instituto de Cardiología y Cirugía Cardiovascular, 17
No. 702, esq. A. El Vedado, Ciudad de La Habana, Cuba.
1 Especialista de I Grado en Cardiología. Investigador Auxiliar.

2 Especialista de II Grado en Anestesiología y Reanimación. Investigador Auxiliar.
3 Especialista de II Grado en Cardiología. Investigador Agregado.
4 Vicedirector Docente del Instituto de Cardiología y Cirugía Cardiovascular.
Investigador Auxiliar.
5 Residente de 3er. Año en Cardiología.
© 2013 2002, Editorial Ciencias Médicas
Calle 23 # 654 entre D y E, Vedado

Ciudad de La Habana, CP 10400
Cuba
ecimed@infomed.sld.cu
doi:10.1093/icvts/ivs181 Advance Access publication 7 May 2012
Outcome and survival analysis of intestinal ischaemia following

cardiac surgery†
Philip Y.K. Pang, Yoong Kong Sin*, Chong Hee Lim, Jang Wen Su and Yeow Leng Chua
Department of Cardiothoracic Surgery, National Heart Centre, Singapore, Singapore
* Corresponding author. National Heart Centre, Mistri Wing, 17 Hospital Avenue, Singapore 168752, Singapore. Tel: +65-6436-7598; fax: +65-6224-3632;
e-mail: kenny.sin.y.k@nhcs.com.sg (Y.K. Sin).
ORIGINAL ARTICLE
Received 14 January 2012; received in revised form 15 March 2012; accepted 1 April 2012
Abstract
OBJECTIVES: Intestinal ischaemia is an uncommon (<1%) but serious complication of cardiac surgery with a mortality rate exceeding
50%. Diagnosis of this potentially lethal condition can be difficult and requires a high index of suspicion. The purpose of this study was
to analyse the outcomes and prognostic factors in patients who develop intestinal ischaemia following cardiac surgery.
METHODS: In a retrospective review from August 1999 to December 2010, we identified 31 out of 9925 (0.31%) consecutive patients
who developed acute intestinal ischaemia following cardiac surgery at our tertiary centre.
RESULTS: The overall mortality was 71.0%. The operative mortality was 65.4% in patients who underwent a laparotomy. Survivors of
this complication had surgical intervention earlier (7.4 ± 4.9 h) compared with the non-survivors (13.9 ± 11.1 h). A total of 35 periopera-
tive variables were analysed. A univariate analysis identified 12 variables associated with an increased risk of mortality. Logistic multi-
variate analysis identified the preoperative logistic EuroSCORE and the base excess at the point of diagnosis of intestinal ischaemia as
significant predictors of mortality. These factors may aid prognostication in this group of patients.
CONCLUSIONS: Despite the high mortality rates associated with intestinal ischaemia following cardiac surgery, early diagnosis and sur-
gical intervention remain the only effective means to reduce mortality.
Keywords: Intestinal ischaemia • Cardiac surgery
INTRODUCTION and database review was performed on 9925 consecutive

patients who had undergone elective and emergency cardiac
Intestinal ischaemia post-cardiac surgery has a reported inci- surgical procedures between August 1999 and December 2010
dence of <1% [1–10]. Early diagnosis can be difficult but is im- at our tertiary referral centre. This study focused exclusively on
portant as prompt surgical intervention can be lifesaving [5, 8]. patients with acute mesenteric ischaemia involving the small
The difficulty in making the diagnosis contributes heavily to and/or large bowel. Patients who developed other gastrointes-
the dismal outcome. The signs and symptoms of abdominal tinal complications (e.g. ischaemic hepatitis, gastrointestinal
pathology are often masked by prolonged mechanical ventilation bleeding, pancreatitis) were excluded. Patients who presented
and sedation especially for critically ill patients after cardiac with acute aortic dissection with evidence of intestinal ischaemia
surgery. A high index of suspicion is required for early diagnosis. prior to cardiac surgery were excluded.
Despite advances in critical care and prompt surgical interven- Preoperative renal insufficiency was defined by serum cre-
tion, the mortality rate of acute intestinal ischaemia remains atinine levels higher than 120 mmol/l. Inotropic support was
high, and ranges from 46.1 to 100% [1–5, 7, 8, 11–14]. defined as infusions of dopamine over 5 mcg/kg/min or any
This study aims to review patients who had undergone cardiac use of adrenaline, noradrenaline or vasopressin. Preoperative
surgery at a tertiary referral centre over an 11-year period, to deter- anticoagulation was administered via intravenous heparin
mine the incidence of intestinal ischaemia following cardiac surgery infusion. Patients with a postoperatively low cardiac output
and investigate their outcomes and possible prognostic factors. were identified as those manifesting signs of cardiogenic
shock and decreased end-organ perfusion (oliguria, altered
mental state, hypoxia, metabolic acidosis) requiring the use of
MATERIALS AND METHODS one or more vasopressors (dopamine, adrenaline, noradren-
aline or vasopressin) beyond the first 24 h of cardiac surgery.
Following approval from the Singhealth Centralised Institutional Some of these patients required further support with
Review Board (reference: 2011/835/C), a retrospective case-note intra-aortic balloon pump (IABP) or extracorporeal membrane
†
Poster presentation at CHEST 2011, Hawaii, 26 October 2011.
oxygenation (ECMO).
216 P.Y.K. Pang et al. / Interactive CardioVascular and Thoracic Surgery
Table 1: Patient demographics and comorbidities
Variable Survivors, n = 9 (%) Non-survivors, n = 22 (%) P-value
Demographics
Age 66.2 ± 8.1 68.3 ± 6.3 0.455
Gender (male/female) 7/2 (77.8/22.2) 12/10 (54.5/45.5) 0.418
LVEF 41.9 ± 16.2 42.3 ± 20.2 0.960
LVEF <30% 3 (33.3) 6 (27.3) 0.528
Comorbidities
Diabetes mellitus 3 (33.3) 11 (50.0) 0.329
Obesity (BMI >25) 5 (55.6) 9 (40.9) 0.363
Renal failure 2 (22.2) 7 (31.8) 0.472
Dialysis 1 (11.1) 1 (4.5) 0.499
Hypertension 8 (88.9) 21 (95.5) 0.503
Stroke 1 (11.1) 2 (9.1) 0.657
Peripheral vascular disease 2 (22.2) 7 (31.8) 0.472
COPD 1 (11.1) 1 (4.5) 0.499
Triple vessel coronary artery disease 7 (77.8) 18 (81.8) 0.577
Previous laparotomy 3 (33.3) 3 (13.6) 0.391
BMI: body mass index; COPD: chronic obstructive pulmonary disease; LVEF: left ventricular ejection fraction.
Statistical analysis was performed using the SPSS version 17 stat-

istical software. Continuous variables were expressed as a mean ± Table 2: Univariate risk analysis for mortality in patients
standard deviation and were compared using the two-tailed t-test. with intestinal ischaemia
Categorical variables, expressed as percentages, were analysed with
the χ 2 or Fisher’s exact test. To identify the risk factors predictive of
Variable Survivors, Non-survivors, P-value
mortality, a univariate analysis of 35 perioperative variables was n = 9 (%) n = 22 (%)
performed. Significant univariate risk factors were examined using a
stepwise logistic regression analysis. A two-tailed P-value <0.05 was Preoperative
used to indicate statistical significance. Emergency surgery 1 (11.1) 9 (40.9) 0.107
Anticoagulation 1 (11.1) 12 (54.5) 0.026
Unstable angina 2 (22.2) 10 (45.5) 0.228
IABP 1 (11.1) 9 (40.9) 0.107
RESULTS Logistic EuroSCORE 5.7 ± 4.8 26.1 ± 20.6 0.007
Intraoperative
Of the total 9925 patients who had undergone cardiac RBC transfusion (units) 0.56 ± 0.53 1.68 ± 1.36 0.023
Postoperative
surgery between August 1999 and December 2010, acute Renal failure 4 (44.4) 19 (86.4) 0.015
intestinal ischaemia was identified in 31 patients (0.31%), Postoperative CVVH 4 (44.4) 18 (81.8) 0.037
diagnosed on either laparotomy or postmortem. Of these, 29 Atrial fibrillation 9 (100.0) 14 (63.6) 0.036
patients (93.5%) had undergone conventional surgery with the Abdominal pain or 9 (100) 10 (54.5) 0.014
use of cardiopulmonary bypass (CPB) and the remaining 6.5% tenderness
Vasopressors beyond 3 (33.3) 15 (68.2) 0.074
had undergone off-pump surgery. Twenty-two patients died, 24 h of surgery
resulting in a mortality rate of 71.0%. All patients had a trans- IABP 3 (33.3) 14 (63.6) 0.124
mural infarction of either the small (64.5%) or large bowel ECMO 0 (0.0) 3 (13.6) 0.244
(19.4%), or both (16.1%). Low cardiac output 5 (55.6) 20 (90.9) 0.024
Dilated bowel loops 9 (100) 12 (54.5) 0.014
The incidence of acute intestinal ischaemia was 0.2% for elect- Bowel wall thickening 2 (22.2) 0 (0.0) 0.022
ive cases compared with 1.2% for emergency cases (P < 0.001). Presence of metabolic 6 (66.7) 21 (95.5) 0.030
Ten patients (32.3%) underwent emergency cardiac surgery, all acidosis
of which were coronary artery bypass grafting (CABG). There was Base excess −2.9 ± 4.1 −8.8 ± 4.6 0.003
no significant difference in the incidence of intestinal ischaemia
following on-pump compared with off-pump surgery (0.32 vs. ECMO: extracorporeal membrane oxygenation; EuroSCORE: European
0.21%, P = 0.562). Of the 31 patients with intestinal ischaemia, 21 system for cardiac operative risk evaluation; CVVH: continuous
veno-venous haemofiltration; RBC: red blood cell.
(67.7%) patients had undergone isolated CABG, 7 (22.6%) had
valve surgery and CABG, 2 (6.5%) had isolated valve surgery and
1 (3.2%) had undergone excision of a left atrial myxoma.
Patient demographics and comorbidities are shown in Table 1.
There were no significant differences in the demographic charac- Factors significantly affecting survival following univariate ana-
teristics between the two groups. Females had a higher mortality lysis are shown in Table 2. The following factors did not signifi-
rate compared with males (83.3 vs. 63.2%, P = 0.418). Pre-existing cantly affect survival: preoperative (emergency surgery, recent
medical conditions were similar and did not significantly myocardial infarction within 90 days, cardiogenic shock, inotrop-
affect survival. ic support, IABP), intraoperative [CPB time, aortic cross-clamp
P.Y.K. Pang et al. / Interactive CardioVascular and Thoracic Surgery 217
High-risk patients with unstable angina or those supported with

Table 3: Multivariate risk analysis for mortality in an IABP preoperatively were identified and anticoagulated with
patients with intestinal ischaemia intravenous heparin. Compared with those who did not require
preoperative anticoagulation, they had a significantly higher
Odds 95% confidence P-value logistic EuroSCORE and also a higher risk of mortality upon
ratio interval developing acute intestinal ischaemia.
Patients who developed new-onset atrial fibrillation post-
Logistic EuroSCORE 20.2 1.31–312.1 0.031 operatively had a lower risk of mortality from acute bowel is-
>10.5%
Base excess <−6.0 12.3 1.04–145.1 0.047
chaemia due to early postoperative anticoagulation and close
Postoperative low 3.49 0.23–53.0 0.368 monitoring. The incidence of other postoperative complications
cardiac output was generally higher in the group of non-survivors compared
with the survivors. Notably, acute renal failure occurring concur-
rently with intestinal ischaemia was shown to be a lethal com-
ORIGINAL ARTICLE
bination, being observed in most of the non-survivors.
The time interval to laparotomy in our series was mainly
(AXC) time] postoperative (re-exploration for bleeding, myocar- limited by the time taken to establish the diagnosis of intestinal
dial infarction, sepsis, IABP, ECMO, use of vasopressors beyond ischaemia. The variety of clinical presentations, which was
24 h of surgery, limb ischaemia, disseminated intravascular coa- further masked by sedation in mechanically ventilated patients,
gulopathy, stroke, anticoagulation, abdominal distension, radio- made the diagnosis of intestinal ischaemia particularly difficult.
graphic evidence of pneumatosis or extraluminal air). Abboud et al. [5] reported the time interval to laparotomy in
Univariate analysis identified the need for preoperative anticoa- their series to be 13.7 ± 19.0 h. Chaudhuri et al. [7] reported a
gulation with intravenous heparin as a risk factor for mortality. relatively longer median interval to laparotomy of 8 days for
These patients requiring anticoagulation also had a higher Mean their series of 15 patients who underwent a laparotomy.
logistic EuroSCORE compared with those who did not require Compared with non-survivors, survivors of intestinal ischaemia
anticoagulation (33.4 vs. 10.7%, P = 0.001) [15]. The non-survivors underwent laparotomy earlier (6.4 ± 3.8 vs. 16.9 ± 10 h, P = 0.519)
required significantly more intraoperative blood transfusion. [8]. Similarly to our findings, this difference was not statistically
A total of 26 (83.9%) of the 31 patients underwent a laparot- significant.
omy following diagnosis. The remaining five patients were either Apart from non-specific dilated bowel loops, there are often
unfit for a laparotomy or declined surgery. The operative mortal- no other signs on plain abdominal X-rays, until a severe intes-
ity in the laparotomy group was 65.4%. The mortality was 100% tinal infarction occurs, during which pneumotosis, frank perfor-
for the five patients who did not undergo surgery. They all died ation or portal venous gas may develop. Although abdominal
within 24 h of the clinical diagnosis, with intestinal ischaemia pain and radiological features of bowel dilatation or wall thicken-
being confirmed post mortem. The mean time to laparotomy ing were more common in the group of survivors, these findings
from the time of diagnosis was shorter in the group of survivors did not lead to a decreased time interval to laparotomy. Rather,
(7.4 ± 4.9 h) compared with the non-survivors (13.9 ± 11.1 h). the improved outcomes may have been due to these findings fa-
However, this result was not statistically significant (P = 0.127). cilitating closer monitoring and prompt diagnosis, allowing surgi-
Further logistic multivariate analysis identified the preopera- cal intervention at an earlier stage of mesenteric ischaemia.
tive EuroSCORE and the base excess at the point of diagnosis of All laparotomies were performed based on sufficient clinical
intestinal ischaemia as the two most important predictors of evidence of acute intestinal ischaemia. To avoid delaying
mortality in this series, as shown in Table 3. surgery, none of these patients underwent mesenteric angiog-
raphy or computed tomography. Furthermore, most of them
were in a state of haemodynamic instability and thus unfit to be
transferred for imaging.
DISCUSSION The logistic multivariate analysis identified the preoperative
EuroSCORE and the base excess at the point of diagnosis of in-
The overall incidence of intestinal ischaemia following cardiac testinal ischaemia as the two most important predictors of mor-
surgery in our series was 0.31%, comparable with the 0.36% (31/ tality in this series. The preoperative logistic EuroSCORE was
8709 patients) described by Mangi but differs from the 0.15% used to risk-stratify patients prior to the initial cardiac surgery.
(17/11 058 patients), 0.25% (16/6452 patients), 0.62% (30/4819 This risk stratification model was found to be applicable in our
patients) and 0.73% (39/5349 patients) reported by other authors study population who developed intestinal ischaemia after
[1–3, 8, 10]. cardiac surgery. Patients with a higher logistic EuroSCORE had
The overall mortality was 71.0% in this series, similar to the an increased risk of mortality from intestinal ischaemia. The se-
64.1 and 67.0% reported by Ghosh et al. [8] and Allen et al. [6], verity of metabolic acidosis at the time of the diagnosis of intes-
respectively. Similarly to our findings, Ghosh et al. [8] reported a tinal ischaemia, as measured with base excess values, was found
100% mortality rate in patients with intestinal ischaemia who did to correlate with the extent of the bowel infarction. Base excess
not undergo a laparotomy. The operative mortality in the lapar- values may also serve as a surrogate marker to reflect the level
otomy group was 65.4%, which is comparable with the 63.6% of critical illness in this group of patients. Severe metabolic acid-
reported by Mangi et al. [3], but lies between the 42.8 and 80.0% osis was observed in cases where an extensive bowel infarction
reported by Abboud et al. [5] and Chaudhuri et al. [7], respective- had occurred, in which the prognosis was invariably dismal.
ly. These findings reinforce the fact that acute intestinal ischae- Persistent metabolic acidosis refractory to bicarbonate infusions
mia after cardiac surgery is an infrequent but catastrophic event. occurred in 37.0% of patients without associated abdominal
218 P.Y.K. Pang et al. / Interactive CardioVascular and Thoracic Surgery
signs or abnormal radiological findings. Thus, intestinal ischaemia ACKNOWLEDGEMENT

must be excluded in patients with unexplained and persistent
metabolic acidosis after cardiac surgery, as it may be the earliest We wish to thank Zakir Hussain Abdul Salam for his help with
and only diagnostic clue of mesenteric ischaemia. the statistical analysis in this paper.
Risk factors for the development of intestinal ischaemia follow-
ing cardiac surgery have been well described. Some of these Conflict of interest: none declared.
factors include advanced age, prolonged CPB or AXC time, IABP
usage, significant inotropic support, peripheral vascular disease,
emergency cardiac surgery and postoperative renal failure [2, 7, 8]. REFERENCES
In contrast, there are few reports of predictors of mortality
from gastrointestinal complications following cardiac surgery. [1] D’Ancona G, Baillot R, Poirier B, Dagenais F, de Ibarra JI, Bauset R et al.
Zacharias et al. [4] analysed a group of patients with gastrointes- Determinants of gastrointestinal complications in cardiac surgery. Tex
Heart Inst J 2003;30:280–5.
tinal complications following cardiac surgery who underwent
[2] Filsoufi F, Rahmanian PB, Castillo JG, Scurlock C, Legnani PE, Adams DH.
abdominal surgery. In this group of 37 patients, 22 patients suf- Predictors and outcome of gastrointestinal complications in patients
fered from intestinal ischaemia, 12 patients had acute cholecyst- undergoing cardiac surgery. Ann Surg 2007;246:323–9.
itis and three patients developed perforated duodenal ulcers. [3] Mangi AA, Christison-Laray ER, Torchiana DF, Warshaw AL, Berger DL.
The operative mortality was 68.2% for those with intestinal Gastrointestinal complications in patients undergoing heart operation:
an analysis of 8709 consecutive cardiac surgical patients. Ann Surg 2005;
ischaemia. The authors reported that preoperative renal failure 241:895–901.
and prolonged CPB or AXC time were factors associated with [4] Zacharias A, Schwann TA, Parenteau GL, Riordan CJ, Durham SJ, Engoren
the increased mortality in this surgically treated group. M et al. Predictors of gastrointestinal complications in cardiac surgery.
However, these factors were not statistically significant in our Tex Heart Inst J 2000;27:93–9.
[5] Abboud B, Daher R, Sleilaty G, Madi-Jebara S, El Asmar B, Achouch R
study.
et al. Is prompt exploratory laparotomy the best attitude for mesenteric
This is a retrospective descriptive study with inherent biases in ischemia after cardiac surgery? Interact CardioVasc Thorac Surg 2008;7:
data collection. Owing to the relatively rare occurrence of intes- 1079–83.
tinal ischaemia, the small sample size underpowered the statis- [6] Allen KB, Salam AA, Lumsden AB. Acute mesenteric ischemia after car-
tical analysis and could have limited the number of statistically diopulmonary bypass. J Vasc Surg 1992;16:391–6.
[7] Chaudhuri N, James J, Sheikh A, Grayson AD, FabriI BM. Intestinal is-
significant variables. A prospective study incorporating a larger chaemia following cardiac surgery: a multivariate risk model. Eur J
sample size would be useful to assess the prognostic value of the Cardiothorac Surg 2006;29:971–7.
risk factors identified. [8] Ghosh S, Roberts N, Firmin RK, Jameson J, Spyt TJ. Risk factors for intes-
tinal ischaemia in cardiac surgical patients. Eur J Cardiothorac Surg 2002;
21:411–6.
[9] Schütz A, Echinger W, Breuer M, Gansera B, Kemkes BM. Acute
mesenteric ischemia after open heart surgery. Angiology 1998;49:
CONCLUSION 267–73.
[10] Leitman IM, Paull DE, Barie PS, Isom OW, Shires GT. Intra-abdominal
complications of cardiopulmonary bypass operations. Surg Gynecol
The incidence of acute intestinal ischaemia following cardiac Obstet 1987;165:251–4.
surgery is 0.31% and is associated with a mortality rate of 71.0%. [11] Krasna MJ, Flancbaum L, Trooskin SZ, Fitzpatrick JC, Scholz PM, Scott GE
It is a diagnosis that must be excluded in all patients who et al. Gastrointestinal complications after cardiac surgery. Surgery 1988;
develop abdominal signs or persistent metabolic acidosis follow- 104:773–80.
[12] Lazar HL, Hudson H, McCann J, Fonger JD, Birkett D, Aldea GS et al.
ing cardiac surgery. Although the clinical and radiological signs
Gastrointestinal complications following cardiac surgery. Cardiovasc Surg
are non-specific, their presence may facilitate an earlier diagno- 1995;3:341–4.
sis and surgical intervention. Severe metabolic acidosis and a [13] Hasan S, Ratnatunga C, Lewis CT, Pillai R. Gut ischaemia following
high preoperative logistic EuroSCORE are factors associated with cardiac surgery. Int Cardiovasc Thorac Surg 2004;3:475–8.
increased mortality. Although mortality rates remain high [14] Huddy SP, Joyce WP, Pepper JR. Gastrointestinal complications in 4473
patients who underwent cardiopulmonary bypass surgery. Br J Surg
despite early surgical intervention, a low threshold for a prompt 1991;78:293–6.
laparotomy remains the most reliable means for a potential re- [15] Roques F, Michel P, Goldstone AR, Nashef SA. The logistic EuroSCORE.
duction in the mortality rate. Eur Heart J 2003;24:882–3.
Warning: The NCBI web site requires JavaScript to function. more...
Sponsored Document from

Lancet
Lancet. 2012 September 22; 380(9847): 1059–1065.
doi: 10.1016/S0140-6736(12)61148-9
PMCID: PMC3493988
Mortality after surgery in Europe: a 7

day cohort study
Rupert M Pearse,a,* Rui P Moreno,b Peter Bauer,c Paolo Pelosi,e Philipp Metnitz,d
Claudia Spies,f Benoit Vallet,g Jean-Louis Vincent,h Andreas Hoeft,i Andrew Rhodes,j,k
and for the European Surgical Outcomes Study (EuSOS) group for the Trials groups of
the European Society of Intensive Care Medicine and the European Society of
Anaesthesiology†
a
Barts and The London School of Medicine and Dentistry, Queen Mary University of
London, London, UK
b
UCINC, Hospital de São José, Centro Hospitalar de Lisboa Central, EPE, Lisbon,
Portugal
c
Section of Medical Statistics, Medical University of Vienna, Vienna, Austria
d
Department of Anaesthesia and General Intensive Care, Medical University of Vienna,
Vienna, Austria
e
IRCCS AOU San Martino-IST, Department of Surgical Sciences and Integrated
Diagnostics, University of Genoa, Genoa, Italy
f
Charité-Universitaetsmedizin, Berlin, Germany
g
Anaesthesiology and Critical Care, University Hospital, Lille, France
h
Erasme Hospital, Université Libre de Bruxelles, Brussels, Belgium
i
Department of Anaesthesiology, University of Bonn, Bonn, Germany
j
St George's Healthcare NHS Trust, London, UK
k
St George's University of London, London, UK
Rupert M Pearse: r.pearse@qmul.ac.uk
*
Correspondence to: Dr Rupert Pearse, Adult Critical Care Unit, Royal London
Hospital, London E1 1BB, UK ; Email: r.pearse@qmul.ac.uk
†
Members listed in appendix
Copyright © 2012 Elsevier Ltd. All rights reserved.
This document may be redistributed and reused, subject to certain conditions.
This document was posted here by permission of the publisher. At the time of the
deposit, it included all changes made during peer review, copy editing, and publishing.
The U. S. National Library of Medicine is responsible for all links within the document
and for incorporating any publisher-supplied amendments or retractions issued
subsequently. The published journal article, guaranteed to be such by Elsevier, is
available for free, on ScienceDirect, at: http://dx.doi.org/10.1016/S0140-
6736(12)61148-9
Go to:
Summary
Background
Clinical outcomes after major surgery are poorly described at the national level.
Evidence of heterogeneity between hospitals and health-care systems suggests potential
to improve care for patients but this potential remains unconfirmed. The European
Surgical Outcomes Study was an international study designed to assess outcomes after
non-cardiac surgery in Europe.
Methods
We did this 7 day cohort study between April 4 and April 11, 2011. We collected data
describing consecutive patients aged 16 years and older undergoing inpatient non-
cardiac surgery in 498 hospitals across 28 European nations. Patients were followed up
for a maximum of 60 days. The primary endpoint was in-hospital mortality. Secondary
outcome measures were duration of hospital stay and admission to critical care. We
used χ2 and Fisher's exact tests to compare categorical variables and the t test or the
Mann-Whitney U test to compare continuous variables. Significance was set at p<0·05.
We constructed multilevel logistic regression models to adjust for the differences in
mortality rates between countries.
Findings
We included 46 539 patients, of whom 1855 (4%) died before hospital discharge. 3599
(8%) patients were admitted to critical care after surgery with a median length of stay of
1·2 days (IQR 0·9–3·6). 1358 (73%) patients who died were not admitted to critical
care at any stage after surgery. Crude mortality rates varied widely between countries
(from 1·2% [95% CI 0·0–3·0] for Iceland to 21·5% [16·9–26·2] for Latvia). After
adjustment for confounding variables, important differences remained between
countries when compared with the UK, the country with the largest dataset (OR range
from 0·44 [95% CI 0·19–1·05; p=0·06] for Finland to 6·92 [2·37–20·27; p=0·0004] for
Poland).
Interpretation
The mortality rate for patients undergoing inpatient non-cardiac surgery was higher than
anticipated. Variations in mortality between countries suggest the need for national and
international strategies to improve care for this group of patients.
Funding
European Society of Intensive Care Medicine, European Society of Anaesthesiology.
Go to:
Introduction
More than 230 million major surgical procedures are undertaken worldwide each year.1
For most patients, risks of surgery are low and yet evidence increasingly suggests that
complications after surgery are an important cause of death.2–5 About 10% of patients
undergoing surgery in the UK are at high risk of complications, accounting for 80% of
postoperative deaths.2–4 If this rate is applicable worldwide, up to 25 million patients
undergo high-risk surgical procedures each year, of whom 3 million do not survive until
hospital discharge. Patients who develop complications but survive to leave hospital
often have reduced functional independence and long-term survival.5–8
Despite obvious differences in procedure-related and patient-related mortality risks,

most surgical patients use one care pathway, sharing standard facilities for preoperative
assessment, anaesthesia, operating rooms, post-anaesthetic recovery, and hospital
wards. This approach is adequate for most patients but might not meet the needs of the
small number of patients at high risk of complications and death. In the USA, evidence
of variations in postoperative mortality within health-care systems suggest the potential
to implement measures that improve patient outcomes.9 Low rates of admission to
critical care for patients at high risk of complications undergoing non-cardiac surgery
are of particular concern,2–4 and might be affected by international differences in the
provision of critical care.10,11 With high volumes of surgery undertaken, even a low rate
of avoidable harm will be associated with many preventable deaths.
International comparative data might provide important insights into delivery of health
care for surgical patients. However, little or no data are available describing provision
of care or outcomes for unselected surgical patients. The objective of the European
Surgical Outcomes Study (EuSOS) was to describe mortality rates and patterns of
critical care resource use for patients undergoing non-cardiac surgery across several
European nations.
Go to:
Methods
Study design and participants
We did this European cohort study between 0900 h (local time) on April 4, 2011, and
0859 h on April 11, 2011. All adult patients (older than 16 years) admitted to
participating centres for elective or non-elective inpatient surgery commencing during
the 7 day cohort period were eligible for inclusion. Patients undergoing planned day-
case surgery, cardiac surgery, neurosurgery, radiological, or obstetric procedures were
excluded because these patients receive care within separate, dedicated pathways.
Participating hospitals (appendix pp 11–68) were a voluntary convenience sample,
identified through membership of the European Society of Intensive Care Medicine and
the European Society of Anaesthesiology and by direct approach from national study
coordinators. Ethics requirements differed by country. In Denmark, centres were
exempt from ethics approval because this study was deemed to be a clinical audit. In all
other nations formal ethics approval was applied for and given. In Finland alone we
were required to obtain written informed consent from individual patients.
Procedures
Local investigators were supported by national coordinators and via a website that
provided key documentation, including the protocol and guidance on study procedures.
We obtained data describing perioperative care facilities once for each hospital at the
beginning of the study. We collected data describing consecutive patients with paper
case record forms, which we made anonymous before entering the information onto a
secure internet-based electronic case record form (OpenClinica, Boston, MA, USA). We
completed an operating theatre case report form for each eligible patient who we then
followed up until hospital discharge for data describing hospital stay, admission to
critical care, and in-hospital mortality. We completed a critical care case record form to
capture data describing the first admission to critical care for any individual patient at
any time during the follow-up period. Example case record forms are available from the
study website.
We selected patient-level variables on the basis that they were objective, routinely
collected for clinical reasons, could be transcribed with a high level of accuracy, and
would be relevant to a risk adjustment model in most patients. We censored critical care
and hospital discharge data at 60 days after surgery. We assessed data for completeness
and then checked for plausibility and consistency with prospectively defined ranges.12
The primary endpoint was in-hospital mortality. Secondary outcome measures were
duration of hospital stay and admission to critical care.
Our aim was to recruit as many participating hospitals as possible and to recruit every
eligible patient in those hospitals. We anticipated that a minimum sample size of 20 000
patients would enable a precise estimate of mortality. This sample size was also
expected to provide a sufficient number of events (>200) for construction of a robust
logistic regression model for mortality.
We used SPSS (version 19.0) for data analysis. Categorical variables are presented as
number (%) and continuous variables as mean (SD) when normally distributed or
median (IQR) when not. We used χ2 and Fisher's exact tests to compare categorical
variables and the t test or the Mann-Whitney U test to compare continuous variables.
Significance was set at p<0·05. We constructed several binary logistic regression
models to identify factors independently associated with hospital mortality and to adjust
for differences in confounding factors between countries. These included a one-level
model and a hierarchical two-level generalised linear mixed model, with patients being
at the first level and hospital at the second. Factors were entered into the model based
on their univariate relation to outcome (p<0·05). All factors were biologically plausible
with a sound scientific rationale and a low rate of missing data. The results of the model
are reported as adjusted odds ratios (OR) with 95% CI. We assessed the models through
sensitivity analyses with three random (disjoint) subsamples of countries and a fourth
sample removing all patients from the largest country in the dataset (the UK). We
explored all possible interacting factors and examined how they might have affected the
final results.
This study is registered with ClinicalTrials.gov, number NCT01203605.
Role of the funding source

The study was funded by the European Society of Intensive Care Medicine and the
European Society of Anaesthesiology who appointed an independent steering
committee (appendix p 11), who were responsible for study design, conduct, and data
analysis. Members of the steering committee had full access to the study data and were
solely responsible for interpretation of the data, drafting and critical revision of the
report, and the decision to submit for publication.
Go to:
Results
We collected data describing patients undergoing in-patient surgery in 498 hospitals
across 28 European nations. Median number of operating theatres in each hospital was
15 (IQR 10–22) and median number of critical care beds was 19 (9–40). Data were
returned for 46 985 cases of which 446 were removed having been identified as
duplicates or having missing critical care or mortality data, leaving 46 539 for analysis
(figure 1). A median number of 83 (39–125) patients were included per hospital and
1045 (455–1732) per country. 281 (56%) hospitals were affiliated to a university,
recruiting 31 132 patients (68% of total, appendix p 2).
Table 1 shows baseline data for all patients. Overall crude mortality was 4·0% and the
median duration of hospital stay was 3·0 days (IQR 1·0–7·0). Prevalence of comorbid
disease, grade of surgery, crude mortality rates, duration of hospital stay, and number of
critical care admissions differed substantially between countries (table 2, appendix p 2).
Table 2 shows unadjusted OR for hospital mortality by country. 3599 patients (8%)
were admitted to critical care at some point during hospital stay, of whom 2555 (71%)
had planned admissions (figure 2). Median stay in critical care was 1·2 days (0·9–3·6).
1358 patients who died were not admitted to critical care at any stage after surgery
(73% of all deaths). 506 patients (14%) admitted to critical care died before hospital
discharge, of whom 218 (43%) died after the first admission to critical care was
complete.
We explored variables associated with hospital mortality in a univariate analysis, the

findings of which were much the same as for a sensitivity analysis of different subsets
of the database (table 1, appendix pp 3–4). We then constructed several binary logistic
regression models to adjust for baseline differences that might explain the unadjusted
OR for individual countries (table 2). We developed both single-level and multilevel
models (appendix pp 5–8) with variables that were significant in the univariate analysis.
The point estimates for the OR did not differ greatly between the one-level and two-
level models, but the hierarchical model consistently provided a more conservative
estimate of country effects across the sensitivity tests (appendix p 9).
We constructed a further model including all significant interacting factors (appendix p

10). Since this increased model complexity did not substantially change the country-
level estimates, we report results of the more parsimonious two-level model without
interactions (figure 3). Factors that were independently associated with mortality and
that we therefore used to adjust for baseline confounders were: country where surgery
was done, urgency of surgery, grade of surgery, surgical procedure category, age,
American Society of Anesthesiologists (ASA) score, metastatic disease, and cirrhosis
(appendix pp 7–8). We entered ASA score rather than the Lee Revised Cardiac Index
because, although the two were highly correlated, less data describing ASA score were
missing.
With the UK study population as the reference category, we identified higher

unexplained rates of mortality in Poland, Romania, Latvia, and Ireland (table 2, figure
3).
Go to:
Discussion
This international prospective study has provided data for a population of more than
46 000 unselected patients undergoing inpatient surgery from 28 European countries.
4% of included patients died before hospital discharge, which was a higher mortality
rate than expected.2,3,6,13–16 We identified substantial differences in crude and risk
adjusted mortality rates between countries. When compared with the UK, the recorded
mortality rates for Poland, Latvia, Romania, and Ireland were higher even after
adjustment for all identified confounding variables. This pattern could relate to cultural,
demographic, socioeconomic, and political differences between nations, which might
affect population health and health-care outcomes.
A major strength of our study was the large number of consecutive unselected patients
enrolled in a multicentre and multinational setting. A vigorous approach to follow-up
for missing and incomplete data provided a high-quality dataset for analysis. The
dataset allowed us to explore probable prognostic factors and to adjust crude mortality
rates to describe differences in outcomes between countries. Our analysis identified
several factors associated with increased mortality. These findings suggest that surgery-
related and patient-related factors interact to increase mortality risk. Only two comorbid
disease categories were identified as independent variables. This finding probably arose
because the ASA score was designed to describe the severity of coexisting medical
disease.
Evidence suggests that critical-care-based cardiorespiratory interventions can improve

outcomes among high-risk surgical patients.17–21 However, in our study, only 5% of
patients underwent a planned admission to critical care with a median stay of about 1
day. Unplanned admissions to critical care were associated with higher mortality rates
than were planned admissions. Remarkably, most patients who died (73%) were not
admitted to critical care at any stage after surgery. Of patients who died after admission
to critical care, 43% did so after the initial episode was complete and the patient had
been discharged to a standard ward. These findings suggest a systematic failure in the
process of allocation of critical care resources. This notion is consistent with previous
reports of a failure to rescue deteriorating surgical patients with a detrimental effect on
patient outcomes22 and the high incidence of myocardial injury in the days after
surgery.23 For some patients with a poor prognosis, postoperative admission to critical
care might have been deemed inappropriate—eg, after palliative surgery for
disseminated malignancy. However, our data suggest these cases are few in number
(<5% of patients had malignancy, table 1). Meanwhile other investigators have
challenged the suggestion that patients should be offered surgery when the standard of
postoperative care is unlikely to be adequate for their needs.2 The low rate of admission
to critical care prevents any detailed comparison of this resource between nations.
Further research is needed to better understand whether early admission to critical care
can improve survival after major surgery.
Despite the large sample size, our study might not be truly representative of current
practice across Europe because only a small proportion of European hospitals took part.
Although in some countries the patient sample was large enough to show national
practice, the high proportion of patients enrolled in university hospitals in other
countries suggests a degree of selection bias. In particular, our data might not show the
true surgical case-mix and standards of care in countries with a small number of
participating hospitals. Although we planned to enrol every eligible patient undergoing
surgery during the study period, we cannot be sure of the exact proportion of eligible
patients included. Nonetheless, assuming the volume of surgery during the cohort week
is typical of the participating hospitals, these centres undertake more than 2·3 million
inpatient surgical procedures each year, which is 1% of the estimated volume of surgery
taking place worldwide.1 Whether truly representative or not, our findings clearly
describe a large cross-section of health care in Europe.
Some of our findings might be indicative of limitations of commonly used risk-

adjustment variables with unexpected patterns of survival across categories for both
ASA score and grade of surgery. This finding could result from the poor ability of
clinicians to discriminate between the less severe categories of these variables. Random
partitioning of the countries into three equal groups and repetition of the modelling
exercise showed much the same results with regards to the OR of the relevant effect
factors, showing some stability of the risk adjustment in subsets of countries. This
stability was further confirmed in more complex models that included interactions
between variables for which none of the interactions with the country factor contributed
significantly to prediction. We identified other interactions that did significantly
contribute to prediction but we did not record a substantial change in country effects
when estimated from the extended model including these interactions. We therefore
decided to use the simpler of the hierarchical models for the final analysis because our
aim had been to construct a parsimonious model that practising clinicians would easily
understand.
As far as we are aware, this was the first large, prospective, international assessment of
surgical outcomes (panel). In some countries, data are available that describe survival
after specific procedures such as vascular, joint replacement, or bowel cancer surgery.24–
26
However, these audits are poorly representative of overall national surgical
populations because high-risk patients are under-represented. The few previous
estimates suggest an overall mortality for unselected inpatient surgery of between 1%
and 2%,2,3,6,13–16 but these values are representative of only a few health-care systems. In
a previous study13 of national registry data from the Netherlands, 30 day mortality was
reported as 1·85%, which is much the same as the crude hospital mortality of 2% for
this country in the EuSOS study. In the UK, a prospective investigation2 with a very
similar methods to EuSOS identified a postoperative critical care admission rate of
6·7%, which is much the same as to the value of 6% for EuSOS in the UK.2 However,
30 day mortality was 1·6% compared with 3·6% for 60 day in-hospital mortality for UK
patients in EuSOS. Reports from nations outside Europe describe 30 day mortality rates
from 1·3% to 2·0%.6,14,15
Panel
Research in context
Systematic review
We searched Medline for original research from the past 10 years describing mortality
rates in large unselected national and international populations of patients undergoing
non-cardiac surgery. We used the search terms “surgery”, “mortality”, and
“complications” and widened our search to include retrospective analyses of health-care
registries and prospective epidemiological studies. Publications were screened by title
and then by abstract for relevance to the objectives of our study. Additionally,
coinvestigators in various European nations searched for publicly available registry
analyses reporting mortality rates for unselected populations of surgical patients. We
identified seven large national studies2,3,6,13–16 describing mortality rates for the
population of interest, three of which involved prospective data collection. No studies
were identified that provided international comparative data. The last search was done
on June 15, 2012.
Interpretation
As far as we are aware, this was the first large prospective international epidemiological
study of unselected non-cardiac surgical patients and as such it provides a new
perspective on mortality after surgery. A few national reports describe mortality rates
from 1·3% to 2·0%.2,3,6,13–16 In our study, the overall crude mortality rate of 4% was
higher than anticipated. We identified important variations in risk-adjusted mortality
rates between nations, and critical care resources did not seem to be allocated to patients
at greatest risk of death. Our findings raise important public health concerns about the
provision of care for patients undergoing surgery in Europe.
Previous investigators have described the differences in provision of health services

across Europe, in particular numbers of critical care beds.10,11 The reported seven-times
greater provision of critical care beds for Germany than for the UK is likely to affect
rates of admission to critical care and postoperative outcomes.10,11,27 This finding is in
keeping with our present data that show a greater rate of admission to critical care after
surgery in Germany than in the UK. Other studies have shown that fewer than a third of
high-risk non-cardiac surgical patients are admitted to critical care after surgery in the
UK despite high mortality rates,2–4 which is consistent with the results of our study;
across Europe 73% of surgical patients who died were never admitted to critical care.
This situation contrasts with perioperative care for cardiac surgical patients who by
definition have severe comorbid disease and undergo major body cavity surgery
followed by routine admission to critical care with mortality rates of less than 2%.28
Several reasons could explain why outcomes for cardiac and non-cardiac surgical
patients differ but the quality of perioperative care is likely to be among the most
important. The heath-care community increasingly recognises the importance of the
entire perioperative care pathway including preoperative assessment, optimisation of
coexisting medical disease, integrated care pathways relevant to the surgical procedure,
WHO surgical checklists, advanced haemodynamic monitoring during surgery, early
admission to critical care, acute pain management and critical-care outreach services,
and hospital discharge planning together with the primary care physician.20,21 Routine
audit and reporting of data for clinical outcomes has also proved a highly effective
instrument for improvement of the quality of perioperative care.29
Our findings suggest both the need and potential to implement measures to improve
postoperative outcomes. In addition to further research in this discipline, the root causes
of this problem could be better understood through increased use of high-quality
registries designed to capture robust data describing quality of care and clinical
outcomes for surgical patients. This step would require increased funding for this
specific area of health services research. The high mortality rate after surgery might be
modified by changes in the organisation of care.20
Go to:
Acknowledgments
This study was funded by the European Society of Intensive Care Medicine and the
European Society of Anaesthesiology. RP is a National Institute for Health Research
(UK) Clinician Scientist.
Go to:
Contributors
All authors were involved in the design and conduct of the study. Data collection and
collation was done by the members of the EuSOS study group. AR, RM, and PB did the
data analysis with input from RP. The report was drafted by RP and AR and revised
following critical review by all authors.
Go to:
Conflicts of interest
We declare that we have no conflicts of interest.
Go to:
Supplementary Material
Supplementary appendix:
Go to:
References
1. Weiser TG, Regenbogen SE, Thompson KD. An estimation of the global volume of
surgery: a modelling strategy based on available data. Lancet. 2008;372:139–144.
[PubMed: 18582931]
2. Findlay G, Goodwin A, Protopappa K, Smith N, Mason M. Knowing the risk: a
review of the peri-operative care of surgical patients. National Confidential Enquiry into
Patient Outcome and Death; London: 2011.
3. Pearse RM, Harrison DA, James P. Identification and characterisation of the high-
risk surgical population in the United Kingdom. Crit Care. 2006;10:R81. [PMCID:
PMC1550954] [PubMed: 16749940]
4. Jhanji S, Thomas B, Ely A, Watson D, Hinds CJ, Pearse RM. Mortality and
utilisation of critical care resources amongst high-risk surgical patients in a large NHS
trust. Anaesthesia. 2008;63:695–700. [PubMed: 18489613]
5. Khuri SF, Henderson WG, DePalma RG, Mosca C, Healey NA, Kumbhani DJ.
Determinants of long-term survival after major surgery and the adverse effect of
postoperative complications. Ann Surg. 2005;242:326–341. [PMCID: PMC1357741]
[PubMed: 16135919]
6. Jencks SF, Williams MV, Coleman EA. Rehospitalizations among patients in the
Medicare fee-for-service program. N Engl J Med. 2009;360:1418–1428. [PubMed:
19339721]
7. Head J, Ferrie JE, Alexanderson K, Westerlund H, Vahtera J, Kivimaki M.
Diagnosis-specific sickness absence as a predictor of mortality: the Whitehall II
prospective cohort study. BMJ. 2008;337:a1469. [PMCID: PMC2563263] [PubMed:
18832415]
8. Derogar M, Orsini N, Sadr-Azodi O, Lagergren P. Influence of major postoperative
complications on health-related quality of life among long-term survivors of esophageal
cancer surgery. J Clin Oncol. 2012;30:1615–1619. [PubMed: 22473157]
9. Ghaferi AA, Birkmeyer JD, Dimick JB. Variation in hospital mortality associated
with inpatient surgery. N Engl J Med. 2009;361:1368–1375. [PubMed: 19797283]
10. Wunsch H, Angus DC, Harrison DA. Variation in critical care services across North
America and Western Europe. Crit Care Med. 2008;36:2787–2793. [PubMed:
18766102]
11. Rhodes A, Ferdinande P, Flaatten H, Guidet B, Metnitz PG, Moreno RP. The
variability of critical care bed numbers in Europe. Intensive Care Med. 2012 published
online July 10.
12. Metnitz PG, Moreno RP, Almeida E. SAPS 3—from evaluation of the patient to
evaluation of the intensive care unit. Part 1: objectives, methods and cohort description.
Intensive Care Med. 2005;31:1336–1344. [PMCID: PMC1315314] [PubMed:
16132893]
13. Noordzij PG, Poldermans D, Schouten O, Bax JJ, Schreiner FA, Boersma E.
Postoperative mortality in The Netherlands: a population-based analysis of surgery-
specific risk in adults. Anesthesiology. 2010;112:1105–1115. [PubMed: 20418691]
14. Yu PC, Calderaro D, Gualandro DM. Non-cardiac surgery in developing countries:
epidemiological aspects and economical opportunities—the case of Brazil. PLoS One.
2010;5:e10607. [PMCID: PMC2868901] [PubMed: 20485549]
15. Glance LG, Lustik SJ, Hannan EL. The surgical mortality probability model:
derivation and validation of a simple risk prediction rule for noncardiac surgery. Ann
Surg. 2012;255:696–702. [PubMed: 22418007]
16. Canet J, Gallart L, Gomar C. Prediction of postoperative pulmonary complications
in a population-based surgical cohort. Anesthesiology. 2010;113:1338–1350. [PubMed:
21045639]
17. Hamilton MA, Cecconi M, Rhodes A. A systematic review and meta-analysis on the
use of preemptive hemodynamic intervention to improve postoperative outcomes in
moderate and high-risk surgical patients. Anesth Analg. 2011;112:1392–1402.
[PubMed: 20966436]
18. Squadrone V, Coha M, Cerutti E. Continuous positive airway pressure for treatment
of postoperative hypoxemia: a randomized controlled trial. JAMA. 2005;293:589–595.
[PubMed: 15687314]
19. Pearse RM, Belsey JD, Cole JN, Bennett ED. Effect of dopexamine infusion on
mortality following major surgery: individual patient data meta-regression analysis of
published clinical trials. Crit Care Med. 2008;36:1323–1329. [PubMed: 18379261]
20. Pearse RM, Holt PJ, Grocott MP. Managing perioperative risk in patients
undergoing elective non-cardiac surgery. BMJ. 2011;343:d5759. [PubMed: 21976704]
21. Pearse RM, Ackland GL. Perioperative fluid therapy. BMJ. 2012;344:e2865.
[PubMed: 22539015]
22. Ghaferi AA, Birkmeyer JD, Dimick JB. Complications, failure to rescue, and
mortality with major inpatient surgery in medicare patients. Ann Surg. 2009;250:1029–
1034. [PubMed: 19953723]
23. Devereaux PJ, Chan MT, Alonso-Coello P. Association between postoperative
troponin levels and 30-day mortality among patients undergoing noncardiac surgery.
JAMA. 2012;307:2295–2304. [PubMed: 22706835]
24. Clement ND, Jenkins PJ, Brenkel IJ, Walmsley P. Predictors of mortality after total
knee replacement: a ten-year survivorship analysis. J Bone Joint Surg Br. 2012;94:200–
204. [PubMed: 22323686]
25. Bilimoria KY, Bentrem DJ, Feinglass JM. Directing surgical quality improvement
initiatives: comparison of perioperative mortality and long-term survival for cancer
surgery. J Clin Oncol. 2008;26:4626–4633. [PubMed: 18574159]
26. Finks JF, Osborne NH, Birkmeyer JD. Trends in hospital volume and operative
mortality for high-risk surgery. N Engl J Med. 2011;364:2128–2137. [PMCID:
PMC3150488] [PubMed: 21631325]
27. Adhikari NK, Fowler RA, Bhagwanjee S, Rubenfeld GD. Critical care and the
global burden of critical illness in adults. Lancet. 2010;376:1339–1346. [PubMed:
20934212]
28. Care quality commission. Heart surgery in the United Kingdom. 2012.
http://heartsurgery.cqc.org.uk/Survival.aspx (accessed June 6, 2012).
29. Rowell KS, Turrentine FE, Hutter MM, Khuri SF, Henderson WG. Use of national
surgical quality improvement program data as a catalyst for quality improvement. J Am
Coll Surg. 2007;204:1293–1300. [PubMed: 17544087]
Go to:
Figures and Tables

Figure 1
Study profile
(A) All patients. (B) Patients admitted to critical care. CRF=case report form.
Figure 2
Planned and unplanned admission to a critical-care unit according to urgency of surgery
Data are n (%) or median (IQR). We collected data describing the first critical care
admission for any individual patient. The data presented do not describe readmission to
critical care. Because of incomplete data for admission planning, 19 admissions to
critical care are not presented in this figure. EuSOS=European Surgical Outcomes
Study. Elective=not immediately life saving; planned within months or weeks.
Urgent=planned surgery within hours or days of the decision to operate. Emergency=as
soon as possible; no delay to plan care; ideally within 24 h.
Figure 3
Adjusted odds ratio for death in hospital after surgery for each country
Table 1
Description of cohort
Survived to
Died in
All patients hospital Odds ratio
hospital p value
(n=46 539) discharge (95% CI)
(n=1864)
(n=44 657)
1·01 (1·01–
Age (years) 56·7 (18·5) 61·0 (18·7) 56·6 (18·5) <0·0001
1·02)
1·15 (1·05–
Men 22 607 968 21 629 0·003
1·26)
0·90 (0·80–
Present smoker 9872 363 9503 0·07
-1·01)
ASA score
1 11 642 362 11 280 Reference ··
Survived to
Died in
hospital p value
(n=1864)
(n=44 657)
0·94 (0·83–
2 21 582 633 20 944 0·36
1·07)
1·51 (1·32–
3 11 574 539 11 025 <0·0001
1·73)
6·75 (5·71–
4 1559 279 1277 <0·0001
7·97)
35·61
5 90 49 41 (23·23– <0·0001
54·59)
Grade of surgery
Minor 12 041 431 11 608 Reference ··
0·93 (0·82–
Intermediate 22 231 741 21 483 0·22
1·05)
1·59 (1·40–
Major 12 170 685 11 476 <0·0001
1·80)
Urgency of
surgery
Elective 35 049 1129 33 908 Reference ··
1·71 (1·52–
Urgent 8923 483 8436 <0·0001
1·91)
3·20 (2·77–
Emergency 2557 249 2303 <0·0001
3·70)
Surgical specialty
1·02 (0·84–
Orthopaedics 12 214 468 11 744 0·85
1·24)
0·76 (0·53–
Breast 1500 43 1456 0·12
1·07)
0·76 (0·59–
Gynaecology 3972 115 3857 0·04
0·99)
1·61 (1·26–
Vascular 2376 140 2233 0·0001
2·05)
Upper 1·88 (1·48–
2228 155 2071 0·0001
gastrointestinal 2·39)
Lower 1·54 (1·25–
4972 284 4683 0·0001
gastrointestinal 1·91)
1·35 (1·04–
Hepato-biliary 2247 113 2134 0·025
1·74)
Plastic or 0·79 (0·59–
2432 73 2356 0·12
cutaneous 1·06)
0·78 (0·61–
Urology 4881 144 4737 0·042
0·99)
Survived to
Died in
hospital p value
(n=1864)
(n=44 657)
0·51 (0·26–
Kidney 463 9 454 0·05
1·01)
0·82 (0·65–
Head and neck 5640 174 5466 0·09
1·03)
Other 3463 132 3329 Reference
Laparoscopic 0·69 (0·59–
5510 160 5350 <0·0001
surgery 0·82)
Comorbid
disorder
3·64 (2·79–
Cirrhosis 498 65 433 <0·0001
4·76)
Congestive heart 2·10 (1·78–
2154 166 1985 <0·0001
failure 2·48)
1·21 (1·05–
COPD 5162 244 4912 0·008
2·48)
Coronary artery 1·73 (1·54–
6274 387 5881 <0·0001
disease 1·94)
Diabetes (taking 1·73 (1·44–
2081 135 1939 <0·0001
insulin) 2·07)
Diabetes (not 1·05 (0·88–
3495 147 3348 0·61
taking insulin) 1·24)
1·91 (1·61–
Metastatic cancer 2173 155 2017 <0·0001
2·27)
1·57 (1·30–
Stroke 2006 120 1884 <0·0001
1·90)
Data are mean (SD) or n unless otherwise specified. Odds ratios were constructed for
in-hospital mortality with univariate binary logistic regression analysis. ASA=American
Society of Anesthesiologists. COPD=chronic obstructive pulmonary disease.
Table 2
Relation between country and in-hospital mortality
Percenta
Media Numbe ge Percenta
Numb
Numb n days r admitted ge died Adjust
er Unadjust
er of in admitt to in ed OR p
died in ed OR
patien hospit ed to critical hospital (95% value
hospit (95% CI)
ts al critical care (95% CI)
al
(IQR) care (95% CI)
CI)
Percenta
Numb
er Unadjust
died in ed OR
hospit (95% CI)
al
(IQR) care (95% CI)
CI)
3·0 9·2% 0·89 1·65
3·2%
Belgium 1486 (1·0– 136 (7·7– 47 (0·65– (0·81– 0·17
(2·3–4·1)
6·0) 10·6) 1·21) 3·40)
4·0 9·4% 2·17 1·89
7·4%
Croatia 1767 (2·0– 166 (8·0– 131 (1·77– (0·94– 0·07
(6·2–8·6)
7·0) 10·8) 2·67) 3·80)
1·0 0·62 0·82
2·2%
Cyprus 45 (1·0– 0 0 1 (0·09– (0·04– 0·90
(0·0–6·7)
3·0) 4·48) 16·70)
4·0 0·64 1·30
Czech 4·8% 2·3%
434 (2·0– 21 10 (0·34– (0·23– 0·77
Republic (2·8–6·9) (0·9–3·7)
9·0) 1·21) 7·46)
2·0 0·90 1·16
3·6% 3·2%
Denmark 1000 (1·0– 36 32 (0·62– (0·52– 0·72
(2·4–4·8) (2·1–4·3)
5·0) 1·29) 2·61)
3·0 0·42 0·60
7·0% 1·5%
Estonia 727 (1·0– 51 11 (0·23– (0·16– 0·45
(5·2–8·9) (0·6–2·4)
6·0) 0·76) 2·28)
2·0 0·54 0·44
4·0% 2·0%
Finland 1071 (1·0– 43 21 (0·35– (0·19– 0·06
(2·8–5·6) (1·1–2·8)
5·0) 0·85) 1·05)
3·0 0·90 1·36
5·8% 3·2%
France 2278 (1·0– 132 73 (0·70– (0·72– 0·34
(4·8–6·8) (2·5–3·9)
6·0) 1·16) 2·56)
4·0 11·6% 0·70 0·85
2·5%
Germany 5284 (2·0– 611 (10·7– 133 (0·57– (0·50– 0·54
(2·1–2·9)
9·0) 12·4) 0·86) 1·43)
3·0 1·01 1·20
3·5% 3·6%
Greece 1803 (2·0– 63 65 (0·78– (0·66– 0·55
(2·7–4·3) (2·7–4·5)
7·0) 1·33) 2·16)
4·0 0·90 1·23
7·1% 3·2%
Hungary 621 (2·0– 44 20 (0·57– (0·43– 0·69
(5·1–9·1) (1·8–4·6)
7·0) 1·43) 3·50)
2·0 9·3% 0·34 0·47
1·2%
Iceland 162 (1·0– 15 (4·8– 2 (0·08– (0·07– 0·46
(0·0–3·0)
4·0) 13·8) 1·37) 3·41)
3·0 1·86 2·61
7·7% 6·4%
Ireland 856 (1·0– 66 55 (1·39– (1·30– 0·007
(5·9–9·5) (4·8–8·1)
6·0) 2·49) 5·27)
Percenta
Numb
er Unadjust
died in ed OR
hospit (95% CI)
al
(IQR) care (95% CI)
CI)
3·0 1·51 1·70
7·5% 5·3%
Italy 2673 (2·0– 200 141 (1·24– (0·97– 0·06
(6·5–8·5) (4·4–6·1)
7·0) 1·84) 2·97)
4·0 21·5% 7·44 4·98
6·3%
Latvia 302 (2·0– 19 65 (16·9– (5·55– (1·22– 0·025
(3·5–9·1)
8·0) 26·2) 9·97) 20·29)
3·0 0·74 1·21
3·7% 2·7%
Lithuania 375 (2·0– 14 10 (0·39– (0·21– 0·83
(1·8–5·7) (1·0–4·3)
5·0) 1·40) 6·95)
3·0 0·55 0·63
Netherlan 7·7% 2·0%
1627 (1·0– 126 32 (0·38– (0·28– 0·26
ds (6·4–9·0) (1·3–2·7)
6·0) 0·78) 1·41)
3·0 0·40 0·51
4·5% 1·5%
Norway 689 (1·0– 31 10 (0·21– (0·17– 0·22
(3·0–6·1) (0·6–2·4)
6·0) 0·75) 1·49)
5·0 17·9% 5·91 6·92
2·0% 0·000
Poland 397 (2·0– 8 71 (14·1– (4·48– (2·37–
(0·6–3·4) 4
7·5) 21·7) 7·79) 20·27)
3·0 1·16 1·43
6·9% 4·1%
Portugal 1489 (1·0– 103 61 (0·88– (0·72– 0·31
(5·6–8·2) (3·1–5·1)
7·0) 1·53) 2·83)
5·0 16·1% 1·97 3·19
6·8%
Romania 1298 (3·0– 209 (14·1– 88 (1·55– (1·61– 0·001
(5·4–8·2)
8·0) 18·1) 2·51) 6·29)
5·0 0·65 1·06
1·2% 2·4%
Serbia 85 (3·0– 1 2 (0·16– (0·11– 0·96
(0·0–3·5) (0·0–5·6)
7·0) 2·67) 10·04)
3·0 11·2% 3·41 2·15
1·9%
Slovakia 1156 (2·0– 22 129 (9·3– (2·76– (0·91– 0·08
(1·1–2·7)
7·0) 13·0) 4·20) 5·07)
3·0 0·81 1·12
2·5% 2·9%
Slovenia 518 (1·0– 13 15 (0·48– (0·30– 0·86
(1·2–3·9) (1·5–4·3)
7·0) 1·37) 4·22)
3·0 12·5% 1·08 1·39
3·8%
Spain 5433 (1·0– 677 (11·6– 208 (0·91– (0·89– 0·15
(3·3–4·3)
7·0) 13·3) 1·28) 2·18)
2·0 0·50 0·58
3·2% 1·8%
Sweden 1314 (1·0– 42 24 (0·33– (0·23– 0·26
(2·2–4·2) (1·1–2·6)
6·0) 0·77) 1·49)
Percenta
Numb
er Unadjust
died in ed OR
hospit (95% CI)
al
(IQR) care (95% CI)
CI)
4·0 0·54 0·86
Switzerla 7·8% 2·0%
1019 (2·0– 79 20 (0·35– (0·25– 0·81
nd (6·1–9·4) (1·1–2·8)
8·0) 0·86) 2·97)
2·0
10 6·3% 3·6%
UK (1·0– 671 378 1·00 ·· ··
630 (5·9–6·8) (3·2–3·9)
6·0)
Odds ratios (OR) referenced against the UK and adjusted for age, American Society of
Anesthesiologists' score, urgency of surgery, grade of surgery (minor, intermediate,
major), surgical specialty, and the presence of either metastatic disease or cirrhosis in a
two-level binary logistic regression model (with patient at the first level and hospital at
the second).
The use of inhibitors of angiotensin-converting

enzyme and its relation to events in the
postoperative period of CABG
O uso de inibidores da enzima conversora de angiotensina e sua relação com eventos no pós-operatório
de cirurgia de revascularização miocárdica
Graciane Radaelli1, Luiz Carlos Bodanese2, João Carlos Vieira da Costa Guaragna3, Anibal Pires
Borges4, Marco Antonio Goldani5, João Batista Petracco6, Jacqueline da Costa Escobar Piccoli7,
Luciano Cabral Albuquerque8
DOI: 10.5935/1678-9741.20110011 RBCCV 44205-1292
Abstract Clinical outcomes after surgery were analyzed between

Background: Angiotensin-converting enzyme (ACE) users and nonusers of ACE inhibitors preoperatively.
inhibitors reduce the chance of death, myocardial infarction Results: Fifty-two percent (n=1,635) of patients received
(MI) and cerebrovascular accident (CVA) in patients with ACE inhibitors preoperatively. The use of ACE inhibitors
coronary disease. However there is no consensus as to its was an independent predictor of need for inotropic support
indication in patients undergoing coronary artery bypass (OR 1.24, 95% CI 1.01 to 1.47, P = 0.01), acute renal failure
grafting (CABG). (OR 1.23, 95% CI 1.01 to 1.73, P = 0.04) and progression to
Objective: To assess the relationship between preoperative atrial fibrillation (OR 1.32, 95% CI 1.02 to 1.7, P = 0.03)
use of ACE inhibitors and clinical outcomes after CABG. postoperatively. The mortality rate among patients receiving
Methods: Retrospective cohort study. We included data or not preoperative ACE inhibitors was similar (10.3% vs.
from 3,139 consecutive patients undergoing isolated CABG 9.4%, P = 0.436), as well as the incidence of myocardial
in Brazilian tertiary care hospital between January 1996 infarction and stroke (15.6% vs. 15.0%, P = 0.694 and 3.4%
and December 2009. Follow-up was until discharge or death. vs. 3.5%, P = 0.963, respectively).
1. Master; Pharmaceuticals. 8. Doctor of Health Sciences, Federal University of Rio Grande do

2. PhD in Cardiology by the Institute of Cardiology of Rio Grande Sul, 2006, Cardiovascular Surgeon at HSL-PUCRS, Porto Alegre,
do Sul / Baylor College of Medicine, Houston / Texas, Professor Brazil.
of Postgraduate Medicine and Health Sciences at the Pontificia
Universidade Catolica do Rio Grande do Sul (PUCRS), Porto
Alegre, Brazil. Work performed at the Pontificia Universidade Catolica do Rio Grande
3. Doctor of Medicine from the Pontificia Universidade Catolica do Sul, Porto Alegre, Brazil.
do Rio Grande do Sul, 2009; Head of Postoperative Cardiac Surgery
HSL-PUCRS, Porto Alegre, Brazil.
4. Medical Resident of the Residency Program in Cardiology, Correspondence Address:
Hospital São Lucas da PUCRS, Porto Alegre, Brazil. Graciane Radaelli
5. Specialist in Cardiovascular Surgery, Chief of Cardiovascular Avenida Ipiranga, 6690 - sala 300 – Serviço de Cardiologia
Surgery of the HSL-PUCRS, Porto Alegre, Brazil. Porto Alegre, Brazil
6. Master of Cardiovascular Surgery, Chief of Cardiac Surgery of the E-mail: gracianeradaelli@yahoo.com.br
HSL-PUCRS, Porto Alegre, Brazil.
7. PhD in Cell and Molecular Biology PUCRS, Adjunct Professor,
Universidade Federal do Pampa - Campus Uruguaiana, Uruguaiana, Article received on May 12th, 2011
RS, Brazil. Article accepted on July 14th, 2011
373
Radaelli G, et al. - The use of inhibitors of angiotensin-converting Rev Bras Cir Cardiovasc 2011;26(3):373-9
enzyme and its relation to events in the postoperative period of
CABG
Conclusion: The use of preoperative ACE inhibitors was 1996 e dezembro de 2009. O seguimento dos pacientes foi
associated with increased need for inotropic support and realizado até a alta hospitalar ou óbito. Desfechos clínicos
higher incidence of acute renal failure and postoperative no pós-operatório foram analisados entre os usuários e os
atrial fibrillation, not associated with increased rates of não-usuários de IECA no pré-operatório.
myocardial infarction, stroke or death. Resultados: Cinquenta e dois por cento (1.635) dos
pacientes receberam IECA no pré-operatório. O uso de IECA
Descriptors: Angiotensin-converting enzyme inhibitors. foi preditor independente da necessidade de suporte
Myocardial revascularization. Coronary disease. Coronary inotrópico (RC 1,24, IC 1,01-1,47; P=0,01), de insuficiência
artery bypass. renal aguda (IRA, RC 1,23, IC 1,01-1,73; P=0,04) e de evolução
para fibrilação atrial (FA, RC 1,32, IC 1,02-1,7; P=0,03) no
pós-operatório. A mortalidade entre os pacientes que
Resumo receberam ou não IECA no pré-operatório foi semelhante
Fundamento: Os inibidores da enzima conversora de (10,3 vs. 9,4%, P=0,436), bem como a incidência de IAM e
angiotensina (IECA) reduzem o risco de óbito, infarto agudo AVE (15,6 vs. 15,0%, P=0,694 e 3,4 vs. 3,5%, P=0,963,
do miocárdio (IAM) e acidente vascular encefálico (AVE) em respectivamente).
portadores de doença coronariana. No entanto, não há Conclusão: O uso pré-operatório de IECA foi associado a
consenso quanto à sua indicação em pacientes que serão maior necessidade de suporte inotrópico e maior incidência
submetidos à cirurgia de revascularização miocárdica (CRM). de IRA e FA no pós-operatório, não estando associado ao
Objetivo: Avaliar a relação entre uso pré-operatório de aumento das taxas de IAM, AVE ou óbito.
IECA e eventos clínicos após realização da CRM.
Métodos: Estudo de coorte retrospectivo. Foram incluídos Descritores: Inibidores da enzima conversora da
dados de 3.139 pacientes consecutivos submetidos à CRM angiotensina. Revascularização miocárdica. Doença das
isolada em hospital terciário brasileiro, entre janeiro de coronárias. Ponte de artéria coronária.
INTRODUCTION inhibitors act as anti-ischemic, through their protective

effects of the vascular bed (antiatherosclerotic,
Inhibitors of angiotensin-converting enzyme (ACE) antithrombotic and anti-inflammatory) [6,7].
inhibitors have been shown to be beneficial in preventing There is growing evidence that ACE inhibitors should
death, myocardial infarction (AMI) and stroke (CVA) in be used in all patients undergoing CABG. The Heart
patients with coronary artery disease (CAD). They additional Outcomes Prevention Evaluation (HOPE) [5] demonstrated
effects of ACE inhibitors are the control of high blood that therapy with ACE inhibitors benefits all patients with
pressure (HBP) and the reduction of morbidity and mortality risk factors for cardiovascular disease as well as patients
in heart failure also, being a mainstay in the treatment of this undergoing CABG. In the perioperative period, ACE
disease [1-3]. In addition, they play an important role in inhibitors reduce inflammation and preserve endothelial
minimizing the occurrence of ischemic events after coronary function, improving long-term graft patency and minimizing
artery bypass grafting (CABG) [4]. the progression of existing atherosclerotic plaques.
The ACE inhibitors should be given early to all patients There are still controversies over the preoperative use
with ventricular dysfunction or myocardial infarction, for of ACE inhibitors in patients undergoing CABG. Studies
the advantages of these drugs in relation to cardiac suggest that the preoperative administration of ACE
remodeling and the hemodynamics improvement inhibitors in cases of CABG helps reduce systemic vascular
(vasodilatation and afterload reduction). The decrease in resistance and the development of vasoplegia in the early
mortality was observed either in clinical studies that postoperative phase, resulting in hypotension and renal
selected patients with AMI and concomitant ventricular dysfunction [8,9]. Other authors [10,11] suggest that
dysfunction (studies SAVE, AIRE and TRACE) as in studies preoperative use of ACE inhibitors does not cause
that did not select a specific group, using ACE inhibitors in hypotension and can be safely used in patients undergoing
all patients with AMI for up to 1-4 years after the event cardiac surgery.
(studies ISIS 4 and GISSI 3). Prolonged use of ACE The aim of this study was therefore to assess the
inhibitors also resulted in cardiovascular benefits (HOPE relationship between the preoperative use of ACE inhibitors
study) [5]. In addition to lowering blood pressure, ACE and clinical events after the completion of CABG.
374
CABG
METHODS performed during the same hospital) or elective (the

patient’s clinical status allowed readmission to hospital later
We evaluated data from consecutive patients to perform CABG).
undergoing CABG at the Hospital São Lucas, Pontificia As for the design, this was a retrospective cohort study.
Universidade Catolica do Rio Grande do Sul (HSL-PUCRS) The present study was assessed by the Research Ethics
in Porto Alegre, Rio Grande do Sul, Brazil, between January Committee of the Pontificia Universidade Catolica do Rio
1996 and December 2009. We excluded those who had no Grande do Sul (Registration 09/04811). Personal information
information about the use or not preoperative ACE obtained was kept confidential data being used exclusively
inhibitors or ARA2, and those undergoing concomitant for research. Because it is an observational study which
valve replacement. In total, 3,139 patients met the inclusion evaluated only clinical and laboratory data of patients with
criteria. Follow-up was during the period of hospitalization, no postoperative intervention or implication about the care
or until hospital discharge or death. provided, it was not applied the Informed Consent Form.
After surgery, patients were admitted to the unit of
Postoperative Cardiac Surgery (POCC) of the hospital. The Statistical analysis
information was taken from structured protocol completed Continuous variables are described as mean ± standard
by doctors or nurses of the unit in the pre-operative, deviation and categorical variables are described in
perioperative and postoperative. The ejection fraction was percentages. The association between preoperative use of
measured by echocardiogram or radiocardiogram during ACE inhibitors and clinical outcomes after surgery was
hospitalization, and this data is available for more than 95% evaluated by Student t test or chi-square as appropriate.
of patients. Patients were included in the ACE inhibitor Multivariate logistic regression analysis was performed to
group should be taking the drug or an angiotensin receptor identify independent determinants of outcomes, and
antagonist 2 (ARA2) for at least 2 weeks. included those with P <0.2. The results are presented in
The primary outcome assessed was the use of inotropic percentages and odds ratio (OR) with confidence interval
support after surgery, defined by need for vasopressor 95% (CI). The level of significance alpha was 0.05.
administration in the immediate postoperative period (24 to All statistical analysis was performed using the
48 hours). Secondary outcomes assessed were: acute renal Statistical Package for the Social Sciences (SPSS) for
failure (ARF), atrial fibrillation (AF), myocardial infarction, Windows, version 12.0.
stroke and death. The preoperative use of ACE inhibitors
or ARA2 was defined as the administration of medication RESULTS
in the same period to 24 hours before CABG. All outcomes,
except for death were evaluated in the immediate Patient characteristics
postoperative period. We evaluated 3,139 patients, being that in 1,635 (52.1%)
The diagnosis of ARF in the postoperative period was we used ACE inhibitors or ARA2 in the pre-operatively
defined by an increase greater than or equal to 50% or (ACE inhibitors group) and in 1,504 (47.9%) we did not.
serum creatinine greater than 0.5 mg/dL above the Baseline characteristics of patients investigated are shown
preoperative value. Postoperative AF was defined as the in Table 1. Apparently the ACE inhibitors group patients
presence of AF of any duration observed in 12-lead were more serious, because they had more comorbidities
electrocardiogram. The diagnosis of postoperative AMI that were related to increased surgical risk as risk score of
was based on the current presence of subepicardial injury CABG recently published by our group [13]. The patients
and the onset of Q wave, current of subendocardial injury in the ACE inhibitor group had a higher number of females,
with increased necrosis markers (troponin I or CK-MB), higher prevalence of advanced heart failure (functional
bundle branch block with new markers also high. Aiming to classes III and IV of the New York Heart Association) and
exclude the elevation of markers secondary to the systolic dysfunction. There was a tendency to increased
procedure, it was only considered the increase of CK-MB number of individuals with chronic AF. There was no
of at least 5 times the reference value or greater than 10% of difference between groups regarding age and the
total CK or troponin I> 10 µg / dL after 12 hours of prevalence of renal disease (ESRD) and chronic obstructive
postoperative values as established by Nascente et al. [12] pulmonary disease (COPD). There was less need for
in the same population. emergency surgery / emergency in the ACE inhibitor group.
On the other hand, the diagnosis of stroke was defined
as new neurological deficit consistent with findings on Outcomes
imaging (CT or MRI of the brain). The data were analyzed The incidence of events evaluated in the post-CABG in
according to surgical priority: emergency (CABG should patients who used and did not use ACE inhibitors in the
be done within a few hours), urgency (CABG should be preoperative period are shown in Figure 1. There was no
375
CABG
Table 1. Baseline characteristics of patients in the preoperative period.

Using ACE Not Using ACE P
Variables inhibitors inhibitors
(n=1635) (%=52.1) (n =1504) (%=47.9)
Age, mean (SD), years 61.21(10.11) 61.60 (10.35) 0.294
Female 585 (35.8) 467 (31.1) 0.006
Hypertension 1361 (83.2) 916 (60.9) <0.001
Smoking 507 (31.0) 574 (38.2) <0.001
Diabetes mellitus 560 (34.3) 369 (24.5) <0.001
Unstable angina 571 (34.9) 575 (38.2) 0.059
NYHA III and IV 303 (18.5) 180 (12.0) <0.001
Previous AMI 766 (46.9) 544 (36.2) <0.001
Ejection fractio
856 (52.6) 995 (66.5)
n>50 % <0.001
609 (37.5) 429 (28.7)
30-50%
161 (9.9) 73 (4.9)
<30%
History of AF 67 (4.1) 42 (2.8) 0.058
COPD 296 (18.1) 297 (19.7) 0.259
IRC 158 (9.7) 133 (8.8) 0.465
Previous heart surgery 60 (3.7) 43 (2.9) 0.241
Emergency surgery 83 (5.1) 147 (9.8) <0.001
Use of intra-aortic balloon 176 (10.8) 115 (7.6) 0.003
ACE: angiotensin-converting enzyme inhibitors; NYHA: classification of heart failure functional class according to
the New York Heart Association, AMI: acute myocardial infarction, AF: atrial fibrillation, COPD: chronic obstructive
pulmonary disease, ARF: acute renal failure
Table 2. Multivariate analysis of predictors of inotropic support

in post-CABG
Variables RC IC 95% P
Preoperative use of ACE inhibitors 1.24 1.04-1.47 0.015
Age 1.03 1.02-1.04 <0.001
NYHA III and IV 1.53 1.22-1.91 <0.001
Emergency surgery 2.02 1.49-2.72 <0.001
Use of intra-aortic balloon 1.51 1.15-1.99 0.003
IRC 1.61 1.23-2.09 <0.001
AF 1.88 1.26-2.83 0.002
Previous AMI 1.27 1.07-1.51 0.006
Previous heart surgery 1.67 1.09-2.55 0.018
ACE: angiotensin-converting enzyme inhibitors; NYHA:
classification of heart failure functional class according to the New
York Heart Association, AMI: acute myocardial infarction, AF:
Fig. 1 - Clinical outcomes after CABG by univariate chi-square. atrial fibrillation, COPD: chronic obstructive pulmonary disease,
ACEI: angiotensin-converting enzyme inhibitors ARF: acute renal failure
AF: atrial fibrillation
ARF: acute renal failure
AMI: acute myocardial infarction
376
CABG
difference in mortality between the groups (10.3% of deaths The preoperative use of ACE inhibitors did not increase
in the group who used ACE inhibitors vs. 9.4% in the group the risk of AMI, stroke or death in post-CABG.
who did not use ACE inhibitors, P = 0.436).
DISCUSSION
Inotropic support
The use of ACE inhibitors was the most used Some studies have shown that blocking the renin-
independent predictor of inotropic support in the angiotensin system (RAS) with ACE inhibitors improves
postoperative period (OR 1.24, 95% CI 1.04 to 1.47, P = ventricular function, prolongs survival and decreases the
0.015). Advanced age, heart failure functional class III or IV size of the infarct in patients after AMI [14] or patients with
New York Heart Association (NYHA), urgent or emergency heart failure [15]. However, it is unclear the role of ACE
surgery, need for intra-aortic balloon and previous history inhibitors in patients undergoing CABG. Surgeons
of chronic renal failure, atrial fibrillation, myocardial attributed the beneficial effects of ACE inhibitors to their
infarction or cardiac surgery were also predictors of the antihypertensive and antiatherogenic properties. Lazar et
need inotropic support after the procedure (Table 2). al. [16] concluded that all patients undergoing CABG
should receive ACE inhibitors preoperatively.
ARF The results of this study suggest that preoperative use
Patients on ACE inhibitors had a higher risk of developing of ACE inhibitors as an independent predictor of the need
postoperative ARF (OR 1.23, 95% CI 1.01 to 1.73, P = 0.042), for inotropic support postoperatively, consistent with
as well as elderly patients, patients with chronic obstructive previously published data [17-20].
pulmonary disease (COPD), heart failure, or CRF basal Perioperative hypotension, for generating a reduction
functional class III or IV heart failure patients, and those in renal perfusion pressure is a known risk factor for the
who required urgent or emergency surgery (Table 3). development of ARF in surgical patients. However, the
association between therapy with ACE inhibitors and renal
Postoperative AF failure after cardiac surgery remains controversial [21]. The
The use of ACE inhibitors, after adjusting for age, was a effect of ACE inhibitors on renal function after surgery
predictor for the development of postoperative AF (OR 1.32, may depend on the prior exposure time. Rady & Ryan [11]
95% CI 1.02 to 1.7, P = 0.032). In addition to age, COPD patients found no significant association between use of ACE
preoperatively also had a higher risk of AF (Table 4). inhibitors and renal failure in post-cardiac surgery in patients
chronically treated with medication. In contrast, Arora et
al. [22] in a large observational study, showed significant
Table 3. Multivariate analysis of predictors of renal dysfunction
association between preoperative use of ACE inhibitors
after CABG.
Variables RC IC 95% P and acute renal failure in the postoperative period of cardiac
Preoperative use of ACE inhibitors 1.23 1.01-1.73 0.042 surgery and abdominal aortic surgery [23]. In addition, there
Age 1.06 1.05-1.08 <0.001 are reports that treatment with ACE inhibitors may increase
COPD 1.7 1.26-2.29 <0.001 the adverse effects during the first 3 months after surgery
IRC 3.8 2.76-5.24 <0.001 did not improve clinical outcome up to 3 years of follow up
NYHA III and IV 1.85 1.37-2.51 <0.001 [24]. On the other hand, the study APRES [25] showed that
Emergency surgery 3.08 2.13-4.46 <0.001 the use of long-term ramipril reduced the composite
ACE: angiotensin-converting enzyme inhibitors; NYHA: classification endpoint of cardiac death, myocardial infarction and heart
of heart failure functional class according to the New York Heart failure in clinical treatment with revascularization. Our study
Association, AMI: acute myocardial infarction, AF: atrial fibrillation, suggested that use of ACE inhibitors in the pre-CABG
COPD: chronic obstructive pulmonary disease, ARF: acute renal failure increases the risk of ARF. As the follow-up occurred during
the hospital stay, not have time to test whether there was a
potential benefit of treatment as clinically important events.
Table 4. Multivariate analysis of predictors of atrial fibrillation
A study by White et al. [26] in patients undergoing
post-CABG.
Variables RC IC 95% P cardiac surgery (CABG and valve) showed a statistically
Preoperative use of ACE inhibitors 1.32 1.02-1.7 0.032 significant association between the use of ACE inhibitors
Age 1.07 1.06-1.09 <0.001 before surgery and reduce postoperative AF. Although we
COPD 1.8 1.36-2.39 <0.001 only evaluated patients undergoing CABG, our results are
ACE: angiotensin-converting enzyme inhibitors; NYHA: classification contradictory to the study cited, since there was an
of heart failure functional class according to the New York Heart increased chance of developing postoperative AF. The
Association, AMI: acute myocardial infarction, AF: atrial fibrillation, blockade of the RAS in patients undergoing CABG
COPD: chronic obstructive pulmonary disease, ARF: acute renal failure contributes to the reduction of systemic vascular resistance
377
CABG
and the vasoplegia in the immediate postoperative period, Outcomes Prevention Evaluation Study Investigators. N Engl
resulting in hypotension, which often requires volume and J Med. 2000;342(3):145-53.
/ or vasoactive drugs [17-20]. It is known that both the
hypotension and volume overload are factors that 2. EURopean trial On reduction of cardiac events with Perindopril
in stable coronary Artery disease Investigators. Efficacy of
contribute to the development of AF postoperatively [27].
perindopril in reduction of cardiovascular events among patients
Our study showed that preoperative use of ACE with stable coronary artery disease: randomised, double-blind,
inhibitors did not increase the risk of AMI, stroke or death placebo-controlled, multicentre trial (the EUROPA study).
in post-CABG. In contrast, Miceli et al. [28] demonstrated Lancet. 2003;362(9386):782-8.
that preoperative therapy with ACE inhibitors has increased
at twice the risk of death in patients undergoing coronary 3. Mill JG, Milanez MC, Busatto VCW, Moraes AC, Gomes
artery bypass grafting. However, it is believed that other MGS. Ativação da Enzima conversora de angiotensina no
studies [29,30] did not show similar results due to small coração após infarto do miocárdio e suas repercussões no
sample size to detect differences in mortality. remodelamento ventricular. Arq Bras Cardiol.
1997;69(2):101-10.
Interrupting the use of ACE inhibitors, or reducing the
dose in the immediate postoperative period, and their 4. Anderson TJ, Elstein E, Haber H, Charbonneau F. Comparative
reintroduction into the postoperative period, may be study of ACE-inhibition, angiotensin II antagonism, and calcium
reasonable alternatives to minimize the acute effects, without channel blockade on flow-mediated vasodilatation in patients
loss of their chronic cardioprotective effects. Therefore, our with coronary disease (BANFF study). J Am Coll Cardiol.
work raises the hypothesis to be tested in future studies. 2000;35(1):60-6.
Our study has some limitations. The fact that it prevents
the best observational evaluation of the intervention, 5. Heart Outcomes Prevention Evaluation Study Investigators.
making a hypothesis-generating study, whose results need Effects of ramipril on cardiovascular and microvascular
outcomes in people with diabetes mellitus: results of the HOPE
further investigation with controlled trials. Patients who
study and MICRO-HOPE substudy. Heart Outcomes
were using ACE inhibitors in the preoperative period had a Prevention Evaluation Study Investigators. Lancet. 2000:
higher number of comorbidities. We have no information 355(9200):253-9.
of the real reason for not prescribing ACE inhibitors in the
preoperative period. The use or nonuse of ACE inhibitors 6. Miyazaki M, Sakonjo H, Takai S. Anti-atherosclerotic effects
before surgery was an option of the doctor who referred of an angiotensin converting enzyme inhibitor and an
the patient to the implementation of CRM. Even with an angiotensin II antagonist in Cynomolgus monkeys fed a high-
indication if we consider the underlying disease that caused cholesterol diet. Br J Pharmacol. 1999;128(3):523-9.
the procedure, nearly half of the sample was not in use. In
7. Brasier A, Recinos A 3rd, Eledrisi MS. Vascular inflammation
these patients, the ACE inhibitor was not prescribed in the
and the renin-angiotensin system. Atherioscler Thromb Vasc
hospital for a short period before surgery. In addition, we Biol. 2002;22(8):1257-66.
did not perform a separate analysis of treatment with ACE
inhibitors or ARA2. During the study period, all 8. Tuman KJ, McCarthy RJ, O’Connor CJ, Holm WE,
revascularization procedures were performed by the same Ivankovich AD. Angiotensin-converting enzyme inhibitors
group of cardiac surgeons, with no significant changes in increase vasoconstrictor requirements after cardiopulmonary
surgical technique. However, there was this period changes bypass. Anesth Analg. 1995;80(3):473-9.
in the pharmacological treatment of ischemic heart disease
that could be related to the incidence of outcomes. 9. Arora P, Rajagopalam S, Ranjan R, Kolli H, Singh M, Venuto
R, et al. Pre-operative use of angiotensin-converting enzyme
inhibitors/angiotensin receptor blockers is associated with
CONCLUSION increased risk for acute kidney injury after cardiovascular
surgery. Clin J Am Soc Nephrol. 2008;3(5):1266-73.
The preoperative use of ACE inhibitors was associated
with increased need for inotropic support postoperatively, 10. Pigott DW, Nagle C, Allman K, Westaby S, Evans RD. Effect
and the higher incidence of ARF and AF not associated with of omitting regular ACE inhibitor medication before cardiac
an increased rate of myocardial infarction, stroke and death. surgery on haemodynamic variables and vasoactive drug
requirements. Br J Anaesth. 1999;83(5):715-20.
REFERENCES 11. Rady MY, Ryan T. The effects of preoperative therapy with
angiotensin-converting enzyme inhibitors on clinical outcome
1. Yusuf S, Sleight P, Pogue J, Bosch J, Davies R, Dagenais G. after cardiovascular surgery. Chest. 1998;114(2):487-94.
Effects of an angiotensin-converting-enzyme inhibitor, ramipril,
on cardiovascular events in high-risk patients. The Heart 12. Nascente RB, Guaragna JCVC, Franca AS, Spiandorello RM,
378
CABG
Werutski G, Azevedo E, et al. Estabelecimento do ponto de 22. Arora P, Rajagopalam S, Ranjan R, Kolli H, Singh M, Venuto
corte da troponina I como marcador de infarto do miocárdio R, et al. Preoperative use of angiotensin-converting enzyme
em cirurgia de revascularização miocárdica. Sci Med. inhibitors/angiotensin receptor blockers is associated with
2005;15(3):142-7. increased risk for acute kidney injury after cardiovascular
surgery. Clin J Am Soc Nephrol. 2008;3(5):1266-73.
13. Cadore MP, Guaragna JCVC, Anacker JFA, Albuquerque LC,
Bodanese LC, Piccoli JCE, et al. Proposição de um escore de 23. Cittanova ML, Zubicki A, Savu C, Montalvan C, Nefaa N,
risco cirúrgico em pacientes submetidos à cirurgia de Zaier K, et al. The chronic inhibition of angiotensin-converting
revascularização miocárdica. Rev Bras Cir Cardiovasc. enzyme impairs postoperative renal function. Anesth Analg.
2010;25(4):447-56. 2001;93(5):1111–5.
14. Pfeffer MA, Greaves SC, Arnold JM, Glynn RJ, LaMotte FS, 24. Rouleau JL, Warnica WJ, Baillot R, Block PJ, Chocron S,
Lee RT, et al. Early versus delayed angiotensin-converting Johnstone D, et al; IMAGINE (Ischemia Management with
enzyme inhibition therapy in acute myocardial infarction. The Accupril post-bypass Graft via Inhibition of the coNverting
healing and early afterload reducing therapy trial. Circulation. Enzyme) Investigators. Effects of angiotensin-converting
1997;95(12):2643-51. enzyme inhibition in low-risk patients early after coronary
artery bypass surgery. Circulation. 2008;117(1):24-31.
15. The SOLVD Investigators. Effect of enalapril on survival in
patients with reduced left ventricular ejection fractions and 25. Kjøller-Hansen L, Steffensen R, Grande P. The Angiotensin-
congestive heart failure. N Engl J Med. 1991;325(5):293-302. converting Enzyme Inhibition Post Revascularization Study
(APRES). J Am Coll Cardiol. 2000;35(4):881-8.
16. Lazar HL. All coronary artery bypass graft surgery patients
will benefit from angiotensin-converting enzyme inhibitors. 26. White CM, Kluger J, Lertsburapa K, Faheem O, Coleman CI.
Circulation. 2008;117(1):6-8. Effect of pre-operative angiotensin converting enzyme
inhibitor or angiotensin receptor blocker use on the frequency
17. Tuman KJ, McCarthy RJ, O’Connor CJ, Holm WE, of atrial fibrillation after cardiac surgery: a cohort study from
Ivankovich AD. Angiotensin-converting enzyme inhibitors the atrial fibrillation suppression trials II and III. Eur J
increase vasoconstrictor requirements after cardiopulmonary Cardiothorac Surg. 2007;31(5):817–20.
bypass. Anesth Analg. 1995;80(3):473-9.
27. Echahidi N, Pibarot P, O’Hara G, Mathieu P. Mechanisms,
18. Argenziano M, Chen JM, Choudhri AF, Cullinane S, Garfein prevention, and treatment of atrial fibrillation after cardiac
E, Weinberg AD, et al. Management of vasodilatory shock surgery. J Am Coll Cardiol. 2008;51(8):793-801.
after cardiac surgery: identification of predisposing factors
and use of a novel pressor agent. J Thorac Cardiovasc Surg. 28. Miceli A, Capoun R, Fino C, Narayan P, Bryan AJ, Angelini
1998;116(6):973-80. GD, et al. Outcome in patients undergoing coronary artery
bypass grafting effects of angiotensin-converting enzyme
19. Raja SG, Fida N. Should angiotensin converting enzyme inhibitor therapy on clinical. J Am Coll Cardiol. 2009.
inhibitors/angiotensin II receptor antagonists be omitted before Disponível em: http://content.onlinejacc.org/cgi/content/full/
cardiac surgery to avoid postoperative vasodilation? Interact j.jacc.2009.07.008v1
Cardiovasc Thorac Surg. 2008;7(3):470-5.
29. Rady MY, Ryan T. The effects of preoperative therapy with
20. Deakin CD, Dalrymple-Hay MJ, Jones P, Monro JL. Effects angiotensin-converting enzyme inhibitors on clinical outcome
of angiotensin converting enzyme inhibition on systemic after cardiovascular surgery. Chest. 1998;114(2):487-94.
vascular resistance and vasoconstrictor requirements during
hypothermic cardiopulmonary bypass. Eur J Cardiothorac 30. Weightman WM, Gibbs NM, Sheminant MR, Whitford EG,
Surg. 1998;13(5):546-50. Mahon BD, Newman MA. Drug therapy before coronary
artery surgery: nitrates are independent predictors of mortality
21. Rosner MH, Okusa MD. Acute kidney injury associated with and beta-adrenergic blockers predict survival. Anesth Analg.
cardiac surgery. Clin J Am Soc Nephrol. 2006;1(1):19-32. 1999;88(2):286-91.
379
Predicting Hospital Mortality and Analysis of
Long-Term Survival After Major Noncardiac
ADULT CARDIAC
Complications in Cardiac Surgery Patients
Parwis B. Rahmanian, MD, David H. Adams, MD, Javier G. Castillo, MD,
Alain Carpentier, MD, PhD, and Farzan Filsoufi, MD
Department of Cardiothoracic Surgery, Heart Center, University Hospital Cologne, Germany, and Department of Cardiothoracic
Surgery, Mount Sinai School of Medicine, New York, New York
Background. This study was designed to investigate the (n ⴝ 111; 1.7%), and gastrointestinal complication (n ⴝ
incidence of and early and midterm outcomes after major 99; 1.5%). Overall operative mortality was 20% and cor-
complications in cardiac surgery patients. We determined related with the number of complications (single, 12.0%;
independent predictors of operative mortality to create a n ⴝ 58 of 485; double, 25.5%; n ⴝ 52 of 204; >3, 40.1%;
model for prediction of outcome. A particular focus was n ⴝ 55 of 137). Ten preoperative and five postoperative
the fate of patients after the occurrence of these predictors of operative mortality were identified and
complications. included in the logistic model, which accurately pre-
Methods. Prospectively collected data of 6,641 patients dicted outcome (C statistic, 0.866). One-year survival was
(mean age, 64 ⴞ 14 years; n ⴝ 2,499 female [38%]) less than 50% in patients with three or more complica-
undergoing cardiac surgery between January 1998 and tions and a length of stay greater than 60 days.
December 2006 were retrospectively analyzed. Outcome Conclusions. With a worsening in the risk profile of
measures were six index complications: respiratory fail- patients undergoing cardiac surgery, an increasing num-
ure, sepsis, dialysis-dependent renal failure, mediastini- ber of patients develop major complications leading to
tis, gastrointestinal complication, and stroke; and their increased length of stay and mortality, which is corre-
impact on operative mortality, hospital length of stay, lated to the number and severity of these complications.
and midterm survival using multivariate regression mod- Our predictive model based on preoperative and postop-
els. The discriminatory power was evaluated by calculat- erative variables allowed us to determine with accuracy
ing the area under the receiver operating characteristic the operative mortality in critically ill patients after
curves (C statistic). cardiac surgery. One-year survival after multiple compli-
Results. A total of 1,354 complications were observed cations and prolonged length of stay remains marginal.
in 826 (12.4%) patients: respiratory failure (n ⴝ 634; 9.5%),
sepsis (n ⴝ 202; 3%), stroke (n ⴝ 163; 2.5%), dialysis- (Ann Thorac Surg 2010;90:1221–9)
dependent renal failure (n ⴝ 145; 2.2%), mediastinitis © 2010 by The Society of Thoracic Surgeons
D uring the last decade the profile of cardiac surgery

patients has worsened, with a trend toward a more
elderly surgical population presenting with multiple co-
the intensive care unit (ICU) that is often complicated by
additional major morbidities. There is little information
with respect to the prevalence and characteristics of this
morbidities such as diabetes, depressed ventricular func- very complex subgroup of patients who have multiple
tion, peripheral vascular disease, renal dysfunction, or pre- complications and survive a prolonged hospital stay. In
vious stroke [1, 2]. In addition, most of these patients addition, most clinical series in cardiac surgery report early
require more complex procedures than isolated coronary results such as 30-day survival, and the status of these
artery bypass grafting (CABG). This accumulation of risk patients beyond this short period is not well determined.
factors in cardiac surgery patients has led to an overall This study was designed to investigate the incidence of
increase in the incidence of major postoperative complica- and early and long-term mortality after major complica-
tions [1, 3]. Most previous studies regarding postoperative
tions in patients undergoing cardiac surgery. In addition
morbidities in cardiac surgery have predominantly focused
we determined independent predictors of in-hospital
on patients presenting with one single complication. Today,
mortality to create a model that could serve as a tool for
with modern intensive care management, most of these
the prediction of survival in this patient population.
patients survive the initial insult with a prolonged stay in
Accepted for publication May 6, 2010.

Material and Methods
Address correspondence to Dr Filsoufi, Department of Cardiothoracic
Surgery, Mount Sinai Hospital, 1190 Fifth Ave, Box 1028, New York, We retrospectively analyzed prospectively collected data
NY 10029-1028; e-mail: farzan.filsoufi@mountsinai.org of 7,174 consecutive patients who underwent cardiac
© 2010 by The Society of Thoracic Surgeons 0003-4975/$36.00

Published by Elsevier Inc doi:10.1016/j.athoracsur.2010.05.015
1222 RAHMANIAN ET AL Ann Thorac Surg
MULTIPLE COMPLICATIONS AFTER CARDIAC SURGERY 2010;90:1221–9
surgery at our institution between January 1998 and Model Development

December 2006. Preoperative demographic information, The ␹2 test, Fisher’s exact test, and Student’s t test were
intraoperative and postoperative variables, and events used as appropriate to evaluate the relationship between
were extracted from a computerized database using the preoperative and postoperative variables and in-hospital
New York State Department of Health data registry. This mortality in univariate analysis. Stepwise multivariate
ADULT CARDIAC
registry represents a mandatory verified peer-reviewed logistic regression was then performed to assess the
data collection system that includes all cardiac surgery influence of these variables as independent risk factors
procedures in New York State [4]. Medical chart review for in-hospital mortality. The cutoff probability value for
was performed to obtain additional information when inclusion and exclusion was 0.10. The Hosmer-
necessary. Follow-up survival information was obtained Lemeshow goodness-of-fit test was used to assess the
by cross-matching the patient’s social security number calibration of the model [6]. A logistic equation was then
with the Web-based social security death index (http:// created using the coefficients of the regression analysis to
ssdi.rootsweb.com/). The protocol was approved by our estimate individual patient’s risk of in-hospital mortality:
local institutional review board and compliant with the
Health Insurance Portability and Accountability Act reg- Exp⌺(XⴱB)⫹intercept(␣)
Probability of dying ⫽
ulations. The approval included a waiver of informed 1 ⫹ Exp⌺(XⴱB)⫹intercept(␣)
consent. where Exp stands for exponential function, XⴱB is the
The following groups of patients were excluded: car- Coefficient B for each single confounding factor, and ␣ is
diac transplantation and elective ventricular assist device the model intercept. After the model was constructed, the
implantation (n ⫽ 201) and preoperative dialysis- probability of dying was calculated for each patient.
dependent renal failure (n ⫽ 254). Patients who died in
the operating room (n ⫽ 24) and those who required an Validation of the Model
assist device for postcardiotomy shock (n ⫽ 54) were also For validation of the model, we included a separate random
excluded. The rationale for excluding assist device and cohort of patients who underwent cardiac surgery between
transplantation procedures was because this patient col- January 2007 and June 2007 at our institution. The same
lective is prone to a different risk of complications related data elements used for the creation of the model were
to the presence of a mechanical device (drive-line infec- available for the internal verification cohort. The probability
tion, stroke) or to immunosuppression and rejection in of dying was calculated for each individual patient. To
case of transplant patients. A total of 6,641 patients were measure and compare the predictive accuracy of the model
included in the statistical analysis. in the study population and internal verification cohort, we
The main outcome measures of this study were six major generated receiver operating characteristic (ROC) curves
postoperative complications (index complications; for defi- and compared their area under the curves (AUC, C statis-
nitions see [4])—namely respiratory failure (RF), sepsis,
dialysis-dependent renal failure, deep sternal wound infec-
tion, gastrointestinal complication (GIC), and stroke—and Table 1. Surgical Procedures Performed During Study Period
their impact on in-hospital mortality (death during same
Procedures n
admission or within 30 days after surgery when dis-
charged). Patients who did not experience any of these Isolated CABG 2,880
index complications were assigned to the no complication Valve procedures 1,436
group, whereas patients with at least one index complica- Single valve 976
tion were allocated to the complications group. Aortic valve replacement 426
Further outcome variables were length of hospital stay Mitral valve repair 301
and midterm survival of discharged patients after the Mitral valve replacement 204
aforementioned complications. Finally, we analyzed the Tricuspid repair or replacement 45
impact of hospital length of stay (LOS) on midterm Double valve 416
mortality. With tricuspid valve 283
Without tricuspid valve 133
Operative Procedures and Perioperative Management Triple valve 44
During the study period a total of 11 cardiac surgeons Aortic procedures 1,252
operated on the patients included in the analysis. All Bentall or valve-sparing 515
procedures were performed using standard anesthetic Ascending arch 400
and surgical techniques. The detail of our perioperative Other 337
management has been previously reported [5]. Combined procedures 1,073
CABG ⫹ mitral valve repair 182
Statistical Analysis
CABG ⫹ mitral valve replacement 91
Normally distributed continuous variables are presented CABG ⫹ aortic valve replacement 366
as mean ⫾ standard deviation and otherwise as median
Other 434
and interquartile range (IQR). Categorical variables are
shown as the percentage of the sample. CABG ⫽ coronary artery bypass grafting.
Ann Thorac Surg RAHMANIAN ET AL 1223
2010;90:1221–9 MULTIPLE COMPLICATIONS AFTER CARDIAC SURGERY
Table 2. Patient Demographics Including a Univariate Comparison of the Association of One or More Complications With
Preoperative Variables
All Patients No Complication Complication(s)
Characteristic (n ⫽ 6,641) Group (n ⫽ 5,815) Group (n ⫽ 826) OR (95% CI) p Value
ADULT CARDIAC
Female sex 2,499 (38) 2,134 (37) 365 (44) 1.4 (1.2–1.6) ⬍0.001
Age (y) 63.9 ⫾ 13.9 63.3 ⫾ 14.0 68 ⫾ 13 ⬍0.001
Age ⬎70 y 2,608 (39) 2,164 (37) 444 (54) 2.0 (1.7–2.3) ⬍0.001
Body mass index kg/m2 26.9 ⫾ 6.2 26.9 ⫾ 5.5 27.2 ⫾ 10.0 0.156
Body mass index ⬎30 kg/m2 1,458 (22) 1,273 (22) 185 (22) 1.1 (0.9–1.3) 0.386
Ejection fraction 0.476 ⫾ 0.145 0.481 ⫾ 0.136 0.446 ⫾ 0.192 ⬍0.001
Ejection fraction ⬍0,30 1,013 (15) 807 (14) 206 (25) 2.1 (1.8–2.5) ⬍0.001
Hypertension 4,498 (68) 3,915 (67) 583 (71) 1.1 (1.0–1.3) 0.060
Diabetes 1,604 (24) 1,389 (24) 215 (26) 1.1 (0.9–1.3) 0.097
Peripheral vascular disease 826 (12) 652 (11) 174 (21) 2.1 (1.7–2.6) ⬍0.001
Cerebrovascular accident 512 (8) 424 (7) 88 (11) 1.5 (1.2–2.0) 0.001
Renal failure 138 (2) 94 (2) 44 (5) 3.4 (2.4–4.9) ⬍0.001
Chronic obstructive 475 (7) 369 (6) 106 (13) 2.2 (1.8–2.7) ⬍0.001
pulmonary disease
Prior myocardial infarction 1,968 (30) 1,676 (29) 292 (35) 1.3 (1.1–1.6) ⬍0.001
Acute myocardial infarction 79 (1) 52 (1) 27 (3) 3.7 (2.3–6.0) ⬍0.001
Congestive heart failure 1,926 (29) 1,543 (27) 383 (46) 2.4 (2.1–2.8) ⬍0.001
Hemodynamic instabilitya 189 (3) 112 (2) 77 (9) 5.2 (3.9–7.1) ⬍0.001
Emergent procedureb 333 (5) 251 (4) 82 (10) 2.4 (1.9–3.2) ⬍0.001
Reoperation 978 (15) 772 (13) 206 (25) 2.2 (1.8–2.6) ⬍0.001
Endocarditis 133 (2) 97 (2) 36 (4) 2.7 (1.8–4.0) ⬍0.001
Aortic calcification 371 (6) 287 (5) 84 (10) 2.2 (1.7–2.8) ⬍0.001
Procedures
Isolated valve procedures 1,436 (22) 1,271 (22) 165 (20) 0.9 (0.7–1.1) 0.118
Isolated CABG procedures 2,880 (43) 2,657 (46) 223 (27) 0.4 (0.4–0.5) ⬍0.001
Combined valve/CABG 1,073 (16) 876 (15) 197 (24) 1.8 (1.5–2.1) ⬍0.001
and other procedures
Aortic procedures 1,252 (19) 1,011 (17) 241 (29) 2.0 (1.7–2.3) ⬍0.001
Other than CABG 3,760 (57) 3,157 (54) 603 (73) 2.3 (1.9–1.7) ⬍0.001
Cross-clamp time (min) 101.7 ⫾ 61.4 100.2 ⫾ 59.8 113.6 ⫾ 71.8 ⬍0.001
CPB time (min) 142.2 ⫾ 77.6 137.4 ⫾ 74.3 179.1 ⫾ 91.8 ⬍0.001
a b
Hemodynamic instability is defined as patient requiring pharmacologic or mechanical support to maintain blood pressure or cardiac output. Emergent
operation is defined as refractory unrelenting cardiac compromise requiring emergency operation.
CABG ⫽ coronary artery bypass grafting; CI ⫽ confidence interval; CPB ⫽ cardiopulmonary bypass; OR ⫽ odds ratio.
tics). Finally, we compared our score with the New York 2,880) underwent isolated CABG, whereas the remaining
State Risk Scores for isolated CABG, isolated valve, and 57% underwent other cardiac procedures (Table 1). These
combined valve and CABG procedures [7, 8] as well as to included 22% (n ⫽ 1,436) isolated valve procedures, 16%
the logistic EuroSCORE [9]. (n ⫽ 1,073) combined valve and CABG procedures, and
A probability value less than 0.05 was considered 19% (n ⫽ 1,252) aortic procedures.
significant for all statistical methods. Midterm survival A total of 1,354 index complications were observed in 826
was analyzed using Kaplan-Meier survival curves. Dif- (12.4%) patients. Univariate risk factors for the development
ferences in patient characteristics were controlled by Cox of one or more index complications are shown in Table 2.
proportional hazard analysis adjusted for age, sex, ejec- The most frequent index complication was RF (n ⫽ 634;
tion fraction, diabetes, peripheral vascular disease, and 9.5%) followed by sepsis (n ⫽ 202; 3%), stroke (n ⫽ 163;
type of procedure. Statistical analyses were performed 2.5%), dialysis-dependent renal failure (n ⫽ 145; 2.2%), deep
with SPSS 18 (SPSS Inc, Chicago, IL). sternal wound infection (n ⫽ 111; 1.7%), and GIC (n ⫽ 99;
1.5%). The distribution of index complications per patient
was as follows: single index complication, n ⫽ 485 (59%);
Results two index complications, n ⫽ 204 (25%); three or more index
A total of 6,641 patients were included in this study. The complications, n ⫽ 137 (16%). Most multiple complications
mean age was 63.9 ⫾ 13.9 years, and 38% (n ⫽ 2,499) of included combinations with RF or sepsis (Table 3).
patients were female. Forty-three percent of patients (n ⫽ The overall in-hospital mortality was 3.9% (n ⫽ 257 of
Table 3. Distribution of Major Postoperative Index Complications and Associated In-Hospital Mortalitya
Frequency Mortality
Group n % of All % of Complications % in Group n %

ADULT CARDIAC
No complication 5,815 87.6 92 1.6

Single 485 7.3 58.7 58 12.0
RF 320 4.8 38.7 66.0 28 8.8
CVA 68 1.0 8.2 14.0 14 20.6
DSWI 34 0.5 4.1 7.0 0 0.0
Sepsis 27 0.4 3.3 5.6 3 11.1
Dialysis 22 0.3 2.7 4.5 7 31.8
GIC 14 0.2 1.7 2.9 6 42.9
Double 204 3.1 24.7 52 25.5
RF ⫹ sepsis 56 0.8 6.8 27.5 15 26.8
RF ⫹ CVA 44 0.7 5.3 21.6 9 20.5
RF ⫹ dialysis 40 0.6 4.8 19.6 17 42.5
RF ⫹ GIC 26 0.4 3.1 12.7 5 19.2
RF ⫹ DSWI 17 0.3 2.1 8.3 0 0.0
Other double complications 21 0.3 2.5 10.3 6 28.6
3 or more complications 137 2.1 16.6 55 40.1
RF ⫹ sepsis ⫹ dialysis 32 0.5 3.9 23.4 19 59.4
RF ⫹ sepsis ⫹ CVA 14 0.2 1.7 10.2 2 14.3
RF ⫹ sepsis ⫹ DSWI 11 0.2 1.3 8.0 1 9.1
RF ⫹ CVA ⫹ dialysis 8 0.1 1.0 5.8 5 62.5
RF ⫹ sepsis ⫹ DSWI ⫹ GIC 8 0.1 1.0 5.8 2 25.0
RF ⫹ sepsis ⫹ GIC 7 0.1 0.8 5.1 2 28.6
RF ⫹ dialysis ⫹ GIC 7 0.1 0.8 5.1 7 100.0
Other complications (ⱖ3) 50 0.8 6.1 36.5 17 34.0
Total 6,641 100.0 257 3.9
a
In-hospital mortality is defined as death during the same admission or within 30 days after surgery when discharged.
CVA ⫽ cerebrovascular accident; DSWI ⫽ deep sternal wound infection; GIC ⫽ gastrointestinal complication; RF ⫽ respiratory failure.
6,641; Table 3). The in-hospital mortality rate among pa- electromechanical dissociation and malignant ventricular
tients of the complications group (ⱖ1 index complication) arrhythmia.
was 20% (n ⫽ 165 of 826) compared with 1.6% (n ⫽ 92 of Univariate and subsequent multivariate logistic re-
5,815) in patients of the no complication group. The overall gression analysis revealed ten preoperative and five
mortality of patients with a single index complication was postoperative independent predictors of in-hospital mor-
12.0% (n ⫽ 58 of 485). The highest mortality among patients tality (Table 4). The result of the Hosmer-Lemeshow
with a single index complication was observed in those who analysis was not statistically significant (p ⬎ 0.05), sug-
experienced GIC (42.9%), dialysis-dependent renal failure gesting a good calibration of the model.
(31.8%), and stroke (20.6%). Patients with two index com-
plications had a twofold increased risk of death compared Predictive Model and Performance
with those with a single index complication (25.5%; n ⫽ 52 The predictive model was created based on these 15
of 204). Patients with two index complications most fre- variables determined by the multivariate analysis. We
quently had RF in combination with another index compli- first calculated the risk of in-hospital mortality of indi-
cation. The highest mortality rate was observed in patients vidual patients and compared the results with the ob-
with RF and dialysis (42.5%), followed by RF and sepsis served mortality using the C statistic. The ROC area
(26.8%), and finally those with RF and stroke (20.5%). The under the curve for the study population was 0.866. We
mortality rate further increased when patients with three or then applied the same model to our internal verification
more index complications were analyzed (40.1%; n ⫽ 55 of cohort (n ⫽ 344; mean age, 64 ⫾ 14 years; 40% [n ⫽ 256]
137). Patients with RF, GIC, and dialysis had a 100% (n ⫽ 7 female) and compared the predicted mortality with the
of 7) mortality rate followed by 62.5% among patients with observed mortality. The distribution of cardiac surgery
RF, stroke, and dialysis-dependent renal failure. Patients of procedures in this cohort was similar to the study popu-
the no complication group had a mortality rate of 1.6% (n ⫽ lation (isolated CABG, 42%; isolated valves, 22%; com-
92). The majority of these patients died within the first 3 bined valve and CABG, 17%; aortic procedures, 19%).
days after surgery (median, 68 hours; IQR, 18 to 168 hours). The predicted and observed mortality of the verification
Causes of death included mainly cardiac failure owing to cohort was 3.9% and 4.1%, respectively. The ROC area
Table 4. Predictors of In-Hospital Mortality in Multivariate Follow-up information was obtained for 98% of 6,384
Analysis surviving patients. The lack of information regarding the
remaining 2% can be explained by invalid social security
Coefficient p
Variable B OR 95% CI Value numbers and patients from abroad. Midterm survival was
significantly reduced in patients with index complications,
ADULT CARDIAC
Preoperative variables with a direct correlation between the number of index
Female sex 0.5 1.7 (1.3–2.3) ⬍0.001 complications and midterm mortality. Postdischarge sur-
Age ⬎70 y 0.4 1.4 (1.1–1.9) 0.016 vival at 1 year in the no complication group was 96.6% ⫾
Ejection fraction ⬍0.30 0.6 1.8 (1.3–2.5) ⬍0.001 0.2% compared with 85.8% ⫾ 1.7% and 72.0% ⫾ 3.6% in
Peripheral vascular 0.7 2.0 (1.4–2.8) ⬍0.001 patients with single or double index complications, respec-
disease tively. Patients with three or more index complications had
Creatinine ⬎2.5 no 0.9 2.4 (1.3–4.4) 0.006 a 1-year survival of only 56.8% ⫾ 5.7% (p ⬍ 0.001). The
dialysis
adjusted survival curves according to the number of index
Acute MI 1.1 2.9 (1.3–6.4) 0.009 complications are shown in Figure 3A. We also observed
Hemodynamic 0.8 2.2 (1.2–3.9) 0.007 an association between LOS and midterm survival.
instability
Patients who were hospitalized for 10 or fewer days
Emergent procedures 0.6 1.8 (1.1–3.1) 0.032
had a 98.0% ⫾ 0.2% 1-year survival rate compared with
Reoperation 0.4 1.5 (1.1–2.1) 0.019
92.1% ⫾ 0.8% and 76.7% ⫾ 2.6% among those with an
Other than CABG 0.5 1.6 (1.2–2.3) 0.006 LOS of 11 to 30 days and 31 to 60 days, respectively.
Postoperative morbidities Almost half of the patients who were hospitalized for
Respiratory failure 0.9 2.5 (1.7–3.6) ⬍0.001 more than 60 days died within the first year after
Sepsis 0.5 1.6 (1.0–2.6) 0.042 discharge (1-year survival, 55.9% ⫾ 4.0%; p ⬍ 0.001; Fig
Stroke 1.2 3.4 (2.1–5.4) ⬍0.001 3B). Patients with a single index complication and an
Renal failure (dialysis) 2.1 8.2 (5.2–12.9) ⬍0.001 LOS greater than 60 days had a midterm mortality of
GIC 1.2 3.2 (1.8–5.6) ⬍0.001 60%, similar to patients with multiple index complica-
Intercept (a) ⫺5.0 tions and prolonged hospital stay, suggesting that the
length of hospitalization was a marker for the severity
CABG ⫽ coronary artery bypass grafting; CI ⫽ confidence interval;
GIC ⫽ gastrointestinal complication; MI ⫽ myocardial infarc- of the complication in these patients.
tion; OR ⫽ odds ratio.
Comment
under the curve for the internal verification cohort was Incidence
0.892 and was not significantly different between the Our study showed an overall incidence of 12% of six
study population and internal verification cohort (p ⫽ index- complications (alone or in combination) after
0.716; Fig 1). When compared with the New York State
Risk scores and the EuroSCORE, it appeared that our
model more accurately predicted mortality as shown in
Figure 2. Using our model, for example, a 50-year-old
man with no additional risk factors who experiences
postoperative RF after isolated CABG has a probability of
dying of 1.7%, whereas a 76-year-old woman with pe-
ripheral vascular disease and an ejection fraction less
than 0.30 who develops the same complication after an
aortic valve replacement with concomitant CABG has a
19.7% mortality risk. If the same patient experiences
additional sepsis and dialysis-dependent renal failure,
her risk of dying would increase to 76.6%.
Length of Hospital Stay and Midterm Survival

The median LOS in the overall population was 7 (IQR, 5
to 11) days. The no complication group had an LOS of 7
(IQR, 5 to 9) days compared with 26 (IQR, 14 to 48) days
in the complications group (p ⬍ 0.001). In the complica-
tions group, the mean LOS was 19 (IQR, 12 to 33), 35
(IQR, 21 to 57), and 50 (IQR, 32 to 85) days in patients with
one, two, and three or more index complications, respec-
tively (p ⬍ 0.001). Thirty-eight percent (n ⫽ 58 of 152) of
patients with two index complications and 72% (n ⫽ 59 of Fig 1. Validation of the predictive model for in-hospital mortality
82) of patients with three or more index complications including preoperative and postoperative factors using receiver oper-
had a hospital LOS greater than 50 days. ating characteristic (ROC) curves.
ADULT CARDIAC
Fig 2. Comparison of our score with the New York State Risk Scores for isolated coronary artery bypass grafting, isolated valve, and combined
valve and coronary artery bypass grafting procedures [7, 8] as well as the EuroSCORE [9]. *p ⬍ 0.001.
cardiac surgery. Respiratory failure and sepsis were the most previous publications regarding the incidence of
most prevalent index complications, whereas GIC was postoperative morbidity [3, 10 –13]. Rankin and col-
the least common. These findings are in accordance with leagues [3] and Shroyer and associates [10] analyzed
terns of complications emerge after prolonged ICU stay.

In our study, 58% of index complications were isolated
whereas two and three or more index complications were
observed in 42% of patients. The most common pattern of
double complications was the association of a single
ADULT CARDIAC
morbidity with RF whereas in patients with three of more
index complications the most frequent combination was
the association of RF, sepsis, and another single morbid-
ity. That is probably a reflection of an acute injury that
required a prolonged ICU stay and was further compli-
cated over time by classic ICU complications. The second
pattern was a combination of index complications that
are a reflection of advanced atherosclerotic disease (eg,
stroke or dialysis with or without GIC). The atheroscle-
rotic burden in these patients is reflected by preoperative
risk factors such as peripheral vascular disease and aortic
calcification, which potentially increase the risk of ab-
dominal organ hypoperfusion and thromboembolic
events, which represent the two main pathophysiologic
mechanisms of ischemic organ injury [11]. In most pa-
tients with multiple index complications, the develop-
ment of postoperative dialysis-dependent renal failure
was a sign of severity and was commonly associated with
the evolution toward multiorgan system failure.
In-Hospital Mortality
Previous studies on the predictors of major complications
after cardiac surgery have identified several risk factors,
such as age, diabetes, advanced atherosclerotic disease,
and low ejection fraction [11, 12, 17–19]. Considering that
recent studies have shown an increase in the prevalence
of these risk factors, it is important to determine the
outcome after these major complications as well as their
impact on an increased use of resources in a contempo-
rary cohort [1, 3, 20]. In our study, we have shown that the
Fig 3. Adjusted long-term survival according to the number of post-
in-hospital mortality depends on the type and the num-
operative index complications (A) and to the length of hospitaliza- ber of index complications. For example, the highest
tion (B). Adjusted for age, sex, ejection fraction, diabetes, peripheral mortality rate was observed after dialysis-dependent
vascular disease, and type of procedure. renal failure (48%) and GIC (34%). In contrast, the in-
hospital mortality was zero among 35 patients with
isolated deep sternal wound infection who required
outcomes of patients from The Society of Thoracic Sur- surgical intervention. One of the significant findings of
geons National Database and reported that RF was the our study was that the overall in-hospital mortality in
most common complication after valve and CABG pro- patients with two and three or more index complications
cedures with an incidence of 14% and 6%, respectively. was twofold (25.5% versus 12%) and threefold (40.1%
Rankin and coworkers [3] further reported that infectious versus 12%), respectively, higher than in patients with a
complications (pneumonia and sepsis) were the second single index complication. In these groups, the highest
most common morbidity with an incidence of 7%. Post- mortality was observed in patients presenting with dial-
operative renal dysfunction is also frequently observed ysis-dependent renal failure in combination with other
in cardiac surgery patients. An increase in serum creati- index complications. In patients requiring dialysis the
nine level may occur in up to 30% of patients after cardiac addition of one or two index complications increased the
surgery [14, 15], and the incidence of renal failure requir- in-hospital mortality from 32% to 44% and 56%, respec-
ing dialysis has been reported to be in the range of 1% to tively. Other studies have confirmed the significant im-
2% [16], similar to what we have seen in our own pact of renal failure alone or in combination with other
experience. Stroke, GIC, and deep sternal wound infec- complications by reporting an operative mortality rang-
tion were other devastating complications with an inci- ing from 30% to 80% [21, 22]. Another finding of this
dence of 2.5%, 1.5%, and 1.7%, respectively [11]. A careful study was that when more than two organ systems were
analysis of our cohort furthermore revealed that major involved, the development of another organ dysfunction
postoperative complications do not occur in isolation but had little additional impact, and the overall mortality in
are often associated with other complications, and pat- this subset remained greater than 40%.
Predictive Model of In-Hospital Mortality index complications, 1-year mortality was 14% compared
An accurate prediction of in-hospital mortality might be with 3% in patients without any index complication. The
important when it comes to decision making about con- survival rates in the complications group further de-
tinuing therapy in these critically ill patients. Most scor- creased with increasing number of complications. Pa-
ing systems used in intensive care medicine such as the tients with two morbidities had a 30% risk of dying
ADULT CARDIAC
Sequential Organ Failure Assessment (SOFA) or Acute during the first year after discharge. The early hazard of
Physiology and Chronic Health Evaluation II (APACHE dying was even worse in patients with three or more
II) score have been developed in general and medical index complications. In this subgroup, almost 50% of
ICU patient populations and not specifically in cardiac discharged patients did not survive the first year. Simi-
surgery patients, who often present particular risk factors larly, hospital LOS was directly associated with de-
such as impaired left ventricular function and general- creased survival after discharge. Patients who were hos-
ized atherosclerotic disease [23]. Furthermore, previous pitalized for 10 days or fewer had 1-year mortality of 2%
studies that have attempted specifically to identify criti- compared with 8% and 14% among those with an LOS of
cally ill cardiac surgery patients who have a high proba- 11 to 30 days and 31 to 60 days, respectively. Almost half
bility of hospital or late mortality have developed predic- of the patients who were hospitalized for more than 60
tive models based on preoperative characteristics only days died within the first year after discharge. Hein and
colleagues [28] similarly reported that patients with an
and are not taking into account postoperative morbidities
ICU length of stay greater than 3 days had a marked
that have their own additional impact on postoperative
decline in survival in the first 6 months after discharge
mortality [24, 25]. In our study, which involves a large
compared with patients with a shorter length of stay in
cohort of patients allowing us to generate and validate a
ICU (70% versus 95% at 6 months). Similarly, Bashour
predictive model, we were able to identify 15 indepen-
and associates [27] observed that patients with an ICU
dent predictors of in-hospital mortality: ten preoperative
length of stay of 10 days or longer had a survival rate of
and five postoperative variables. Most preoperative vari-
60% at 1 year. These findings suggest that 30-day survival
ables identified by our analysis, such as age, sex, and low
and hospital discharge are insufficient variables to mea-
ejection fraction, have also been shown by other studies
sure outcome after cardiac surgery, and therefore it is
to negatively impact surgical outcome [12, 20, 26]. In
preferable to include 1-year survival in clinical outcome
addition we included postoperative variables, which,
research. This more complete measure would allow phy-
once they occur, have an independent additional effect
sicians to better evaluate the outcome of critically ill
on in-hospital mortality. Among these index complica- patients after cardiac surgery and might be helpful to
tions, dialysis-dependent renal failure was the strongest further determine the resource allocation and long-term
predictor of in-hospital mortality. Using our logistic health-care costs.
model, we were able to predict in-hospital mortality of
individual patients after the occurrence of major postop- Strengths and Limitations
erative complications. The accuracy of our model was The data analyzed in this study were obtained from the
shown by the C statistic, which showed similar areas New York State Department of Health registry, a state-
under the ROC curves for both the study population mandated database with external audit, and therefore
and the validation cohort or internal verification co- provide very accurate information about perioperative
hort. When our model was compared with the New variables. The study also includes a large and heteroge-
York State Risk Score for isolated CABG, valve, and neous group of patients who underwent a variety of
combined valve and CABG procedures [7, 8], and with surgical procedures, and therefore the findings are ap-
the logistic EuroSCORE [9], it predicted mortality also plicable to a broad spectrum of cardiac surgery patients.
more accurately. Despite the effectiveness of our model, However, this was a retrospective observational study,
we would like to emphasize that any scoring system and therefore conclusions are necessarily limited in their
generates an estimation, and misclassification rates as application. In this study we did not include postopera-
high as 15% have been observed using predictive models tive low cardiac output as a postoperative morbidity in
[27]. Furthermore, clinical scoring systems need to be our analysis. Some previous studies have used postoper-
readjusted frequently to address time-related changes in ative administration of inotropic medication to define
patient demographics and improvements in medical postoperative low cardiac output or cardiogenic shock.
therapy. Nonetheless, if applied with caution, it could be However, with the greater number of patients with
a useful tool in the interaction with the patient and his or impaired left ventricular function, today the use of ino-
her family to provide an estimate of hospital survival and tropic medication in cardiac surgery patients is common
to help decision making about the length and invasive- and we believe that this is not a precise enough variable
ness of ICU therapy in critically ill patients. to define postoperative cardiogenic shock. Furthermore,
the dosage and duration of inotropic medication is often
One-Year Survival based on individual judgment and may be a very subjec-
The burden of postoperative morbidity continued be- tive variable. Other hemodynamic variables such as
yond hospitalization. Among discharged patients, sur- measured cardiac output or central venous oxygen satu-
vival was significantly different depending on the num- ration are more reliable, but these data were not captured
ber of index complications. In patients with one of the six by our database. It is important, however, to mention that
in our study we analyzed preoperative variables such as 9. Roques F, Michel P, Goldstone AR, Nashef SA. The logistic
ejection fraction, history of congestive heart failure, EuroSCORE. Eur Heart J 2003;24:881–2.
10. Shroyer AL, Coombs LP, Peterson ED, et al. The Society of
emergent procedure, and hemodynamic instability on Thoracic Surgeons: 30-day operative mortality and morbid-
admission, which are strong predictors of postoperative ity risk models. Ann Thorac Surg 2003;75:1856 – 65.
low cardiac output, and most of them remained indepen- 11. Filsoufi F, Rahmanian PB, Castillo JG, Scurlock C, Legnani
ADULT CARDIAC
dent determinants of in-hospital mortality in our predic- PE, Adams DH. Predictors and outcome of gastrointestinal
tive model. Finally, clinical outcome analysis focused on complications in patients undergoing cardiac surgery. Ann
Surg 2007;246:323–9.
postoperative mortality and morbidity, and we were not 12. Canver CC, Chanda J. Intraoperative and postoperative risk
able to provide information on late complications, quality factors for respiratory failure after coronary bypass. Ann
of life, and cause of death after discharge. Thorac Surg 2003;75:853– 8.
13. Mangano CM, Diamondstone LS, Ramsay JG, Aggarwal A,
Conclusions Herskowitz A, Mangano DT. Renal dysfunction after myo-
cardial revascularization: risk factors, adverse outcomes, and
With a worsening in the risk profile of patients undergo- hospital resource utilization. The Multicenter Study of Peri-
ing cardiac surgery, an increasing number of patients operative Ischemia Research Group. Ann Intern Med 1998;
experience major postoperative complications. Owing to 128:194 –203.
14. Antunes PE, Prieto D, Ferrao de Oliveira J, Antunes MJ.
the continuous progress in intensive care management of
Renal dysfunction after myocardial revascularization. Eur
these complications, a growing number of patients sur- J Cardiothorac Surg 2004;25:597– 604.
vive the acute injury and become chronically ill. Our 15. Kuitunen A, Vento A, Suojaranta-Ylinen R, Pettila V. Acute
study allowed us to determine the incidence and the renal failure after cardiac surgery: evaluation of the RIFLE
pattern of these complications. Patients with postopera- classification. Ann Thorac Surg 2006;81:542– 6.
16. Thakar CV, Arrigain S, Worley S, Yared JP, Paganini EP. A
tive complications required prolonged hospitalization clinical score to predict acute renal failure after cardiac
and presented with increased mortality, which was cor- surgery. J Am Soc Nephrol 2005;16:162– 8.
related with the number and severity of these complica- 17. Mangi AA, Christison-Lagay ER, Torchiana DF, Warshaw
tions. A significant number of patients with major mor- AL, Berger DL. Gastrointestinal complications in patients
undergoing heart operation: an analysis of 8709 consecutive
bidities were discharged to rehabilitation facilities. These
cardiac surgical patients. Ann Surg 2005;241:895–901; discus-
complications also affected postdischarge survival, which sion 904.
was as low as 50% at 1 year in patients with three or more 18. Salazar JD, Wityk RJ, Grega MA, et al. Stroke after cardiac
complications or an initial LOS greater than 60 days. The surgery: short- and long-term outcomes. Ann Thorac Surg
development of a scoring system based on preoperative 2001;72:1195–202.
19. Rajakaruna C, Rogers CA, Suranimala C, Angelini GD,
and postoperative variables allowed us to determine with Ascione R. The effect of diabetes mellitus on patients under-
accuracy the in-hospital mortality in critically ill patients going coronary surgery: a risk-adjusted analysis. J Thorac
after cardiac surgery. Cardiovasc Surg 2006;132:802–10.
20. Topkara VK, Cheema FH, Kesavaramanujam S, et al. Coro-
nary artery bypass grafting in patients with low ejection
References fraction. Circulation 2005;112(9 Suppl):I-344 –50.
21. Bove T, Calabro MG, Landoni G, Aletti G, Marino G,
1. Ferguson TB Jr, Hammill BG, Peterson ED, DeLong ER, Crescenzi G, et al. The incidence and risk of acute renal
Grover FL. A decade of change—risk profiles and outcomes failure after cardiac surgery. J Cardiothorac Vasc Anesth
for isolated coronary artery bypass grafting procedures, 2004;18:442–5.
1990 –1999: a report from the STS National Database Com- 22. Chertow GM, Levy EM, Hammermeister KE, Grover F,
mittee and the Duke Clinical Research Institute. Society of Daley J. Independent association between acute renal failure
Thoracic Surgeons. Ann Thorac Surg 2002;73:480 –90. and mortality following cardiac surgery. Am J Med 1998;104:
2. McKhann GM, Grega MA, Borowicz LM Jr, Baumgartner 343– 8.
WA, Selnes OA. Stroke and encephalopathy after cardiac 23. Ryan TA, Rady MY, Bashour CA, Leventhal M, Lytle B, Starr
surgery: an update. Stroke 2006;37:562–71. NJ. Predictors of outcome in cardiac surgical patients with
3. Rankin JS, Hammill BG, Ferguson TB Jr, et al. Determinants prolonged intensive care stay. Chest 1997;112:1035– 42.
of operative mortality in valvular heart surgery. J Thorac 24. Tu JV, Jaglal SB, Naylor CD. Multicenter validation of a risk
Cardiovasc Surg 2006;131:547–57. index for mortality, intensive care unit stay, and overall
4. Adult Cardiac Surgery in New York State 2005–2007: New hospital length of stay after cardiac surgery. Steering Com-
York State Department of Health. Available at: http:// mittee of the Provincial Adult Cardiac Care Network of
www.health.state.ny.us/statistics/diseases/cardiovascular/ Ontario. Circulation 1995;91:677– 84.
heart_disease/docs/2005–2007_adult_cardiac_surgery. 25. Hein OV, Birnbaum J, Wernecke KD, Konertz W, Jain U,
pdf. Accessed May 1, 2010. Spies C. Three-year survival after four major post-cardiac
5. Rahmanian PB, Adams DH, Castillo JG, Vassalotti J, Filsoufi operative complications. Crit Care Med 2006;34:2729 –37.
F. Early and late outcome of cardiac surgery in dialysis- 26. Scott BH, Seifert FC, Grimson R, Glass PS. Octogenarians
dependent patients: single-center experience with 245 con- undergoing coronary artery bypass graft surgery: resource
secutive patients. J Thorac Cardiovasc Surg 2008;135:915–22. utilization, postoperative mortality, and morbidity. J Cardio-
6. Hosmer DW, Lemeshow S. Applied logistic regression. New thorac Vasc Anesth 2005;19:583– 8.
York: Wiley, 1989. 27. Bashour CA, Yared JP, Ryan TA, et al. Long-term survival
7. Hannan EL, Wu C, Bennett EV, et al. Risk stratification of and functional capacity in cardiac surgery patients after
in-hospital mortality for coronary artery bypass graft sur- prolonged intensive care. Crit Care Med 2000;28:3847–53.
gery. J Am Coll Cardiol 2006;47:661– 8. 28. Hein OV, Birnbaum J, Wernecke K, England M, Konertz W,
8. Hannan EL, Wu C, Bennett EV, et al. Risk index for predict- Spies C. Prolonged intensive care unit stay in cardiac sur-
ing in-hospital mortality for cardiac valve surgery. Ann gery: risk factors and long-term-survival. Ann Thorac Surg
Thorac Surg 2007;83:921–9. 2006;81:880 –5.
Original Cardiovascular 51
Acute Renal Dysfunction Does Not Develop More

Frequently Among Octogenarians Compared to
Septuagenarians after Cardiac Surgery
Michael Ried 1 Assad Haneya 1 Philipp Kolat 1 Tobias Potzger 1 Thomas Puehler 1 Cristof Schmid 1
Claudius Diez 1
1 Department of Cardiothoracic Surgery, University Medical Center Address for correspondence and reprint requests Michael Ried, M.D.,
Regensburg, Regensburg, Germany Department of Cardiothoracic Surgery, University Medical Center
Regensburg, Franz-Josef-Strauß Allee 11, Regensburg 93053, Germany
Thorac Cardiovasc Surg 2012;60:51–56. (e-mail: micha.ried@t-online.de).
Abstract Background We tested the hypothesis that octogenarians develop more frequently
renal dysfunction compared with septuagenarians after cardiac surgery.

Methods A retrospective, observational study on an age-, gender- and operation-
matched cohort of 598 patients, (299 octogenarians vs. 299 septuagenarians) who
underwent cardiac surgery between January 2006 and August 2009, was performed.
Kidney function was estimated with the abbreviated Modification in Renal Disease
equation and acute kidney injury was defined as a decrease of glomerular filtration
rate 50%.
Results Operations included 246 coronary, 198 isolated valve, and 154 combined
coronary and valve procedures. Mean logistic EuroSCORE was 8.5% in septuagenarians
and 13.2% in octogenarians. Octogenarians had significantly more frequent and estimat-
ed GFR < 60 mL/min/1.73 m2 (44 vs. 34.4%, p ¼ 0.02). The incidence of dialysis-
dependent acute kidney failure did not differ between both groups (6.7 vs. 5.4%, p ¼
0.60). Postoperative decline of glomerular filtration rate < 25% occurred significantly
more often in septuagenarians (40 vs. 30%, p ¼ 0.02). Septuagenarians with a preopera-
tive GFR < 60 mL/min/1.73 m2 had a higher 30-day mortality compared with patients with
Keywords a GFR > 60 mL/min/1.73 m2 (10.9 vs. 3.1%, p ¼ 0.02). Overall, 30-day mortality in
► octogenarians octogenarians was 7.7% without significant differences with respect to preoperative GFR.
► cardiac surgery Conclusions Octogenarians do not develop acute kidney failure more frequently than
► acute kidney injury their matched septuagenarian counterparts. They can be operated on at an acceptable
► renal dysfunction risk for morbidity and mortality. Preoperative impaired renal function is associated with
► mortality higher risk for mortality in septuagenarians.
In Germany, more than every second patient undergoing Previous studies reported higher rates for operative mor-
cardiac surgery is older than 70 years (50.8% in 2009). tality, postoperative complications, and increased hospital
Furthermore, the increasing life expectancy of the population costs in octogenarians undergoing cardiac surgery compared
and the higher incidence of cardiovascular disease with with younger patients.3–5 However, continuous advances in
advancing age have resulted in a significant higher number cardiac anesthesia, operative techniques, myocardial protec-
of patients > 80 years (11.8 in 2009 compared with 10.3 in tion, and perioperative care have led to acceptable mortality
2008) admitted for cardiac surgery.1,2 rates and satisfactory midterm results in patients 80 years.6
received Copyright © 2012 by Thieme Medical DOI http://dx.doi.org/

April 26, 2011 Publishers, Inc., 333 Seventh Avenue, 10.1055/s-0031-1295567.
accepted after revision New York, NY 10001, USA. ISSN 0171-6425.
July 12, 2011 Tel: +1(212) 584-4662.
published online
December 29, 2011
52 Acute Renal Dysfunction Ried et al.
Not only survival does reflect intention of treatment, but kidney injury are independent predictors of mortality in
there is also evidence for better functional status and im- elderly patients. Postoperative acute kidney injury was de-
proved quality of life for this elderly population after cardiac fined as a decrease of glomerular filtration rate 50% or the
surgery.7–9 This led to the application of cardiac surgical need of dialysis due to oliguria ( 5 mL/kg/24 h urine output).
procedures with increasing frequency in octogenarians. Operative mortality was defined as in-hospital and within
Additionally, the number of patients with impaired renal 30 days after surgery.
function requiring cardiac surgery has steadily increased and
acute kidney injury still remains a frequent complication in 5 Operative Technique
to 30% of the patients after cardiac surgery.10,11 Several The procedures were performed with either full or partial
studies determined preoperative and postoperative renal upper median sternotomy. All operations were performed
dysfunction as an independent predictor of increased peri- with the aid of standard cardiopulmonary bypass employing
operative morbidity with reduced survival rates.12–14 There- single cannulation of the ascending aorta and right atrium or
fore, this concomitant disease is an important component of bicaval cannulation when necessary for mitral valve proce-
perioperative care especially in the elderly patient.15 Howev- dures. The operative technique varied among surgeons with
er, to our knowledge, no large series has yet determined respect to kind of perfusion, myocardial protection, temper-
whether renal dysfunction is a risk factor for morbidity and ature, and left ventricular venting. The selection of valve
mortality in octogenarians compared with septuagenarians, prosthesis was based on surgeon’s and patient’s preference.
who are quite prevalent among the surgical population in Cardiac anesthesia was performed according to the institu-

developed countries. tion’s guidelines.
In the present report, we tested the hypothesis that
octogenarians develop more frequently kidney failure com- Statistical Analysis
pared with a matched cohort of septuagenarians undergoing Statistical analysis was performed with SPSS 16.0 (SPSS Inc,
cardiac surgical procedures. Chicago, IL) for Windows (Microsoft Corp, Redmond, WA) and
Stata SE 10.1 for Windows (StataCorp, College Station, TX).
Descriptive statistics were used to describe patient character-
Patients and Methods
istics and to compare variables. Continuous data were first
Patients and Study Design tested for normal distribution by Q-Q plots. Means and
We studied 598 consecutive patients 70 years who standard deviations were computed for normally distributed
underwent elective cardiac surgery between January 2006 continuous variables, whereas medians and interquartile
and August 2009 . The study had a retrospective, observa- ranges were used to describe nonnormally distributed con-
tional, and nonrandomized design. Preoperative demo- tinuous data. The Mann–Whitney U-test was used for com-
graphic profiles, operative and postoperative data were parison of nonnormally distributed data and Student’s t-test
obtained from the institutional database and medical re- was used for normally distributed data. Categorical data are
cords. The requirement of individual patient consent was shown as frequency distributions (n) and percentages (%).
waived because of the study’s retrospective design and the Fisher’s exact test was used for categorical variables in a 2 2
data collection from routine care. The study sample com- table or χ²-test in a 2 3 table. Logistic regression analysis
prised data of 299 patients 80 years (including all helped to examine the relationship between potential risk
octogenarians undergoing cardiac surgery during study factors and operative mortality. All significant variables from
period), who were compared with a control group of 299 the univariate analysis were included to determine indepen-
patients between 70 and 79 years. Patients in the control dent risk factors for adverse outcome. Goodness of fit was
group were matched for gender and operative procedure. tested with the Hosmer–Lemeshow-test and a receiver oper-
Preoperative variables were defined as in the European ating characteristic analysis (ROC) was used to evaluate the
System for Cardiac Operative Risk Evaluation (Euro- discriminatory performance. Differences with a p-value of
SCORE).15 Preoperative serum creatinine (SCr) was mea- < 0.05 were considered to be statistically significant.
sured in μmol/L at the day of hospital admission and
then converted to mg/dL (1 mg/dL ¼ 88.4 μmol/L). The
Results
estimated glomerular filtration rate (eGFR) was calculated
with the abbreviated Modification of Diet in Renal Disease Study Sample Demographics
(MDRD) for mula and expressed in mL/min/1.73 m 217: Detailed preoperative patient characteristics among the
MDRD-eGFR ¼ 186.3 SCr –1.154 age –0.203 0.742 (if study sample are listed in ►Table 1. The mean age of
female) octogenarians was 82 2.1 years (range 80 to 89 years)
Postoperative course was followed up in terms of cardiac compared with 74 2.8 years in septuagenarians (p < 0.05)
and especially renal status, detecting early mortality with with men and women being equally represented. Between
respect to age and renal impairment. The primary endpoint to both groups, no significant differences were found for preva-
this study was to determine if older age is associated with an lence of atrial fibrillation, COPD, diabetes mellitus, diabetic
increased risk of postoperative decrease in renal function nephropathy, and preoperative myocardial infarction. The
(decline of e GFR 50%) or acute renal failure. The secondary ejection fraction was nearly similar in both groups and within
purpose was to determine if renal insufficiency and acute the normal range. The difference regarding the mean logistic

Acute Renal Dysfunction Ried et al. 53
Table 1 Demographic Data Among the Study Sample
Variable Control Group (Age 70–79) Study Group (Age 80) p Value
Patients, No. 299 299
Age, mean SD [years] 74 2.8 82 2.1 < 0.001
Female (%) 137 (46) 137 (46) 1.00
COPD (%) 26 (8.7) 25 (8.4) 0.89
Atrial fibrillation (%) 40 (13.4) 43 (14.4) 0.81
Diabetes mellitus (%) 33 (11) 25 (8.4) 0.33
Diabetic nephropathy (%) 13 (4.4) 11 (3.7) 0.68
Ejection fraction, mean SD (%) 61 13.5 59 13.5 0.34
Logistic EuroSCORE† (%) 8.5 (7.7 to 9.4) 13.2 (12.1 to 14.4) < 0.001

Myocardial infarction preoperatively (%) 72 (24.1) 85 (28.4) 0.27
SCr on admission‡ (mg/dL) 1 (0.8; 1.2) 1.1 (0.9;1.3) 0.045
Estimated GFR < 60 103 (34.4) 131 (44) 0.022
(mL/min/1.73 m2) (%)

Used for multivariate analysis.
†
Data are shown as mean with 95% confidence interval.
‡
Data are shown as median plus interquartile range.
COPD, chronic obstructive pulmonary disease; GFR, glomerular filtration rate; SCr, serum creatinine; SD, standard deviation.
EuroSCORE reflected the higher estimated perioperative risk Operative Data

in the octogenarians (13.2 vs. 8.5%; p < 0.05), which is ►Table 2 summarizes the operative procedures within the
mainly based on the advanced age. two study groups. The number of bypass grafting and isolated
Preoperative kidney function was significantly impaired in valve replacements did not differ. The differences for cardio-
patients older than 80 years. Median serum creatinine on pulmonary bypass time and aortic cross-clamp time were
admission was slightly higher in octogenarians (p ¼ 0.045) small and with little clinical relevance, but due to the large
and estimated median glomerular filtration rate differed sample size statistically significant between both groups
between both groups (62 mL/min/1.73 m2 vs. 70 mL/min/ (p < 0.05).
1.73 m2; p < 0.05). Total 44% of octogenarians had a preoper-
ative estimated GFR < 60 mL/min/1.73 m2 compared with Renal Complications and Mortality
34% in septuagenarians (p ¼ 0.022) (►Fig. 1). The incidence of postoperative temporary dialysis-depen-
dent acute kidney failure was 5.4% in the control group and
6.7% in the study group (p ¼ 0.60). Postoperative decline of
glomerular filtration rate > 50% was insignificant, but a
decline < 25% occurred significantly more often in septua-
genarians (40 vs. 30%; p ¼ 0.02) (►Fig. 2).
In-hospital mortality was significantly higher in septuage-
narians with a GFR < 60 mL/min/1.73 m2 than in patients with a
GFR > 60 mL/min/1.73 m2 (11.9 vs. 2.6%; p < 0.05). Accordingly,
septuagenarians with a preoperative GFR < 60 mL/min/1.73 m2
had a higher 30-day mortality compared with patients with a
GFR > 60 mL/min/1.73 m2 (10.9 vs. 3.1%; p ¼ 0.02) (►Fig. 3). On
the other hand, overall in-hospital and 30-day mortality in
octogenarians was 6.7 and 7.7% without significant differences
with respect to preoperative GFR. A logistic regression model
was developed to identify independent predictors for 30-day
mortality (►Table 3). Multivariate logistic regression showed
that irrespective of age (OR 1.02, 95% CI 0.95 to 1.09), an eGFR
< 60 mL/min/1.73 m2 (OR 2.2, 95% CI 1.12 to 4.33) and a
prolonged aortic cross-clamp time (OR 1.01, 95% CI 1.00 to 1.02)
were independently associated with increased mortality. The
Figure 1 Bar graph showing the preoperative estimated renal func-
final model showed excellent goodness of fit (Hosmer-Leme-
tion (eGFR) stratified for each age group and eGFR class. Proportion of
patients with an eGFR < 60 mL/min/1.73 m 2 (black bars) is significantly show test; χ² ¼ 2.49, p ¼ 0.48) and had fair discriminatory
higher (p ¼ 0.022) in octogenarians compared with septuagenarians. power (AUC ¼ 0.74; 95% CI 0.67 to 0.82).

Table 2 Perioperative Data
Variable Control Group (Age 70–79) Study Group (Age 80) p Value
Isolated CABG 123 (41.1) 123 (41.1) 1.00
Isolated Valve 99 (33.1) 99 (33.1) 1.00
CABG þ Valve 77 (25.8) 77 (25.8) 1.00

Bypass time (min) 95 (76; 122) 88 (70; 114) 0.012
Aortic cross clamp time (min)† 60 (46; 80) 57 (43; 75) 0.019
In-hospital mortality (%) 18 (6) 20 (6.7) 0.87
30-day mortality (%) 18 (6) 23 (7.7) 0.52
ICU stay (d)† 1 (1; 4) 2 (1; 4) 0.87
†
Hospital stay (d) 12 (10; 16) 13 (10; 17) 0.24
Postoperative respiratory failure (%) 25 (8.4) 27 (9) 0.89
Ventilation time (h)† 12 (9; 18) 12 (9; 19) 0.25
Re-Intubation (%) 18 (6) 23 (7.7) 0.52

Tracheotomy (%) 8 (2.7) 10 (3.3) 0.81
‡
Postoperative central neurological event (%) 8 (2.7) 16 (5.4) 0.14
Redo-thoracotomy (%) 30 (10) 16 (5.4) 0.05
Patients without postoperative transfusion of PRBC (%) 109 (36.5) 100 (33.5) 0.91

Data are shown as median with interquartile range.
†
Used for multivariate analysis.
‡
Including PRIND, TIA and Stroke.
CABG, coronary artery bypass grafting; ECC, extracorporeal circulation; ICU, intensive care unit; PRBC, packed red blood cell.
Morbidity and Mortality otomy, or noninvasive ventilation was the most common
Detailed postoperative characteristics are also shown complication of patients 80 years (9 vs. 8.4%; p ¼ 0.89).
in ►Table 2. The overall in-hospital mortality was 6% in Postoperative stroke was observed in 2.7% of septuagenar-
septuagenarians compared with 6.7% in patients 80 years ians and in 5.4% of octogenarians, but without statistical
(p ¼ 0.87). Thirty-day mortality was slightly higher in significance.
octogenarians than in the control group (7.7 vs. 6%; p ¼ However, although we observed a lower actual 30-day
0.52). Length of the ICU and hospital stay were somewhat mortality in both groups, they did not statistically differ from
longer in the study group (p ¼ 0.87 and p ¼ 0.24). Postop- the expected mortality (septuagenarians: 6 vs. 8.5%, p ¼ 0.37;
erative respiratory failure with the need of prolonged octogenarians: 7.7 vs. 13.2%, p ¼ 0.10) as calculated with the
mechanical ventilation, re-intubation, subsequent trache- EuroSCORE.
Figure 2 Postoperative decline of estimated GFR in octogenarians

and septuagenarians. There is a statistical significant difference in the Figure 3 Thirty-day mortality (%) of both study and control group
< 25% decline group. with respect to preoperative GFR.

Acute Renal Dysfunction Ried et al. 55
Table 3 Independent Risk Factors for 30-Day Mortality
Variable Odds Ratio 95% Confidence Interval p Value

Age 1.02 0.95–1.09 0.60
Aortic cross-clamping (min) 1.01 1.00–1.02 0.026
Estimated GFR < 60 2.20 1.12–4.33 0.023
(mL/min/1.73 m2)
Discussion be more used to their chronic-reduced renal function and

they might not be suffering so often from a worse renal
Acquired heart disease is a leading cause of death among the function postoperative and without causing any relevant
elderly population. Despite relevant comorbidities, such as clinical consequences. Postoperative acute kidney injury
renal insufficiency, COPD, peripheral vascular disease, and may currently also be well defined according to the RIFLE
degenerative cerebral disease, the demand for cardiac surgery criteria (Risk (increase in Serum Creatinine (SCr) >1.5),
in patients > 70 years has steadily increased over the last Injury (increase in SCr > 2.0), Failure (increase in SCr >
decade and a further percent rise is even expected in the next 3.0), and Loss (end-stage kidney disease).22

years.1,2 Our early mortality rate was 6.7% in hospital and 7.7%
But octogenarians are reported to have increased inciden- within 30 days in octogenarians who underwent isolated or
ces of postoperative complications and significantly higher combined cardiac surgical procedures. These results are
rates of mortality compared with younger patients after within the range of mortality rates (4.4 to 13%) previously
primary cardiac surgery.16,17 On the other hand, cardiac reported for octogenarians6,8,23,24 and do not differ from the
surgery in selected octogenarians can be performed with expected mortality calculated with the EuroSCORE. It is
good survival rates and complication rates that were similar suggested that advanced age per se is not associated with
to those of septuagenarians.5,9 Thus, the decision for surgery an excessive operative risk.25 Our logistic regression model
still remains complex in this high-risk group of patients. could confirm this finding and also identified renal insuffi-
Although many studies have reported that renal dysfunc- ciency as potential risk factor for an adverse outcome. Al-
tion is one of the most important risk factors for perioperative though advanced age and renal dysfunction are commonly
morbidity and mortality after cardiac surgery,10,13–15 there used predictors for risk stratification in cardiac surgery, they
are no studies which examined especially the incidence of alone define not a contraindication for an operative treat-
acute kidney injury and the effects of impaired renal function ment. Instead, individual patient characteristics and comor-
on mortality in octogenarians compared with a control group bidities form the basis for adequate and individual risk
of septuagenarians. Previous studies described slightly higher stratification for patients at advanced age. Early risk identifi-
rates of postoperative renal failure in octogenarians and cation, enhanced renal protection strategies, and careful
identified it as an independent risk factor for early mortali- individual patient selection may further improve early and
ty.6,18 Generally, even a small increase in serum creatinine late outcome of septuagenarians and octogenarians after
within 48 hours in the postcardiac surgery setting seems to cardiac operations. The EuroSCORE was developed among
confer an adverse prognosis with a higher risk of death.11 patients having cardiac surgery in 1995. Thus, a current risk
Because serum creatinine seems to be an inadequate screen- stratification systems such as the EuroSCORE requires further
ing method for renal dysfunction in the elderly population,19 development to adjust for an increasing number of patients
we used the Modification of Diet in Renal Disease (MDRD) with advanced age and more complex comorbidities.
formula for the estimated glomerular filtration rate (eGFR).
This is the most widely used formula and is recommended by
the National Kidney Foundation Kidney Disease Outcomes
Strengths and Limitations of the Study
Quality Initiative chronic kidney disease guidelines. Chronic This was a contemporary single-center analysis of developing
kidney disease is defined as an eGFR less than 60 mL/min/1.73 postoperative acute kidney failure in octogenarians com-
m2.20 Outcome depends on the severity of renal impairment pared with septuagenarians. We are aware of the potential
and a strong correlation between age and deterioration of restrictions of our study. All data were revalidated by another
renal function was suggested.12,21 Postoperative requirement coauthor before analysis. This procedure should have reduced
of dialysis (10% septuagenarians, 7% octogenarians) predicted errors, but did not eliminate them completely. The obvious
a worse outcome and is suggested to be a prominent marker limitation is its retrospective and nonrandomized design. All
of death.5 Although octogenarians in our sample had preop- patients from our institutional database were screened to
eratively an impaired renal function (eGFR < 60 mL/min/1.73 select and match patients by gender and operation. Thus,
m2) more frequently, we found no significant differences there may be a risk of selection bias. We were not able to
between both study groups regarding postoperative renal obtain detailed data on perioperative blood volume manage-
complications, particularly the incidence of temporary dialy- ment and fluid balance. Furthermore, we could not provide
sis-dependent acute kidney failure. Thus, octogenarians may data on long-term survival. This study was conducted at a

single institution, which may limit the generalization of its moderate to severe preoperative renal dysfunction without dialy-
results to other centers. sis. Interact Cardiovasc Thorac Surg 2008;7(1):90–95
11 Lassnigg A, Schmid ER, Hiesmayr M, et al. Impact of minimal
increases in serum creatinine on outcome in patients after cardio-
Conclusions thoracic surgery: do we have to revise current definitions of acute
renal failure? Crit Care Med 2008;36(4):1129–1137
Octogenarians do not develop acute kidney failure more 12 van Straten AHM, Soliman Hamad MA, van Zundert AAJ, Martens
frequent than their matched septuagenarian counterparts EJ, Schönberger JPAM, de Wolf AM. Preoperative renal function as
after cardiac surgery. They can be operated on at an accept- a predictor of survival after coronary artery bypass grafting:
comparison with a matched general population. J Thorac Cardio-
able risk for morbidity and mortality. Preoperative-impaired
vasc Surg 2009;138(4):971–976
renal function seems to be associated with higher risk for
13 Diez C, Mohr P, Kuss O, Osten B, Silber RE, Hofmann HS. Impact of
mortality in septuagenarians. Therefore, it is clinically impor- preoperative renal dysfunction on in-hospital mortality after
tant to detect elderly patients with an impaired renal function solitary valve and combined valve and coronary procedures.
before surgery. Selected octogenarians should not be denied Ann Thorac Surg 2009;87(3):731–736
the benefits of early operative treatment for coronary heart 14 Simon C, Luciani R, Capuano F, et al. Mild and moderate renal
dysfunction: impact on short-term outcome. Eur J Cardiothorac
and/or valve disease. Future multicenter studies should de-
Surg 2007;32(2):286–290
termine treatment strategies that may improve the postop- 15 Nashef SA, Roques F, Michel P, Gauducheau E, Lemeshow S,
erative outcome of elderly patients, especially in patients Salamon R. European system for cardiac operative risk evaluation
aged older than 70 years (septuagenarians) and preoperative (EuroSCORE). Eur J Cardiothorac Surg 1999;16(1):9–13

presence of renal dysfunction. 16 Bardakci H, Cheema FH, Topkara VK, et al. Discharge to home
rates are significantly lower for octogenarians undergoing cor-
onary artery bypass graft surgery. Ann Thorac Surg 2007;83(2):
483–489
References 17 Böning A, Lutter G, Mrowczynski W, et al. Octogenarians under-
1 Gummert JF, Funkat A, Beckmann A, et al. Cardiac surgery in going combined aortic valve replacement and myocardial revas-
Germany during 2009. A report on behalf of the German Society cularization: perioperative mortality and medium-term survival.
for Thoracic and Cardiovascular Surgery. Thorac Cardiovasc Surg Thorac Cardiovasc Surg 2010;58(3):159–163
2010;58(7):379–386 18 Zingone B, Gatti G, Rauber E, et al. Early and late outcomes of
2 Friedrich I, Simm A, Kötting J, Thölen F, Fischer B, Silber RE. Cardiac cardiac surgery in octogenarians. Ann Thorac Surg 2009;87(1):71–
surgery in the elderly patient. Dtsch Arztebl Int 2009;106 78
(25):416–422 19 Swedko PJ, Clark HD, Paramsothy K, Akbari A. Serum creatinine is
3 Alexander KP, Anstrom KJ, Muhlbaier LH, et al. Outcomes of cardiac an inadequate screening test for renal failure in elderly patients.
surgery in patients > or ¼ 80 years: results from the National Arch Intern Med 2003;163(3):356–360
Cardiovascular Network. J Am Coll Cardiol 2000;35(3):731–738 20 National Kidney Foundation. K/DOQI clinical practice guidelines
4 Kolh P, Kerzmann A, Lahaye L, Gerard P, Limet R. Cardiac surgery in for chronic kidney disease: evaluation, classification, and stratifi-
octogenarians; peri-operative outcome and long-term results. Eur cation. Am J Kidney Dis 2002;39(2, Suppl 01):S1–S266
Heart J 2001;22(14):1235–1243 21 Brown JR, Cochran RP, MacKenzie TA, et al; Northern New England
5 Engoren M, Arslanian-Engoren C, Steckel D, Neihardt J, Fenn-Buderer Cardiovascular Disease Study Group. Long-term survival after
N. Cost, outcome, and functional status in octogenarians and sep- cardiac surgery is predicted by estimated glomerular filtration
tuagenarians after cardiac surgery. Chest 2002;122(4):1309–1315 rate. Ann Thorac Surg 2008;86(1):4–11
6 Melby SJ, Zierer A, Kaiser SP, et al. Aortic valve replacement in 22 Chiappini B, Camurri N, Loforte A, Di Marco L, Di Bartolomeo R,
octogenarians: risk factors for early and late mortality. Ann Thorac Marinelli G. Outcome after aortic valve replacement in octoge-
Surg 2007;83(5):1651–1656, discussion 1656–1657 narians. Ann Thorac Surg 2004;78(1):85–89
7 Fruitman DS, MacDougall CE, Ross DB. Cardiac surgery in octoge- 23 Collart F, Feier H, Kerbaul F, et al. Valvular surgery in octogenar-
narians: can elderly patients benefit? Quality of life after cardiac ians: operative risks factors, evaluation of Euroscore and long term
surgery. Ann Thorac Surg 1999;68(6):2129–2135 results. Eur J Cardiothorac Surg 2005;27(2):276–280
8 Huber CH, Goeber V, Berdat P, Carrel T, Eckstein F. Benefits of 24 Kolh P, Kerzmann A, Honore C, Comte L, Limet R. Aortic valve
cardiac surgery in octogenarians—a postoperative quality of life surgery in octogenarians: predictive factors for operative and
assessment. Eur J Cardiothorac Surg 2007;31(6):1099–1105 long-term results. Eur J Cardiothorac Surg 2007;31(4):600–606
9 Jokinen JJ, Hippeläinen MJ, Hänninen T, Turpeinen AK, Hartikainen 25 Craver JM, Puskas JD, Weintraub WW, et al. 601 octogenarians
JEK. Prospective assessment of quality of life of octogenarians after undergoing cardiac surgery: outcome and comparison with youn-
cardiac surgery: factors predicting long-term outcome. Interact ger age groups. Ann Thorac Surg 1999;67(4):1104–1110
Cardiovasc Thorac Surg 2008;7(5):813–818
10 Filsoufi F, Rahmanian PB, Castillo JG, Chikwe J, Carpentier A, Adams
DH. Early and late outcomes of cardiac surgery in patients with

Schloss et al. Journal of Cardiothoracic Surgery 2011, 6:103
http://www.cardiothoracicsurgery.org/content/6/1/103
RESEARCH ARTICLE Open Access
Impact of aprotinin and renal function on

mortality: a retrospective single center analysis
Brian Schloss1, Parul Gulati2, Lianbo Yu2, Mahmoud Abdel-Rasoul2, William O’Brien3, Jon Von Visger4 and
Hamdy Awad1*
Abstract
Background: An estimated up to 7% of high-risk cardiac surgery patients return to the operating room for
bleeding. Aprotinin was used extensively as an antifibrinolytic agent in cardiac surgery patients for over 15 years
and it showed efficacy in reducing bleeding. Aprotinin was removed from the market by the U.S. Food and Drug
Administration after a large prospective, randomized clinical trial documented an increased mortality risk associated
with the drug. Further debate arose when a meta-analysis of 211 randomized controlled trials showed no risk of
renal failure or death associated with aprotinin. However, only patients with normal kidney function have been
studied.
Methods: In this study, we look at a single center clinical trial using patients with varying degrees of baseline
kidney function to answer the question: Does aprotinin increase odds of death given varying levels of preoperative
kidney dysfunction?
Results: Based on our model, aprotinin use was associated with a 3.8-fold increase in odds of death one year later
compared to no aprotinin use with p-value = 0.0018, regardless of level of preoperative kidney dysfunction after
adjusting for other perioperative variables.
Conclusions: Lessons learned from our experience using aprotinin in the perioperative setting as an antifibrinolytic
during open cardiac surgery should guide us in testing future antifibrinolytic drugs for not only efficacy of
preventing bleeding, but for overall safety to the whole organism using long-term clinical outcome studies,
including those with varying degree of baseline kidney function.
Keywords: complex cardiac surgery, aprotinin, bleeding, renal dysfunction, mortality, antifibrinolytic drugs
Background Connecticut), an antifibrinolytic agent, has been used

Approximately one million cardiac surgeries are per- extensively since a study showed that it reduced the
formed in the United States every year. Of these, about need for blood transfusions during repeat cardiac sur-
200,000 can be classified as complex procedures, such gery [2]. Since then, other clinical trials have confirmed
as repeat coronary artery bypass grafting (CABG), valve aprotinin’s efficacy in reducing the need for blood trans-
replacements, and combined CABG with valve repairs/ fusions during these high-risk cardiac procedures [3,4].
replacements. One of the reasons these procedures are The safety of aprotinin was brought into question in
labeled as complex is because they carry a significant 2006 when a study revealed an increased risk of renal
increased risk of perioperative bleeding. An estimated failure, myocardial infarction, and stroke [5]. In 2008,
2.98% to 6.96% of high-risk cardiac operation patients aprotinin was removed from the market after a large
return to the operating room due to bleeding [1]. Apro- prospective, randomized clinical trial documented an
tinin (Bayer Pharmaceutical Corporation, West Haven, increased mortality risk associated with the drug [6].
Further debate arose when a meta-analysis of 211 ran-
domized controlled trials showed no increased risk of
* Correspondence: Hamdy.Elsayed-Awad@osumc.edu
1
Department of Anesthesiology, The Ohio State University Medical Center, renal failure or death associated with aprotinin [4]. The
(410 West 10th Avenue), Columbus, (43210), USA ongoing debate about aprotinin’s safety prompted us to
Full list of author information is available at the end of the article
© 2011 Schloss et al; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons
Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in
any medium, provided the original work is properly cited.
Schloss et al. Journal of Cardiothoracic Surgery 2011, 6:103 Page 2 of 5
examine clinical outcomes from our own institution that Table 1 Baseline Characteristics of Patients, According to
asked two specific questions: 1) What association does Treatment Group
aprotinin have on all causes of mortality given varying No Aprotinin Aprotinin p-value
levels of preoperative kidney dysfunction, and 2) What Mean (SD) Mean (SD)
association does the drug have on all causes of mortality Age 61.1 (12.7) 64.5 (14.1) 0.045
rates one year later in the same group of patients? Height (cm) 171.6 (8.9) 171.6 (10.6) 0.999
Based on the previously mentioned studies, we proved Weight (kg) 91.1 (22.9) 86.0 (20.5) 0.062
that aprotinin increases the odds of death regardless of Lowest HCT on bypass 23.1 (4.8) 21.6 (4.7) 0.01
the level of a preoperative kidney dysfunction in these Blood glucose (on bypass) 231.6 (61.6) 245.8 (74.6) 0.10
respective cohort patients. Blood glucose 202.1 (59.9) 216.5 (80.01) 0.11
(48 hrs postoperative)
CPB duration 75.9 (56.9) 113.8 (64.3) < 0.001
Methods
RBC transfusion on bypass 1.27 (1.92) 2.13 (2.48)
This retrospective, single-center study compared aproti-
Prime volume 575.0 (205.2) 588.6 (324.3) 0.71
nin versus no aprotinin use during complex cardiac sur-
UOP on pump 261.5 (241.9) 287.3 (290.9) 0.51
gery between October 2003 and October 2005. The
No Aprotinin Aprotinin p-value
study was conducted at The Ohio State University Med- No. (%) No. (%)
ical Center in Columbus, Ohio. A total of 1,644 complex Male 79 (68.1) 94 (70.15) 0.73
cardiac procedures were done during this two-year per- Mild renal failure 45 (38.79) 45 (33.33) 0.55
iod, which included repeat CABG, valve replacements, Moderate renal failure 39 (33.62) 45 (33.33)
and combined CABG with valve repairs/replacements. Severe renal failure 32 (27.59) 45 (33.33)
Non-complex cases, such as primary coronary artery Diabetes 49 (42.24) 51 (38.06) 0.50
bypass surgery, were specifically excluded in an attempt COPD none 100 (86.21) 115 (85.19) 0.69
to avoid the statistical bias that aprotinin tends to be COPD mild 7 (6.0) 7 (5.2)
used in more complex surgeries, which inherently carry COPD mod 8 (6.9) 9 (6.7)
a greater morbidity and mortality risk [7]. The patient COPD severe 1 (0.9) 4 (3.0)
received the drug or no drug per the surgeons’ request. Hypertension 90 (77.6) 110 (81.5) 0.44
After obtaining approval from our institutional review Diabetes 49 (42.2) 51 (38)
board, we retrieved perioperative data from our institu- Peripheral vascular disease 17 (14.7) 28 (20.7) 0.21
tion’s thoracic surgery, perfusion, and general electronic Myocardial infarction 44 (37.9) 52 (38.5) 0.92
medical record databases. From the 1,644 cases, the Congestive heart failure 42 (36.2) 83 (61.5) < 0.001
Center for Biostatistics randomly selected 251 with vary- One year mortality 20 (17.2) 43 (32.1) 0.007
ing degrees of renal dysfunction for analysis.
COPD - chronic obstructive pulmonary disease; CPB - cardiopulmonary bypass;
Twenty data points per patient were collected, includ- HCT - hematocrit; RBC - red blood cell; UOP - urine output
ing preoperative and postoperative kidney function,
patient demographics, medical comorbidities, intrao-
perative variables, aprotinin administration, postopera- using Student t tests or Wilcoxon rank sum test where
tive hemodialysis requirements, and one-year mortality appropriate. The relationship between one-year mortal-
(Table 1). Glomerular filtration rate, a measure for kid- ity as the outcome variable and aprotinin treatment was
ney function, was estimated using the Modification of analyzed using a multivariate logistic regression model
Diet in Renal Disease study equation formula. This took that adjusted for other variables determined to be signif-
into account serum creatinine measurement, age, sex, icantly related to mortality. Additional variables were
and race. The estimated glomerular filtration rate checked for potential confounding or effect modification
(eGFR) calculation was recorded in milliliters per min- but did not make it into the final model as their effect
ute. The decision to treat with aprotinin was based on on the relationship between aprotinin treatment and the
surgeon preference. Follow-up data, including all causes outcome was minimal. We also adhered to the general
of mortality at one year, were obtained using records guideline to include no more than one variable per 10
from the electronic medical record database. patients in the group that experienced the outcome
event of interest. All statistical analyses were performed
Statistical analysis using SAS 9.2 (SAS, Carey, N.C.).
Categorical demographic and clinical characteristics of
patients were compared between the treatment and con- Results
trol groups using Chi-squared or Fisher exact tests as A total of 1,644 patients underwent complex cardiac
appropriate (Table 1). Continuous characteristics were surgery between 2003 and 2005. From this population,
compared between the treatment and control groups 251 were randomly selected for analysis. From this
group, 39 were excluded from the final model because Table 3 Results of Multivariate Logistic Regression for
of incomplete data sets. A total of 212 subjects were One Year Postoperative Mortality in 184 Patients*
included in our final statistical model. Effect Point Estimate 95% CI p-value
Statistical analysis revealed three factors, other than Aprotinin vs. control 6.474 2.270 18.469 0.0005
aprotinin, that were significantly associated with mortal- Diabetes vs. no Diab 2.304 0.941 5.645 0.0679
ity (Table 2). These factors were controlled in a multi- urine_output 0.996 0.994 0.999 0.0019
variate logistic regression model. Based on this model, Prime_Vol 1.002 1.000 1.003 0.0391
aprotinin use was associated with a 3.8-fold increase in Blood_Gluc_High_48_h 1.007 1.000 1.013 0.0498
odds of death compared to no aprotinin use (p = RBC_trans 1.422 1.176 1.720 0.0003
0.0018) regardless of the level of preoperative kidney Initial_Creatinine_m 0.595 0.364 0.974 0.0388
dysfunction after adjusting for other perioperative vari- CI - confidence interval; Gluc - glucose; RBC - red blood cell; Vol - volume
ables. The other three variables found to be significantly *Excluded were 67 patients with missing values for at least one of the
associated with death were diabetes, packed red blood covariates in the model. The Hosmer-Lemeshow goodness of fit chi-square
test statistic was 3.62 (P = 0.89).
cell transfusion on cardiopulmonary bypass, and 48-
hour postoperative eGFR. In our model, diabetes was
associated with a 2.2-fold increase in odds of death associated with worsening preoperative kidney dysfunc-
compared with non-diabetic patients (p = 0.0312) (Table tion [8,9], our model found no interaction between
3). For packed red blood cell transfusion, the odds of aprotinin and preoperative kidney dysfunction. Thus,
death increased by 28% for every unit given while on our data supports the hypothesis that aprotinin use was
bypass (p = 0.0018). Lastly, for every one unit (ml/min) associated with a 3.8-fold increase in odds of death
increase in eGFR, the odds of death decreased by 2.4%. compared to no aprotinin use (p = 0.0018) regardless of
As expected, patients had varying levels of kidney dys- the level of a preoperative kidney dysfunction after
function preoperatively, though level of kidney dysfunc- adjusting for other perioperative variables.
tion did not differ significantly between the two groups. This finding coincides with the Blood Conservation
In our model, preoperative kidney function as a contin- Using Antifibrinolytics in a Randomized Trial [6] and
uous or categorical variable did not significantly interact further supports the decision to remove aprotinin from
with aprotinin. This indicates that the increased odds of the market. Our results support the opinion that the
death in aprotinin-treated patients were the same across ability of aprotinin to reduce blood loss during complex
each level of kidney function. cardiac surgery does not outweigh the risk of death
The rates of postoperative hemodialysis were low for associated with the drug. Furthermore, the continued
both drug and no drug groups. Only one patient in the availability of the lysine analogues, such as aminocaproic
no drug group and two patients in the aprotinin group acid and tranexamic acid, lends little credibility to the
required postoperative hemodialysis. continued use of aprotinin. Though there is no clear
data showing the lysine analogues to be equally as effi-
Discussion cacious as aprotinin [7,10], the risks of aprotinin likely
The primary new finding in our study is that aprotinin do not outweigh the benefits.
use, irrespective of the level of preoperative renal dys- There has been a certain degree of controversy over
function, was associated with a 3.8-fold increase in odds aprotinin’s effect on kidney function postoperatively.
ratio of death one year postoperatively. While previous Several previous retrospective studies have shown apro-
studies have documented an increase in mortality tinin to be associated with an increased risk of post-
operative renal failure [11,12]. In contrast, the only
prospective, randomized, placebo-controlled trial to
Table 2 Multivariate Logistic Regression for One Year investigate aprotinin’s effects on kidney function post-
Postoperative Mortality in 184 Patients Showing operatively showed no significant difference between
Aprotinin Increased Death* drug and control groups [13]. Furthermore, several ret-
Effect Odds Ratio (95% CI) p-value rospective studies have shown no increased incidence of
Aprotinin vs. control 3.830 (1.649, 8.893) 0.0018 dialysis in aprotinin-treated patients compared with con-
Diabetes 2.236 (1.075, 4.651) 0.0312 trol [14,15]. Our study showed no increase in hemodia-
PRBC transfusion on bypass 1.282 (1.096, 1.498) 0.0018 lysis in the aprotinin versus no aprotinin group.
eGFR 48 hours postoperatively 0.976 (0.958, 0.994) 0.0079 However, given the low rates of hemodialysis in both of
CI - confidence interval; eGFR - estimated glomerular filtration rate; PRBC - our groups, it is possible that we missed a significant
packed red blood cells rise in risk of renal failure. It should also be noted that
* Excluded were 39 patients with missing values for at least one of the
covariates in the model. The Hosmer-Lemeshow goodness of fit chi-square
we were unable to obtain definitive causes of death dur-
test statistic was 3.62 (P = 0.89). ing our data collection.
There were several other variables that were found to Acknowledgements

We wish to acknowledge Thomas E. Williams, Jr., MD, PhD, clinical associate
be significantly associated with mortality. We discuss
professor of surgery in the Division of Cardiothoracic Surgery at The Ohio
these, as it was necessary to include them in our final State University Medical Center and Keri Hudec, technical editor with the
statistical model. We found that the presence of dia- Department of Anesthesiology at The Ohio State University Medical Center.
We also wish to thank Olivia Zvara and Melissa Harben for assisting with
betes, increased blood transfusions on bypass, and
data collection.
decreased eGFR all were associated with increased mor-
tality. These findings are not surprising, and similar Author details
1
Department of Anesthesiology, The Ohio State University Medical Center,
mortality associations have been demonstrated in the lit-
(410 West 10th Avenue), Columbus, (43210), USA. 2Center for Biostatistics,
erature for diabetes [16-19] and red blood cell transfu- The Ohio State University Medical Center, (2012 Kenny Road), Columbus,
sion [20-25]. (43210), USA. 3Perfusion Services, The Ohio State University Medical Center,
(452 W. 10th Avenue), Columbus, (43210), USA. 4Department of Nephrology,
Our findings, while significant and new, have some
The Ohio State University Medical Center, (395 West 12th Avenue),
limitations. The most significant of these limitations is Columbus, 43210, USA.
the retrospective nature of the study. Also, the small
Authors’ contributions
number of deaths and our inability to document a cause
BS analyzed the data and wrote the manuscript. PG, LY, MA, and WO
of death may have potentially masked an increased risk analyzed the data. JV analyzed the data and wrote the manuscript. HA
of renal dysfunction as well as any other major organ designed the study, analyzed the data and wrote the manuscript.
All authors read and approved the final manuscript.
system failures. In addition, our preoperative baseline of
acute kidney dysfunction is non-standardized since there Competing interests
is no agreed upon definition of grades of acute renal The authors declare that they have no competing interests.
dysfunction [26]. The aprotinin was administered to the
Received: 17 May 2011 Accepted: 30 August 2011
patient per surgeon preference. Published: 30 August 2011
Conclusions References
1. Society of Thoracic Surgery National Database: STS data analyses - January
The exact mechanism of increased mortality associated
1999: operative mortality summaries.[http://www.sts.org/national-
with aprotinin is unknown; however, it has been specu- database], [Accessed October 30, 2009].
lated that it may be related to its systemic effect on 2. Henry DA, Moxey AJ, Carless PA, O’Connell D, McClelland B, Henderson KM,
Sly K, Laupacis A, Fergusson D: Anti-fibrinolytic use for minimising
inflammation, oxidative stress, and bioactive molecules
perioperative allogeneic blood transfusion. Cochrane Database Syst Rev
in addition to the complex interaction between the 2007, 4:CD001886-CD001886.
homeostatic, complement, and cytokine pathways in the 3. Laupacis A, Fergusson D: Drugs to minimize perioperative blood loss in
cardiac surgery: meta-analyses using perioperative blood transfusion as
setting of ischemia and reperfusion [27,28]. Ray and
the outcome. Anesth Analg 1997, 85:1258-1267.
Stein [29] theorize that aprotinin’s potential increase in 4. Sedrakyan A, Treasure T, Elefteriades JA: Effect of aprotinin on clinical
mortality may simply be due to an over effectiveness of outcomes in coronary artery bypass graft surgery: a systematic review
and meta-analysis of randomized clinical trials. J Thorac Cardiovasc Surg
antifibrinolysis when compared to the lysine analogues.
2004, 128:442-448.
Perhaps the balance between procoagulatory and antic- 5. Mangano DT, Tudor IC, Dietzel C: The risk associated with aprotinin in
oagulatory effects is tipped too far in one direction by cardiac surgery. N Engl J Med 2006, 354:353-365.
6. Markin JM: Lessons learned in antifibrinolytic therapy: the BART Trial.
aprotinin. Novel pathways may exist that could be exert-
Semin Cardiothorac Vasc Anesth 2009, 13:127-131.
ing deleterious affects on major organ systems. Regard- 7. Bidstrup BP, Harrison J, Royston D, Taylor KM, Treasure T: Aprotinin therapy
less of speculation, it is clear that further research is in cardiac operations: a report on use in 41 cardiac centers in the
United Kingdom. Ann Thorac Surg 1993, 55:971-976.
necessary to elucidate the exact mechanisms that cur-
8. Blumberg N: Deleterious clinical effects of transfusion
rent and future antifibrinolytics exert on the whole immunomodulation: proven beyond a reasonable doubt. Transfusion
organism, including long-term effects. A more complete 2005, 45:33S-40S.
9. Mangano CM, Diamondstone LS, Ramsay JG, Aggarwal A, Herskowitz A,
understanding of their pathways will help to design
Mangano DT: Renal dysfunction after myocardial revascularization: risk
drugs in the future that are as effective as they are safe. factors, adverse outcomes, and hospital resource utilization. Ann Intern
We believe that the lessons learned from our experience Med 1998, 128:194-203, The Multicenter Study of Perioperative Ischemia
Research Group.
of using aprotinin as an antifibrinolytic during open
10. Carless PA, Moxey AJ, Stokes BJ, Henry DA: Are antifibrinolytic drugs equal in
heart surgery in the past should guide us in testing reducing blood loss and transfusion in cardiac surgery? A meta analysis of
future drugs, which will not only focus on their efficacy randomized head to head trials. BMC Cardiovasc Disord 2005, 5:19.
11. Augoustides JG, Pochettino A, Ochroch EA, Cowie D, Weiner J,
of preventing bleeding, but on overall safety to the
Gambone AJ, Pinchasik D, Bavaria JE, Jobes DR: Renal dysfunction after
whole organism using long-term clinical outcome thoracic aortic surgery requiring deep hypothermic circulatory arrest:
studies. definition, incidence, and clinical predictors. J Cardiothorac Vasc Anesth
2006, 20:673-677.
12. Karkouti K, Beattie WS, Dattilo KM, McCluskey SA, Ghannam M, Hamdy A,
Wijeysundera DN, Fedorko L, Yau TM: A propensity score case-control
List of Abbreviations
comparison of aprotinin and tranexamic acid in high-transfusion-risk
CABG: coronary artery bypass grafting; eGFR: estimated glomerular filtration
cardiac surgery. Transfusion 2006, 46:327-338.
rate
13. Lemmer JH Jr, Stanford W, Bonney SL, Chomka EV, Karp RB, Laub GW,
Rumberger JA, Schaff HV: Aprotinin for coronary artery bypass grafting:
effect on postoperative renal function. Ann Thorac Surg 1995, 59:132-136.
14. Levy JH, Pifarre R, Schaff HV, Horrow JC, Albus R, Spiess B, Rosengart TK,
Murray J, Clark RE, Smith P: A multicenter, double blind, placebo-
controlled trial of aprotinin for reducing blood loss and the requirement
for donor-blood transfusion in patients undergoing repeat coronary
artery bypass grafting. Circulation 1995, 92:2236-2244.
15. Fergusson DA, Hébert PC, Mazer CD, Fremes S, MacAdams C, Murkin JM,
Teoh K, Duke PC, Arellano R, Blajchman MA, Bussières JS, Côté D, Karski J,
Martineau R, Robblee JA, Rodger M, Wells G, Clinch J, Pretorius R, BART
Investigators: A comparison of aprotinin and lysine analogues in high-
risk cardiac surgery. N Engl J Med 2008, 358:2319-2331.
16. Szabo Z, Hakanson E, Svedjeholm R: Early postoperative outcome and
medium-term survival in 540 diabetic and 2239 nondiabetic patients
undergoing coronary artery bypass grafting. Ann Thorac Surg 2002,
74:712-719.
17. Engoren MC, Habib RH, Zacharias A, Schwann TA, Riordan CJ, Durham SJ:
Effect of blood transfusion on long term survival after cardiac operation.
Ann Thorac Surg 2002, 74:1180-1186.
18. Fransen E, Maessen J, Dentener M, Senden N, Buurman W: Impact of blood
transfusions on inflammatory mediator release in patients undergoing
cardiac surgery. Chest 1999, 116:1233-1239.
19. Habib RH, Zacharias A, Schwann TA, Riordan CJ, Durham SJ, Shah A:
Adverse effects of low hematocrit during cardiopulmonary bypass in the
adult: should current practice be changed? J Thorac Cardiovasc Surg 2003,
125:1438-1450.
20. Carson JL, Schloz PM, Chen AY, Peterson ED, Gold J, Schneider SH:
Diabetes mellitus increases short term mortality and morbidity in
patients undergoing coronary artery bypass graft surgery. J Am Coll
Cardiology 2002, 40:418-423.
21. Sobieski MA, Slaughter MS, Hart DE, Pappas PS, Tatooles AJ: Prospective
study on cardiopulmonary bypass prime reduction and its effect on
intraoperative blood product and hemoconcentrator use. Perfusion 2005,
20:31-37.
22. van de Watering LM, Hermans J, Houbiers JG, van den Broek PJ, Bouter H,
Boer F, Harvey MS, Huysmans HA, Brand A: Beneficial effects of leukocyte
depletion of transfused blood on postoperative complications in
patients undergoing cardiac surgery: a randomized clinical trial.
Circulation 1998, 97:562-568.
23. Kristeller JL, Roslund BP, Stahl RF: Benefits and risks of aprotinin use
during cardiac surgery. Pharmacotherapy 2008, 28:112-124.
24. Lazar HL, McDonnell M, Chipkin SR, Furnary AP, Engelman RM, Sadhu AR,
Bridges CR, Haan CK, Svedjeholm R, Taegtmeyer H, Shemin RJ, Society of
Thoracic Surgeons Blood Glucose Guideline Task Force: The Society of
Thoracic Surgeons practice guidelines series: blood glucose
management during adult cardiac surgery. Ann Thorac Surg 2009,
87:663-669.
25. DeFoe GR, Ross CS, Olmstead EM, Surgenor SD, Fillinger MP, Groom RC,
Forest RJ, Pieroni JW, Warren CS, Bogosian ME, Krumholz CF, Clark C,
Clough RA, Weldner PW, Lahey SJ, Leavitt BJ, Marrin CA, Charlesworth DC,
Marshall P, O’Connor GT: Lowest hematocrit on bypass and adverse
outcomes associated with coronary artery bypass grafting. Northern
New England Cardiovascular Disease Study Group. Ann Thorac Surg 2001,
71:769-776.
26. Conlon PJ, Stafford-Smith M, White WD, Newman MF, King S, Winn MP,
Landolfo K: Acute renal failure following cardiac surgery. Nephrol Dial
Transplant 1999, 14:1158-1162.
27. McEvoy MD, Reeves ST, Reves JG, Spinale FG: Aprotinin in cardiac surgery: Submit your next manuscript to BioMed Central
a review of conventional and novel mechanisms of action. Anesth Analg and take full advantage of:
2007, 105:949-962.
28. Asopa S, Elahi M: Hemodilution during cardiopulmonary bypass: optimal
• Convenient online submission
hematocrit for safe outcome. Acute Card Care 2007, 9:63-64.
29. Ray WA, Stein CM: The aprotinin story - is BART the final chapter? N Engl • Thorough peer review
J Med 2008, 358:2398-2400. • No space constraints or color figure charges
doi:10.1186/1749-8090-6-103 • Immediate publication on acceptance
Cite this article as: Schloss et al.: Impact of aprotinin and renal function • Inclusion in PubMed, CAS, Scopus and Google Scholar
on mortality: a retrospective single center analysis. Journal of
Cardiothoracic Surgery 2011 6:103. • Research which is freely available for redistribution
Submit your manuscript at

www.biomedcentral.com/submit
PERIOPERATIVE HEMATOLOGY: TO BLEED OR NOT TO BLEED
Pharmacologic tools to reduce bleeding in surgery

Sam Schulman1,2
1Department of Medicine, McMaster University, and 2Thrombosis and Atherosclerosis Research Institute,
Hamilton, ON
Strategies to reduce blood loss and the need for transfusions in surgery include enhancement of coagulation,
inhibition of fibrinolysis, and an improved decision algorithm for transfusion based on bedside monitoring of global
hemostasis. The synthetic antifibrinolytic drug tranexamic acid has emerged as an effective alternative in this respect
for orthopedic and cardiac surgery. Although it seems less effective than aprotinin, it has not been associated with the
increased risk of mortality of the latter. Thromboelastography to monitor the global hemostatic capacity and to guide
the appropriate use of blood components in cardiac surgery is also effective in reducing the need for transfusion.
Patients on antithrombotic drug therapy may need reversal before surgery to avoid excessive blood loss, or
intraoperatively in cases of unexpected bleeding. Available options are protamine for unfractionated or low-molecular-
weight heparin, recombinant activated factor VII for fondaparinux, prothrombin complex concentrate for vitamin K
antagonists and possibly for oral factor Xa inhibitors, dialysis and possibly activated prothrombin complex concentrate
for oral thrombin inhibitors, desmopressin for aspirin and possibly for thienopyridines, and platelet transfusions for the
latter.
Introduction Desmopressin
Reduction of operative blood loss is important to avoid reoperation, Another pharmacologic pathway that increases coagulant activity is
to avoid or minimize the need for blood transfusions, and to diminish the vasopressin analog D-arginine-deamino-vasopressin, which
postoperative anemia that could lead to congestive heart failure, elevates the levels of FVIII and VWF. There is also a transient
delayed wound healing and mobilization, and other complications. increase in fibrinolytic activity by the release of tissue plasminogen
Allogeneic blood transfusions have been associated with increased activator. In 19 trials with 1387 patients, there was no reduction in
mortality and risk of infection,1 and major bleeding events in patients the number of patients transfused with blood (relative risk
with acute coronary syndromes have been linked to worse prognoses, [RR] ⫽ 0.96; 95% CI, 0.87-1.06).5 Desmopressin did reduce the
with a 5-fold increment of mortality within 30 days.2 Although strict blood loss by a weighted mean of 242 mL (95% CI, ⫺388-⫺96) but
selection and screening of blood donors and donated blood have not the risk for reoperation (RR ⫽ 0.69; 95% CI, 0.26-1.83). There
drastically reduced the risk of transfusion-transmitted diseases, there is was no evidence of harm from desmopressin, but the positive effects
still a small residual risk. Additional concerns are febrile nonhemolytic were not convincing. It is possible that desmopressin has a more
transfusion reactions and transfusion errors. important role in patients with known disorders of the primary
hemostasis.
In addition to optimization of surgical techniques to reduce bleeding,
the use of hemostatic agents has also been explored. There are here Fibrinogen
essentially 2 types of hematostatic agents, procoagulant and antifibrino- In a randomized pilot trial, 20 patients undergoing elective coronary
lytic, as shown by the examples in Figure 1. The former has the artery bypass surgery with a fibrinogen level ⬍ 3.8 g/L received 2 g
intuitive peril of being prothrombotic, particularly in the state of of fibrinogen or no infusion before surgery.6 Postoperative blood
postoperative acute-phase reaction, with increased levels of factor VIII loss was 32% lower in the fibrinogen group (565 vs 830 mL,
(FVIII), VWF, and fibrinogen as well as fibrinolytic shut-down.3 respectively; P ⫽ .01) and there was one subclinical graft occlusion in
this group. Larger studies are obviously required to confirm these data.
Hemostatic agents
rFVII Antifibrinolytic agents
Recombinant activated FVIIa (rFVIIa) has been studied in random- Pharmacologic diminution of fibrinolytic activity can be achieved
ized trials in cardiac surgery, spinal surgery, liver transplanta- with the relatively specific synthetic lysine analogs tranexamic acid
tion, liver resection, retropubic prostatectomy, posttraumatic (TXA) and epsilon aminocaproic acid (EACA) or with a broad-
reconstruction of the pelvis, and dental extractions in patients spectrum serine protease inhibitor (eg, aprotinin). The lysine
with liver cirrhosis, all of which are off-label use.4 There is analogs block the lysine-binding sites on plasminogen, thereby
indeed an increased risk for arterial thromboembolic complica- preventing its activation to plasmin.
tions with rFVIIa compared with placebo (odds ratio [OR] ⫽ 1.68;
95% confidence interval [CI], 1.20-2.36). This increased risk was Aprotinin was associated with increased mortality compared with
observed particularly in elderly patients (OR ⫽ 3.02; 95% CI, TXA in a large randomized trial in high-risk cardiac surgery
1.22-7.48) and the risk is distributed over all off-label indications.4 (RR ⫽ 1.55; 95% CI, 0.99 - 2.42),7 which led to the withdrawal of
Therefore, prophylaxis with rFVIIa should not be used to reduce aprotinin from the market in 2008. EACA has been associated with
blood loss. hypotension, cardiac arrhythmias, myopathy, and rhabdomyolysis,
Hematology 2012 517

Figure 1. Reduction of blood loss. Blood loss reduction can be achieved by the enhancement of fibrin formation with rFVIIa to increase the formation
of FIXa and FXa(a), with desmopressin to stimulate the release of FVIII and VWD(b), or by providing more fibrinogen(c). Another strategy is to decrease
the degradation of fibrin by inhibiting the conversion of plasminogen to plasmin with aprotinin, TXA, or EACA(d).
but is the only available antifibrinolytic agent in some places. TXA the case for aprotinin and then only in revision of total hip prosthesis.12
is the most promising alternative due to a favorable benefit-risk Each of the 2 agents reduced blood loss and there was no increase in
ratio, experience from several decades of use for most types of thromboembolic events.
bleeding or surgery in patients with congenital or acquired bleeding
disorders,8,9 and the 10 times lower dose required than with EACA. For total knee replacement, 19 studies specifically with TXA were
The drug has been studied mainly in orthopedic and cardiac surgery. eligible and the need for transfusion was reduced (risk ratio ⫽ 0.39;
95% CI, 0.32-0.48) and there was a mean reduction of postoperative
In a systematic review of randomized trials in the setting of elective blood loss (245 mL; 95% CI, 213-278)13 and total blood loss
surgery and with any of the 3 antifibrinolytic drugs compared with
placebo or with direct comparisons between them, 252 studies with Table 1. Efficacy and safety of antifibrinolytic agents in elective
more than 25 000 patients were included.10 The main results are surgery based on a systematic review10
summarized in Table 1. The investigators pointed out the heteroge-
Comparison RR 95% CI
neity of the data, probably due to publication bias. There has also
been a range of doses used in the trials (Table 2). Need for blood transfusion
Aprotinin vs control 0.66 0.60-0.72
TXA vs control 0.61 0.53-0.70
Orthopedic surgery EACA vs control 0.81 0.67-0.99
Antifibrinolytic agents in general reduce the need for blood transfusion Aprotinin vs TXA or EACA 0.90 0.81-0.99
in elective knee- or hip-replacement surgery, as demonstrated in a Need for reoperation
systematic review.11 The RR compared with control or placebo was Aprotinin vs control 0.46 0.34-0.62
0.52 (95% CI, 0.42-0.64) and no increase in the RR for venous TXA vs control 0.80 0.55-1.17
thromboembolism could be identified (RR ⫽ 0.95; 95% CI, 0.80-1.10). EACA vs control 0.32 0.11-0.99
Each of the 3 antifibrinolytic agents was associated with a significantly Mortality
smaller blood loss, which was most prominent with aprotinin. Meta- Aprotinin vs control 0.81 0.63-1.06
analyses have also been performed separately for various orthopedic Aprotinin vs TXA or EACA 1.39 1.02-1.89
surgeries. Thirteen studies on total hip arthroplasty were evaluated and, Myocardial infarction
Aprotinin vs control 0.87 0.69-1.11
although there was a reduction in the need for blood transfusion when
Aprotinin vs TXA or EACA 1.11 0.82-1.50
results for TXA and aprotinin were pooled, individually, this was only
518 American Society of Hematology

Table 2. Dose ranges used with the antifibrinolytic drugs in taking many hours, this could be followed by a maintenance
randomized controlled trials infusion of 1 mg/kg/h.
Agent Bolus dose Maintenance dose
Reversal of antithrombotic agents
Aprotinin 25 000 KIU/kg or 2.5-5 ⫻ 105 KIU/h
5 ⫻ 105-2 ⫻ 106 KIU
Antithrombotic drug use is becoming more prevalent due to the
EACA 75-150 mg/kg or 12.5-30 mg/kg/h or aging population with its higher prevalence of atrial fibrillation, and
80 mg-15 g 1-2 g/h to increasing awareness of the need for prophylaxis against isch-
TXA 2.5 - 100 mg/kg 0.25-4 mg/kg/h emic stroke in this condition. Antiplatelet agents, often in combina-
tions, are used long-term after insertion of drug-eluting stents.
KIU indicates kallikrein-inhibiting units.
Ideally, a planned interruption, possibly with the use of a bridging
agent with short half-life, has been arranged for surgery or other
(591 mL; 95% CI, 536-647). These benefits were gained without invasive procedures. This is discussed in the article by Ortel.18 For
any increase in the risk for deep vein thrombosis with TXA. One emergency surgery in a patient known to be treated with an
study had a high incidence of deep vein thrombosis (45%-48%), antithrombotic agent and elective surgery with planned continuation
which was detected with a radioisotope test and no thromboprophy- of the antithrombotic drug(s) but with unexpected bleeding ten-
laxis had been given.14 The heterogeneity in the data was substan- dency, there is a need for identification of the most responsible drug,
tial except for the positive effect in a subset receiving high-dose but in the second scenario, it is also important to determine whether
( ⬎ 4 g) TXA, in whom it was consistent. it is the drug, the surgical condition, or an underlying bleeding
disorder that is the main cause.
For hip fracture surgery, one randomized trial compared TXA at a
dose of 15 mg/kg with placebo in 110 patients.15 The study drug was Intraoperative assessment of cause of bleeding
given at start of surgery and repeated 3 hours later. There was a It should be relatively easy for the surgeon to exclude local etiology,
trend to reduced need for blood transfusions with TXA versus such as abnormal blood vessels, tumor, infection or other conditions
placebo (42% and 60%, respectively; P ⫽ .06). This difference causing technical difficulties, from the cause of bleeing. The
became statistically significant when the investigators pooled their anesthesiologist should explore if any medical or laboratory informa-
results with a previous study in hip-fracture surgery (OR ⫽ 0.47; tion that would indicate the type of systemic congenital or acquired
95% CI, 0.26-0.85). The investigators also pointed out that in their bleeding tendency has been missed. A blood sample for emergent
study, unlike studies of TXA for elective orthopedic surgery, there screening for disseminated intravascular coagulation should be sent
was an increase in a composite of different vascular events with off, but the lag time until results are obtained is often problematic.
16% versus 6% on placebo, but this was not statistically significant Point-of-care instruments for either single tests (eg, activated
(hazard ratio ⫽ 2.96; 95% CI, 0.80-11.0). clotting time or prothrombin time) or for global coagulation analysis
such as thromboelastography are then very helpful. The objective is
In major spine surgery, 18 trials with 966 patients were eligible for to obtain information on what part of the hemostatic system is
inclusion in a meta-analysis of the effect of the 3 antifibrinolytic failing. Thromboelastography to guide the transfusion of blood
agents.16 There was a decrease in both the blood loss and the need components has been compared with standard management, (ie,
for transfusions with each of the agents, and this was most sending samples to the laboratory), in 5 randomized controlled
prominent with EACA. No increase in the risk of venous thrombo- studies.19-23 Simple pooling of the results regarding the proportion
embolism was detected. of patients transfused with RBCs, plasma, or platelets shows a
benefit for point-of-care monitoring (Table 3). The mean postopera-
Cardiac surgery tive or total blood loss was also numerically lower in this group in
All 3 antifibrinolytic agents were evaluated in a meta-analysis of all studies. Although evidence is so far lacking for improved clinical
49 trials in cardiac surgery and indirect comparisons derived from outcomes, conservation of blood products is also of importance and
studies against placebo or control and direct comparisons from therefore employment of intraoperative coagulation assessment
head-to-head trials were performed.17 The need for transfusion was with thromboelastography is recommended.
reduced with TXA (RR ⫽ 0.70; 95% CI, 0.61-0.80), EACA
(RR ⫽ 0.75; 95% CI, 0.58-0.96), and aprotinin (RR ⫽ 0.66; Heparins
95% CI, 061-0.72), although the need for reoperation was only Unfractionated heparin is completely reversed by protamine sulfate,
reduced significantly with aprotinin (RR ⫽ 0.48; 95% CI, 0.34- with which it forms a stable salt complex. Protamine is given at a
0.67). Although aprotinin did not increase the risk for death dose of 1 mg/100 units of heparin and the circulating amount of the
compared with placebo (RR ⫽ 0.93; 95% CI, 0.69-1.25), the point latter should be estimated based on last dose and half-life. The
estimate was higher in the indirect comparison with TXA maximum infusion rate is 50 mg/10 min and the effect of reversal
(RR ⫽ 1.69; 95% CI, 0.70-4.10), with a strong statistical trend in
the direct comparison (RR ⫽ 1.43; 95% CI, 0.98-2.08) and with Table 3. Pooled data of the number of patients transfused with
similar numbers in the direct comparison against EACA. There was blood components in randomized trials comparing
no increase in the risk of myocardial infarction with these agents. thromboelastography (TEG)– guided transfusions with standard
care
Dosing of TXA Blood Studies,
A wide range of doses of TXA has been used in the trials and there is component n TEG Control OR (95% CI) P
no evidence to promote a higher over a lower dose. The dose
typically used for patients with congenital bleeding disorders of RBCs 4 100/214 128/212 0.58(0.38-0.86) .004
Plasma 4 43/250 69/250 0.54(0.35-0.84) .006
10 mg/kg bolus is suggested also for bleeding prophylaxis for
Platelets 5 36/267 65/264 0.48(0.30-0.75) .001
cardiac and major orthopedic surgery. For complex procedures
Hematology 2012 519

should be monitored with bedside activated clotting time. Overdose Table 5. Overview of the antidotes for different antithrombotic drugs
of protamine can cause hypotension, bronchoconstriction, platelet Antithrombotic agent Antidote Dose
aggregation, and consumption with bleeding.24 Low-molecular-
weight heparin is only partly reversed by protamine, but for Unfractionated heparin Protamine 1 mg/100 units
practical purposes, this method is still useful.24 Low-molecular-weight heparin Protamine 1 mg/100 units
Fondaparinux rFVIIa 90 ␮g/kg
Direct Xa inhibitors PCC? 25-50 IU/kg
The pentasaccharide fondaparinux and heparinoids (danaparoid) Direct thrombin inhibitors APCC? or hemodialysis 25-50 IU/kg
are not reversed by protamine. Fondaparinux has a half-life of Aspirin DDAVP (desmopressin) 0.3 ␮g/kg
18-20 hours, which becomes much longer in severe renal failure. Clopidogrel DDAVP (or platelets) 0.3 ␮g/kg
Studies in human volunteers and in single-patient cases have Prasugrel Platelets
demonstrated the beneficial reversing effect of rFVIIa at a dose of Ticagrelor ? ?
90 ␮g/kg.25,26
APCC indicates activated PCC; and DDAVP, D-arginine-deamino-vasopressin.
Vitamin K antagonists
Vitamin K antagonists should in this emergency setting be reversed given at a dose of 0.3 ␮g/kg slowly either IV or subcutaneously and
with prothrombin complex concentrate (PCC), because vitamin K is preferably combined with 10 mg/kg of TXA. Desmopressin
takes far too long (6-12 hours) to take effect and the volume of should be avoided in patients with active coronary artery disease
plasma needed (2.3 L) to reverse other than slight prolongations of and TXA is contraindicated in renal or ureteral procedures due to
the prothrombin time will cause volume overload.27 Sufficient the risk for clot formation and hydronephrosis. Desmopressin might
hemostasis is for most surgeries achieved with a prothrombin time also be effective for patients treated with the thienopyridines
corresponding to an international normalized ratio (INR) of 1.5. The clopidogrel35 or prasugrel, but platelet transfusion is another
dose of PCC is based on actual INR and body weight (Table 4). alternative. Although the inhibiting effect of aspirin and thienopyri-
When completely normal hemostatic function (INR ⱕ 1.2) is of dines on platelet function lasts for approximately 1 week, this
essence, as in neurosurgery, it is advisable to increase the dose of pertains only to platelets that were in circulation when the last dose
PCC by another 10 IU/kg. In a study of 160 patients treated emergently was given and to which the drug has bound irreversibly. Within a
with PCC, 44 had emergency surgery as indications; all of these few hours after that, there will only be a minimal concentration of
patients had good hemostatic effect without any thromboembolic the active metabolite and new platelets will not be inhibited.
complications in this subset.28 In a meta-analysis of 27 studies on PCC,
there was a weighted mean of thromboembolic complications of 1.4% The newest antiplatelet agent, ticagrelor, provides a dilemma. The
(95% CI, 0.8-2.1).29 The population was a mixture of patients having fact that it is “reversible” pertains only to its binding to the P2Y12
had emergency surgery and those with active bleeding. receptor on the platelets.36 Ticagrelor is active and remains so, with
a half-life of approximately 8 hours, and during this time also
New oral anticoagulants inhibits transfused platelets; there is so far no antidote. In case of
bleeding, the only option is supportive therapy, including blood
For the new oral anticoagulants, there are so far no approved or
transfusions, and to await the natural elimination of the drug, which
clinically validated antidotes. The oral thrombin inhibitor dabiga-
is not dialyzable.
tran can be eliminated by hemodialysis,30 which has been applied
successfully in at least one case with massive bleeding immediately
after open-heart surgery.31 Activated PCC, which in an animal Conclusion
model was effective at counteracting bleeding induced with another A large number of studies have demonstrated strategies to reduce
oral thrombin inhibitor, Melagatran, should be evaluated in clinical blood loss and the need for transfusions in major surgery. There is
situations.32 now a need for knowledge translation to clinical practice so that
these methods can be implemented widely. For the established
The oral anti-FXa inhibitor rivaroxaban has also shown good anticoagulants, there are effective and generally available antidotes
response in terms of laboratory parameters to PCC,33 but clinical (Table 5). The new antithrombotic agents will pose some difficulties
validation is required. Most of the drugs in this class have high in case of need for urgent surgery. However, the half-life of most of
plasma protein binding and are unlikely to be removed with dialysis. these new agents is relatively short and there is also active research
to identify suitable antidotes for them.
Reversal of platelet inhibitors
For the majority of surgical procedures except neurosurgery and Disclosures
prostate surgery, concomitant medication with aspirin is not a Conflict-of-interest disclosure: The author has received honoraria
significant problem. There can be a modest but manageable increase from Boehringer Ingelheim and Bayer Healthcare. Off-label drug
in blood loss. If aspirin was not interrupted and there is intra- or use: PCC and activated PCC for the reversal of new oral
postoperative bleeding that is difficult to manage, the vasopressin anticoagulants.
analog desmopressin can improve hemostasis.34 Desmopressin is
Correspondence
Table 4. Dose of PCC (IU/kg) necessary to reverse vitamin K Sam Schulman, Thrombosis Service, HHS-General Hospital, 237
antagonists Barton St E, Hamilton, ON, L8L 2X2, Canada; Phone: 905-527-
0271, ext 44479; Fax: 905-521-1551; e-mail: schulms@mcmaster.ca.
Actual INR Target INR ⬇ 1.5 Target INR ⱕ 1.2
Therapeutic (INR 2.0-3.0) 20 30
References
Supratherapeutic (INR 3.0-5.0) 30 40
1. Rogers MA, Blumberg N, Heal JM, Hicks GL Jr. Increased risk
Very excessive (INR ⬎ 5.0) 40 50
of infection and mortality in women after cardiac surgery
520 American Society of Hematology

related to allogeneic blood transfusion. J Womens Health structured use of routine laboratory tests or near-patient assess-
(Larchmt). 2007;16(10):1412-1420. ment with clinical judgement in the management of bleeding
2. Eikelboom JW, Mehta SR, Anand SS, Xie C, Fox KA, Yusuf S. after cardiac surgery. Br J Anaesth. 2004;92(2):178-186.
Adverse impact of bleeding on prognosis in patients with acute 21. Cui Y, Hei F, Long C, et al. Perioperative monitoring of
coronary syndromes. Circulation. 2006;114(8):774-782. thromboelastograph on blood protection and recovery for
3. D’Angelo A, Kluft C, Verheijen JH, Rijken DC, Mozzi E, severely cyanotic patients undergoing complex cardiac surgery.
Mannucci PM. Fibrinolytic shut-down after surgery: impairment Artif Organs. 2010;34(11):955-960.
of the balance between tissue-type plasminogen activator and its 22. Shore-Lesserson L, Manspeizer HE, DePerio M, Francis S,
specific inhibitor. Eur J Clin Invest. 1985;15(6):308-312. Vela-Cantos F, Ergin MA. Thromboelastography-guided trans-
4. Levi M, Levy JH, Andersen HF, Truloff D. Safety of recombi- fusion algorithm reduces transfusions in complex cardiac
nant activated factor VII in randomized clinical trials. N Engl surgery. Anesth Analg. 1999;88(2):312-319.
J Med. 2010;363(19):1791-1800. 23. Westbrook AJ, Olsen J, Bailey M, Bates J, Scully M, Salamon-
5. Carless PA, Henry DA, Moxey AJ, et al. Desmopressin for sen RF. Protocol based on thromboelastograph(TEG) out-
minimising perioperative allogeneic blood transfusion. Co- performs physician preference using laboratory coagulation
chrane Database Syst Rev. 2004;1:CD001884. tests to guide blood replacement during and after cardiac
6. Karlsson M, Ternstrom L, Hyllner M, et al. Prophylactic surgery: a pilot study. Heart Lung Circ. 2009;18(4):277-288.
fibrinogen infusion reduces bleeding after coronary artery 24. Schulman S, Bijsterveld NR. Anticoagulants and their reversal.
bypass surgery. A prospective randomised pilot study. Thromb Transfus Med Rev. 2007;21(1):37-48.
Haemost. 2009;102(1):137-144. 25. Bijsterveld NR, Moons AH, Boekholdt SM, et al. Ability of
7. Fergusson DA, Hebert PC, Mazer CD, et al. A comparison of recombinant factor VIIa to reverse the anticoagulant effect of
aprotinin and lysine analogues in high-risk cardiac surgery. the pentasaccharide fondaparinux in healthy volunteers. Circu-
N Engl J Med. 2008;358(22):2319-2331. lation. 2002;106(20):2550-2554.
8. Ong YL, Hull DR, Mayne EE. Menorrhagia in von Willebrand 26. Lisman T, Bijsterveld NR, Adelmeijer J, et al. Recombinant
disease successfully treated with single daily dose tranexamic factor VIIa reverses the in vitro and ex vivo anticoagulant and
acid. Haemophilia. 1998;4(1):63-65. profibrinolytic effects of fondaparinux. J Thromb Haemost.
9. Cap AP, Baer DG, Orman JA, Aden J, Ryan K, Blackbourne 2003;1(11):2368-2373.
LH. Tranexamic acid for trauma patients: a critical review of 27. Holland LL, Brooks JP. Toward rational fresh frozen plasma
the literature. J Trauma. 2011;71(1 Suppl):S9-S14. transfusion: The effect of plasma transfusion on coagulation
10. Henry DA, Carless PA, Moxey AJ, et al. Anti-fibrinolytic use test results. Am J Clin Pathol. 2006;126(1):133-139.
for minimising perioperative allogeneic blood transfusion. 28. Majeed A, Eelde A, Ågren A, Schulman S, Holmström M.
Cochrane Database Syst Rev. 2011;3:CD001886. Thromboembolic safety and efficacy of prothrombin complex
11. Kagoma YK, Crowther MA, Douketis J, Bhandari M, Eikelboom concentrates in the emergency reversal of warfarin coagulopa-
J, Lim W. Use of antifibrinolytic therapy to reduce transfusion in thy. Thromb Res. 2012;129(2):146-151.
patients undergoing orthopedic surgery: a systematic review of 29. Dentali F, Marchesi C, Pierfranceschi MG, et al. Safety of
randomized trials. Thromb Res. 2009;123(5):687-696. prothrombin complex concentrates for rapid anticoagulation
12. Gill JB, Rosenstein A. The use of antifibrinolytic agents in total reversal of vitamin K antagonists. A meta-analysis. Thromb
hip arthroplasty: a meta-analysis. J Arthroplasty. 2006;21(6): Haemost. 2011;106(3):429-438.
869-873. 30. Stangier J, Rathgen K, Stahle H, Mazur D. Influence of renal
13. Alshryda S, Sarda P, Sukeik M, Nargol A, Blenkinsopp J, impairment on the pharmacokinetics and pharmacodynamics of
Mason JM. Tranexamic acid in total knee replacement: a oral dabigatran etexilate: an open-label, parallel-group, single-
systematic review and meta-analysis. J Bone Joint Surg Br. centre study. Clin Pharmacokinet. 2010;49(4):259-268.
2011;93(12):1577-1585. 31. Warkentin TE, Margetts P, Connolly S, Lamy A, Ricci C.
14. Tanaka N, Sakahashi H, Sato E, Hirose K, Ishima T, Ishii S. Recombinant factor VIIa(rFVIIa) and hemodialysis to manage
Timing of the administration of tranexamic acid for maximum massive dabigatran-associated postcardiac surgery bleeding.
reduction in blood loss in arthroplasty of the knee. J Bone Joint Blood. 2012;119(9):2172-2174.
Surg Br. 2001;83(5):702-705. 32. Elg M, Carlsson S, Gustafsson D. Effects of activated prothrom-
15. Zufferey PJ, Miquet M, Quenet S, et al. Tranexamic acid in hip bin complex concentrate or recombinant factor VIIa on bleed-
fracture surgery: a randomized controlled trial. Br J Anaesth. ing time and thrombus formation during anticoagulation with a
2010;104(1):23-30. direct thrombin inhibitor. Thromb Res. 2001;101(3):145-157.
16. Gill JB, Chin Y, Levin A, Feng D. The use of antifibrinolytic 33. Eerenberg ES, Kamphuisen PW, Sijpkens MK, Meijers JC,
agents in spine surgery. A meta-analysis. J Bone Joint Surg Am. Buller HR, Levi M. Reversal of rivaroxaban and dabigatran by
2008;90(11):2399-2407. prothrombin complex concentrate: a randomized, placebo-
17. Henry D, Carless P, Fergusson D, Laupacis A. The safety of controlled, crossover study in healthy subjects. Circulation.
aprotinin and lysine-derived antifibrinolytic drugs in cardiac 2011;124(14):1573-1579.
surgery: a meta-analysis. CMAJ. 2009;180(2):183-193. 34. Levi M, Eerenberg E, Kamphuisen PW. Bleeding risk and
18. Ortel T. Perioperative management of patients on chronic reversal strategies for old and new anticoagulants and antiplate-
antithrombotic therapy. Hematology Am Soc Hematol Educ let agents. J Thromb Haemost. 2011;9(9):1705-1712.
Program. 2012;2012:529-535. 35. Leithäuser B, Zielske D, Seyfert UT, Jung F. Effects of
19. Ak K, Isbir CS, Tetik S, et al. Thromboelastography-based desmopressin on platelet membrane glycoproteins and platelet
transfusion algorithm reduces blood product use after elective aggregation in volunteers on clopidogrel. Clin Hemorheol
CABG: a prospective randomized study. J Card Surg. 2009; Microcirc. 2008;39(1-4):293-302.
24(4):404-410. 36. Nawarskas JJ, Clark SM. Ticagrelor: a novel reversible oral
20. Avidan MS, Alcock EL, Da Fonseca J, et al. Comparison of antiplatelet agent. Cardiol Rev. 2011;19(2):95-100.
Hematology 2012 521

Shahin et al. Critical Care 2011, 15:R162
http://ccforum.com/content/15/4/R162
RESEARCH Open Access
The relationship between inotrope exposure, six-

hour postoperative physiological variables, hospital
mortality and renal dysfunction in patients
undergoing cardiac surgery
Jason Shahin1*, Benoit deVarennes2, Chun Wing Tse3, Dan-Alexandru Amarica3 and Sandra Dial4,5
Abstract
Introduction: Acute haemodynamic complications are common after cardiac surgery and optimal perioperative
use of inotropic agents, typically guided by haemodynamic variables, remains controversial. The aim of this study
was to examine the relationship of inotrope use to hospital mortality and renal dysfunction.
Material and methods: A retrospective cohort study of 1,326 cardiac surgery patients was carried out at two
university-affiliated ICUs. Multivariable logistic regression analysis and propensity matching were performed to
evaluate whether inotrope exposure was independently associated with mortality and renal dysfunction.
Results: Patients exposed to inotropes had a higher mortality rate than those not exposed. After adjusting for
differences in Parsonnet score, left ventricular ejection fraction, perioperative intraaortic balloon pump use, bypass
time, reoperation and cardiac index, inotrope exposure appeared to be independently associated with increased
hospital mortality (adjusted odds ratio (OR) 2.3, 95% confidence interval (95% CI) 1.2 to 4.5) and renal dysfunction
(adjusted OR 2.7, 95% CI 1.5 to 4.6). A propensity score-matched analysis similarly demonstrated that death and
renal dysfunction were significantly more likely to occur in patients exposed to inotropes (P = 0.01).
Conclusions: Postoperative inotrope exposure was independently associated with worse outcomes in this cohort
study. Further research is needed to better elucidate the appropriate use of inotropes in cardiac surgery.
Introduction oxygen consumption and delivery, are often used to

Low cardiac output syndrome is a common complica- guide therapy. Inotropic agents are frequently used and
tion in patients undergoing cardiac surgery [1]. The titrated to achieve certain target levels of these physiolo-
aetiology is multifactorial and is believed to be related gical variables [3].
to a combination of myocardial ischaemia, reperfusion However, there is no consensus regarding low cardiac
injury, cardioplegia-induced myocardial dysfunction and output syndrome in terms of both the physiological para-
preexisting cardiac disease [2]. The administration of meters that define it and the interventions used to treat it
fluids and inotropes and the use of ventricular assist [2]. The use of inotropes has been described to be both
devices are common therapies in the management of centre- and physician-dependent and as being adminis-
low cardiac output syndrome. Physiological variables tered to as few as 5% or to as many as 100% of patients
such as cardiac index, mixed venous oxygen saturation, undergoing elective coronary bypass surgery [4-6].
oxygen delivery and lactate levels, obtained either from Although inotropes have been demonstrated to improve
a cardiac output monitoring device or by measuring haemodynamics and measured physiological variables
[2], they may be a source of increased mortality and mor-
bidity as they can increase cardiac arrhythmias and
* Correspondence: jason.shahin@gmail.com ischaemia [7]. Very few randomised, controlled trials
1
Division of Critical Care, McGill University Health Centre, 687 Pine Avenue
West, Montreal, QC, H3A 1A1, Canada comparing the different agents used and their effects on
Full list of author information is available at the end of the article
© 2011 Shahin et al.; licensee BioMed Central Ltd. This is an open access article distributed under the terms of the Creative Commons
Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in
any medium, provided the original work is properly cited.
Shahin et al. Critical Care 2011, 15:R162 Page 2 of 10
clinical outcomes have been performed in patients who Dobutamine exposure was defined as delivery of any
undergo cardiac surgery [2]. In a recent observational dose as long as it was administered for at least three
study, the receipt of dobutamine perioperatively was hours in the ICU. Epinephrine exposure was defined as a
associated with increased mortality [8]. In addition, minimum duration of three hours in the ICU if the dose
investigators in randomised, controlled trials of inotropes was < 5 μg/minute or any duration if doses ≥ 5 μg/min-
in patients with heart failure and left ventricular dysfunc- ute were used. Consistent with the definitions used by
tion have reported increased side effects and increased other authors [2], norepinephrine and vasopressin were
mortality [9]. not considered inotropes.
While the use of inotropes as part of a protocol to tar- The primary study outcomes were hospital mortality
get physiological parameters within the first six hours and the occurrence of postoperative renal dysfunction.
after cardiac surgery has been shown to improve out- Renal dysfunction was defined by an increase in creati-
comes in patients with sepsis [10], it is unclear whether nine ≥ 200% from baseline in the first five postoperative
their use in the postoperative cardiac care setting is days or new renal replacement therapy at any point dur-
favourable with respect to morbidity and mortality. This ing hospitalisation. These definitions are consistent with
study was undertaken to evaluate the relationship of stages 2 and 3 of the Acute Kidney Injury Network clas-
inotrope use to morbidity and mortality in a cohort of sification system [11]. Secondary outcomes were ICU
consecutive patients undergoing cardiac surgery. length of stay and hospital length of stay.
Data were collected regarding patient age, sex, Parsonnet
Materials and methods score, medical history, procedure-related variables and
Setting and study population six-hour postoperative physiological variables. Medical his-
This study was conducted at two adult tertiary care uni- tories were abstracted from patient records. Conditions
versity-affiliated hospitals. Data were collected retrospec- considered were any prior cardiac surgery, hypertension,
tively between 1 January 2005 and 31 December 2005 by diabetes, atrial fibrillation, preoperative hospitalisation for
trained reviewers using standardised data collection sheets. heart failure, preoperative renal dysfunction, preoperative
Consecutive patients who had undergone coronary artery dialysis, preoperative left ventricular ejection fraction and
bypass graft (CABG) surgery, valve replacement or repair left ventricular dysfunction. Preoperative renal dysfunction
or combined CABG and valvular or aortic procedures was defined by a preoperative creatinine level ≥ 150 μmol/
were included in the study. Patients who had undergone a L. Left ventricular dysfunction was defined as left ventricu-
heart transplant, pulmonary thromboendarterectomy or lar ejection fraction < 30%. Procedure-related variables
placement of a ventricular assistance device were consisted of cardiac procedures, perioperative intraaortic
excluded. balloon pump use, reoperation, emergency operation,
All patients had been admitted postoperatively to the mean bypass time and severe postoperative bleeding. Car-
ICU. A Swan-Ganz catheter had been used perioperatively diac procedures were separated into a CABG-only group
at both hospitals to guide patient resuscitation. Serum lac- and a group who had undergone other procedures, defined
tate and mixed venous oxygen saturation levels had been as (1) valve repair or replacement or (2) combined CABG
measured in all patients at one site, and selected patients and valve repair or replacement. An emergency operation
at the other site. To avoid bias, only the data from the was defined as a cardiac procedure occurring within 24
patients treated at the hospital with routinely measured hours of acute coronary syndrome or immediately after a
serum lactate and mixed venous oxygen saturation were percutaneous intervention. Reoperation was defined as the
used for the analysis of those variables. need for reoperation 48 hours after the initial cardiac pro-
All research was conducted in keeping with the princi- cedure. Severe postoperative bleeding was defined as
ples outlined in the Declaration of Helsinki. The research blood loss > 2.5 L within 24 hours after the initial cardiac
ethics committee of the McGill University Health Centre procedure. The six-hour postoperative physiological vari-
Research Institute approved the study. The hospital’s ables consisted of mean arterial pressure, cardiac index,
ethics committee waived the need for informed consent as mixed venous oxygen saturation and serum lactate. These
the data were collected retrospectively. tests were performed in the ICU six hours after the
operation.
Exposure, covariates and end points
The primary drug exposure studied was postoperative Data analysis
inotrope use. Epinephrine, milrinone and dobutamine Preoperative variables, intraoperative variables and out-
were the main inotropes used in the study centres. The comes in patients exposed or unexposed to inotropes
definition of ‘postoperative inotrope use’ varied for each were compared using Student’s t-test and the Wilcoxon
medication. Milrinone exposure was defined as adminis- rank-sum test for continuous variables and a c2 test for
tration of any dose for any length of time in the ICU. categorical variables. Multivariable logistic regression
was performed to test the association between inotrope after stratifying by the six-hour physiological variables. All
exposure and outcomes after adjusting for possible con- data processing and analyses were performed using SAS
founding variables. A forward stepwise procedure was version 9.2 software (SAS Institute, Cary, NC, USA).
first used to assess inotrope exposure in the model.
Variables were kept in the model if they were believed Results
to be clinically important or if they altered the associa- In total, 1,326 patients were initially included in the
tion between inotrope exposure and outcome by ≥ 10%. study. Their mean age was 66 years, with 10% of the
Covariates included in the model were Parsonnet score, cohort being older than 80 years of age and more than
left ventricular dysfunction, perioperative intraaortic bal- two-thirds being male. The majority of the procedures
loon pump use, bypass time, reoperation, postoperative were CABG operations, and the mean Parsonnet score
bleeding, preoperative renal dysfunction, treatment hos- (± SD) was 13.4 ± 10.6 (Table 1). Over 97% of the
pital, aprotinin use and cardiac index < 2.2 [3]. Variables patients had Swan-Ganz catheters inserted. Fifty percent
included in the Parsonnet score were not included in of patients were exposed to inotropes intraoperatively,
the model, other than a low left ventricular ejection and forty-two percent were exposed postoperatively. The
fraction, as we thought this might strongly influence hospital mortality rate was 7.8%, and renal dysfunction
decision making with respect to inotrope use, and its occurred in 8.3% of the patients (Table 2). As shown in
weighting in the Parsonnet score might not reflect this. Figure 1, hospital mortality increased progressively with
Colinearity between left ventricular ejection fraction and Parsonnet scores > 20.
Parsonnet score was excluded as the variance inflation Twelve patients who died within six hours of admis-
factor was < 2.5. sion to the ICU were excluded from further analysis,
A second analytic method, propensity score matching, leaving a total of 1,314 patients finally included in the
was performed to evaluate the association between ino- study.
trope exposure and outcomes. A one-to-many greedy Patients exposed to inotropes differed from those
five-to-one digit technique was performed to match one unexposed in that they were older; had higher Parsonnet
control in the no agent group (control group) by one scores, more comorbid illnesses and lower preoperative
case (inotrope-exposed). A ‘greedy five-to-one digit left ventricular ejection fractions; were more likely to
match’ means that the cases were first matched to con- have undergone valvular or combined procedures; and
trols on five digits of the propensity score. For those that had longer bypass times. However, at six hours after
did not match, cases were then matched to controls on ICU admission, aside from serum lactate, which was
four digits of the propensity score. This continued down higher in the group of patients exposed to inotropes,
to a one-digit match on propensity scores for those that postoperative physiological parameters were very similar.
remained unmatched. If a one-digit match was not possi- The hospital mortality rate was seven times higher in
ble, the case remained unmatched and was not included the group of patients exposed to inotropes compared to
in the matched case control analysis. Matching variables those not exposed. Similarly, the rate of severe renal
included Parsonnet score, low left ventricular ejection failure was significantly higher in the patients exposed
fraction, age, sex, bypass time, procedure type, CABG to inotropes (Table 2).
only or other procedures, and perioperative intraaortic We also examined the association of hospital mortality
balloon pump use. In this matched sample, baseline char- and renal failure with the six-hour postadmission phy-
acteristics and outcomes were compared between ino- siological variables. As expected, the mortality was lower
trope exposed and unexposed groups by performing in patients with higher measured cardiac indices, mixed
paired t-tests for continuous variables and McNemar’s venous saturation, normal lactate values and higher cal-
test for categorical values. culated oxygen delivery at six hours. However, the mor-
Because of previous literature correlating patient out- tality was significantly lower in patients unexposed to
comes with physiological goals [10] and recommendations inotropes even in the presence of six-hour measured
that certain levels of physiologic variables be targeted [3], physiological variables lower than recommended thresh-
we examined the relationship between measured six-hour old values. The odds of dying in the hospital were four
physiologic variables and outcomes. Specifically, oxygen to eight times higher in the inotrope exposed group
delivery, cardiac index, serum lactate and mixed venous than in the unexposed group for similar levels of mea-
oxygen saturation were examined. The six-hour post- sured physiologic variables (Figure 2). A similar relation-
operative physiological variables were categorised with ship was found for renal dysfunction (Figure 3).
cutoffs based on recommendations in the literature [3,10], As a low preoperative left ventricular ejection fraction
with the exception of oxygen delivery, which was based on may have increased the likelihood that a patient would be
the median value derived from the data. We examined the treated with inotropes, the effect of inotropes on mortal-
relationship between inotrope exposure and outcomes ity in the 232 patients with preoperative left ventricular
Table 1 Baseline characteristics for total cohort and patient subgroupsa

Characteristics Total cohort (N = 1,326) Inotrope-unexposed (n = 783) Inotrope-exposed (n = 531)
Demographics
Mean age (± SD) 66.1 (11.0) 64.6 (10.7) 68.3 (11.0)
Age ≥ 80 years, n (%) 123 (9.3) 51 (6.5) 72 (13.6)
Females, n (%) 404 (30.5) 188 (24.0) 207 (39.0)
Mean Parsonnet score (± SD) 13.4 (10.6) 10.5 (8.3) 17.2 (11.7)
Medical history
Prior cardiac surgery, n (%) 78 (5.9) 26 (3.3) 48 (9.0)
Hypertension, n (%) 904 (68.2) 534 (68.2) 362 (68.2)
Diabetes, n (%) 417 (31.5) 220 (28.1) 193 (36.4)
Atrial fibrillation, n (%) 135 (10.2) 53 (6.8) 80 (15.1)
Preoperative CHF, n (%) 179 (13.5) 79 (10.1) 99 (18.6)
Preoperative renal dysfunction, n (%) 94 (7.1) 31 (4.0) 61 (11.5)
Preoperative dialysis, n (%) 21 (1.6) 9 (0.7) 12 (2.3)
Mean LVEF (± SD) 49.7 (11.8) 53.3 (11.8) 44.4 (15.8)
LVEF < 30%, n (%) 241 (18.2) 74 (9.5) 162 (30.5)
Procedure-related variables
CABG only, n (%) 912 (68.8) 601 (76.8) 303 (57.1)
Other procedure, n (%) 414 (31.2) 182 (23.2) 228 (42.9)
Perioperative IABP, n (%) 132 (10.0) 30 (3.8) 96 (18.1)
Received inotropes intraoperative, n (%) 666 (50.7) 219 (28.0) 447 (84.2)
Emergency operation, n (%) 132 (10.0) 58 (7.4) 74 (13.9)
Reoperation, n (%) 78 (5.9) 20 (2.6) 35 (6.6)
Mean bypass time, minutes (± SD) 101.2 (45.7) 86.8 (32.8) 120 (52)
Severe postoperative bleeding, n (%) 50 (3.8) 15 (1.1) 35 (6.6)
a
CHF, congestive heart failure; LVEF, left ventricular ejection fraction; CABG, coronary artery bypass graft; IABP, intraaortic balloon pump; SD, standard deviation.
Table 2 Postoperative variables and outcomesa

Variables and outcomes Total cohort Inotrope-unexposed (n = 783) Inotrope-exposed
(N = 1,314) (n = 531)
Postoperative inotrope and vasopressor use, n (%)
Norepinephrine 674 (51.3) 279 (35.6) 395 (74.4)
Vasopressin 73 (5.6) 5 (0.6) 68 (12.8)
Epinephrine 308 (23.29) 0 308 (58.0)
Dobutamine 75 (5.7) 0 75 (14.1)
Milrinone 332 (25.3) 0 332 (62.5)
Six-hour postoperative physiological variables
Mean arterial pressure, mmHg (± SD) 75.9 (10.0) 76.7 (10.0) 74.5 (9.7)
Mean oxygen delivery, mL/minute/m2 353 (101) 363 (100) 341 (98)
(± SD)
Mean cardiac index, L/min (± SD) 2.9 (0.7) 2.9 (0.7) 2.8 (0.7)
Mixed venous oxygen saturation, n (%) 70.7 (9.0) 71.3 (8.2) 70.1 (9.3)
Mean serum lactate, μmol/L (± SD) 3.2 (2.7) 2.1 (1.6) 4.1 (3.0)
Outcomes
Died, n (%) 103 (7.8) 15 (1.9) 76 (14.3)
Renal dysfunction, n (%) 105 (8.3) 25(3.2) 87 (16.8)
Median ICU length of stay, days (IQR) 1.1(1.8) 1.0 (0.9,1.6) 2.1 (1.0,4.7)
Median hospital length of stay, days (IQR) 8 (6) 7.0 (5.0,9.0) 10.0 (6.0,18.0)
a
IQR, interquartile range; SD, standard deviation.

their six-hour physiological variables were lower than

the targets recommended in the literature [3,10]. The

relationship between inotrope exposure and poor out-

comes remained significant after adjusting for differ-
ences in Parsonnet score, left ventricular ejection
fraction, perioperative intraaortic balloon pump use,

bypass time, reoperation and cardiac index. In addition,

an analysis using propensity score matching produced

similar results.

The demonstration of poorer outcomes in patients
exposed to inotropes is consistent with the results demon-

strated in several previous studies. An observational study

of patients who received dobutamine after cardiac surgery
demonstrated increased cardiac morbidity after the data

were adjusted for confounders [8]. Milrinone has also

been described as being associated with an increased risk
Figure 1 Hospital mortality by Parsonnet score in a cohort of
1,326 cardiac surgery patients. Error bars indicate 95% confidence
of postoperative atrial fibrillation in a cardiac surgery
intervals (95% CIs). population [12]. Further evidence that inotropes may be
harmful can be found in the heart failure literature. Ran-
domised, controlled trials of patients with decompensated
fractions < 30% was examined. In this subgroup, patients heart failure treated with phosphodiesterase inhibitors ver-
who received inotropes had significantly higher mortality sus placebo revealed that those in the treatment group
than patients who did not, with an odds ratio OR of 14.7 experienced more episodes of hypotension and cardiac
(95% confidence interval (95% CI) 2.0 to 11.1). arrhythmia and had higher mortality rates [13-15].
After adjusting for differences in Parsonnet score, left Furthermore, despite promising initial results, levosimen-
ventricular ejection fraction, perioperative intraaortic dan, a new class of inotrope, was not shown to be superior
balloon pump use, bypass time, reoperation and cardiac to dobutamine in a randomised, controlled trial of patients
index, exposure to inotropes was associated with with acute heart failure and left ventricular ejection frac-
increased hospital mortality (adjusted OR 2.3, 95% CI tions < 30% [16]. Researchers who conducted a systematic
1.2 to 4.5; P = 0.01) (Table 3). Receipt of inotropes was review of controlled trials of b-adrenergic agents com-
also significantly associated with increased odds of renal pared to either placebo or an active agent in patients with
dysfunction (adjusted OR 2.7, 95% CI 1.6 to 4.7; P < heart failure concluded that there is very little evidence
0.001) (Table 4). There were no differences in these that treatment improves symptoms or patient outcomes
associations when treatment hospital, preoperative renal and may in fact be harmful [9]. Inotrope use was also asso-
dysfunction or aprotinin use was included in the regres- ciated with increased mortality in a recent observational
sion analysis. As these variables did not confound the study of heart failure with an effect size similar in magni-
association between the exposures and outcomes, they tude to that found in our study [17].
were not included in the final regression model. Since the 1970s, a number of randomised trials have
Using greedy one-to-five matching, 123 inotrope- been undertaken in medical and surgical patients to inves-
exposed patients were matched to 123 unexposed patients tigate whether targeting specific resuscitation goals, such
using one-digit matching only. Only preoperative left ven- as cardiac output and oxygen delivery, would improve
tricular ejection fraction was statistically significantly dif- patient outcomes [18]. Achieving the prespecified goals
ferent between the two groups, although equal numbers of often involved the use of inotropic medications to increase
patients had left ventricular ejection fractions < 30%. Hos- cardiac output and oxygen delivery. Such therapy, referred
pital mortality, renal dysfunction, ICU and hospital length to as ‘goal-directed therapy’, has been associated with
of stay were significantly worse in the patients exposed to improved outcomes, primarily in patients with sepsis and
inotropes (Table 5). in certain high-risk surgical patient populations [10,19].
However, these trials differed with regard to patient mix,
Discussion physiologic targets, therapies used and management of
In this observational study, postoperative inotrope expo- control arms [20].
sure was associated with increased hospital mortality Four controlled studies have been published in which a
and renal dysfunction in cardiac surgery patients. Hospi- goal-directed therapy protocol was used in the cardiac
tal mortality and renal dysfunction were consistently surgery setting [21-23]. The trials differed with regard to
lower in patients unexposed to inotropes, even when targeted goals, therapeutic protocols and use of inotropic
Figure 2 Hospital mortality stratified by (A) oxygen delivery, (B) cardiac index, (C) serum lactate and (D) mixed venous oxygen
saturation. Two groups of patients are represented (inotrope-exposed and inotrope-unexposed). The adjusted odds ratios for the association
between inotrope exposure and mortality were 8.5 (95% CI 4.8 to 15.0) after adjusting for oxygen delivery and 7.7 (95% CI 4.4 to 13.7) after
adjusting for cardiac index. The adjusted odds ratios for the association between inotropes exposure and mortality were 5.7 (95% CI 2.4 to 13.5)
after adjusting for serum lactate and 5.3 (95% CI 2.4 to 11.4) after adjusting for mixed venous oxygen saturation. Error bars indicate 95% CI. Data
from only one hospital were used for analyses of serum lactate and mixed venous oxygen saturation, resulting in wider 95% CIs.
medications. Two of the trials, which employed mainly increased myocardial oxygen consumption and cardiac
fluid infusions, demonstrated improved physiological arrhythmia. Both of these side effects may lead to poor
goals with minimal catecholamine use. The largest trial, cardiac performance [7]. Alternatively, low cardiac out-
which targeted mixed venous oxygen saturation, demon- put may be due to mechanical obstruction, as in cardiac
strated shortened hospital stay and less morbidity, but tamponade, which may require surgical intervention.
these outcomes were associated with increased catechola- The use of inotropes in these situations may transiently
mine use. The last trial, which enrolled 30 patients in improve the haemodynamic state but ultimately lead to
total, demonstrated no significant difference in outcomes further harm, as appropriate diagnosis and treatment
but did require a more intense inotrope regimen to attain may be either delayed or missed altogether. Further-
the specified goals. Despite demonstrating improved clin- more, catecholamine use has been associated with
ical outcomes, all four studies were underpowered to reduced metabolic efficiency by promoting fatty acid
detect any difference in mortality. Furthermore, two of oxidation over that of glucose. This may be a further
the four protocols required greater catecholamine doses impediment to optimal cardiac performance. Catechola-
to achieve their goals. mine use has also been associated with bacterial growth,
Several mechanisms may explain the increased mortal- increased bacterial virulence, biofilm formation, insulin
ity observed in patients exposed to inotropes. The two resistance and hyperglycaemia, all of which may contri-
most common side effects of inotropic medications are bute to poor outcomes [25].
Figure 3 Renal dysfunction stratified by (A) oxygen delivery, (B) cardiac index, (C) serum lactate and (D) mixed venous oxygen
saturation. Two groups of patients are represented (inotrope-exposed and inotrope-unexposed). The adjusted odds ratios for the association
between inotrope exposure and renal dysfunction were 6.0 (95% CI 3.8 to 9.5) after adjusting for oxygen delivery and 5.9 (95% CI 3.7 to 9.6) after
adjusting for cardiac index. The adjusted odds ratios for the association between inotrope exposure and mortality were 5.6 (95% CI 2.8 to 11.2)
after adjusting for serum lactate and 5.3 (95% CI 2.5 to 10.9) after adjusting for mixed venous oxygen saturation. Error bars indicate 95% CI. Data
from only one hospital were used for analyses of serum lactate and mixed venous oxygen saturation, resulting in wider 95% CIs.
Our study has several strengths. First, our results are less Conclusions
likely to be biased by a single centre’s practice pattern, as The results of our study demonstrate that inotrope
the patients were recruited from two centres. Second, as exposure was associated with increased hospital mortal-
we routinely collected postoperative physiological data, we ity and renal dysfunction in cardiac surgery patients.
were able to adjust for important haemodynamic variables. The observational nature of the data and the potential
Finally, because extensive preoperative, intraoperative and for confounding precludes any final conclusions about a
postoperative data were collected, we were able to control causal relationship. Nevertheless, the significant practice
for many potential confounding factors. variations reported in the literature, and the consistency
The study’s main limitation is that it is an observa- of our results with those reported in the cardiac surgery
tional study, and thus the associations could be due to and heart failure literature, demonstrate the need for
residual confounding. Specifically, we may not have fully future research [4-6,26]. As inotropes may be associated
adjusted the data for confounding by indication and with increased morbidity and mortality, adequately pow-
confounding due to severity of illness. To minimise con- ered, randomised, controlled trials are needed to clarify
founding, we performed multiple different analyses, the risks and benefits of inotrope use in cardiac surgery
including a propensity-matched analysis. patients.
Table 3 Multivariable analysis of association between inotrope exposure and hospital mortalitya
Variable Crude odds ratio Adjusted odds ratio (95% CI) P value
Inotrope status
Inotrope unexposed (ref) 1.0 1.0
Inotrope exposed 9.1 2.3 (1.2 to 4.5) 0.01
Parsonnet score
0 to 9 (ref) 1.0 1.0
10 to 14 2.4 1.8 (0.7 to 4.6) 0.24
15 to 19 3.1 1.8 (0.7 to 4.9) 0.23
20 to 29 8.3 4.6 (2.0 to 10.9) < 0.001
≥ 30 24.6 11.2 (4.6 to 27.5) < 0.001
Ejection fraction
LVEF ≥ 30% 1.0
LVEF < 30% 2.2 1.5 (0.9 to 2.7) 0.15
Intraaortic balloon pump use
No perioperative IABP use 1.0
Perioperative IABP use 8.2 3.3 (1.9 to 5.8) < 0.001
Bypass timeb 1.6 1.3 (1.1 to 1.4) < 0.001
Reoperative status
No reoperation 1.0
Reoperation 6.1 4.7 (2.3 to 9.3) < 0.001
Cardiac index at six hours, L/min
Cardiac index ≥ 2.2 (ref) 1.0 1.0
Cardiac index < 2.2 2.8 1.7 (0.93 to 1) 0.09
a
LVEF, left ventricular ejection fraction; IABP, intraaortic balloon pump; ref, reference value; CI, confidence interval; bmodelled linearly as an increase in bypass
time of 30 minutes.
Table 4 Multivariable analysis of association between inotrope exposure and renal dysfunctiona
Variable Crude odds ratio Adjusted odds ratio (95% CI) P value
Inotrope status
Inotrope unexposed (ref) 1.0 1.0
Inotrope exposed 7.5 2.7 (1.5 to 4.6) < 0.001
Parsonnet score
0 to 9 (ref) 1.0 1.0
10 to 14 1.5 1.2 (0.7 to 2.4) 0.59
15 to 19 2.1 1.4 (0.7 to 2.9) 0.35
20 to 29 3.6 2.3 (1.2 to 4.3) 0.01
≥ 30 6.5 2.8 (1.3 to 6.1) 0.007
Ejection fraction
LVEF ≥ 30% 1.0
LVEF < 30% 1.6 1.0 (0.6 to 1.7) 0.97
Intraaortic balloon pump use
No perioperative IABP use 1.0
Perioperative IABP use 5.4 2.7 (1.6 to 4.7) < 0.001
Bypass timeb 1.5 1.2 (1.1 to 1.4) < 0.001
Reoperative status
No reoperation 1.0
Reoperation 4.5 2.3 (1.2 to 4.5) 0.02
Renal dysfunction
Normal preoperative renal function 1.0
Preoperative renal dysfunction 3.7 1.7 (0.8 to 3.6) 0.14
Cardiac index at six hours, L/min
Cardiac index ≥ 2.2 (ref) 1.0 1.0
Cardiac index < 2.2 1.7 1.0 (0.6 to 1.9) 0.88
a b
LVEF, left ventricular ejection fraction; IABP, intraaortic balloon pump; ref, reference value; CI, confidence interval; modelled linearly as an increase in bypass
time of 30 minutes.
Table 5 Baseline characteristics and outcomes for propensity-matched groupsa

Characteristics No inotropes (n = 123) Inotropes (n = 123) P value
Demographics
Mean age (± SD) 67.2 (11.5) 67.3 (10.9) 0.97
Age ≥ 80 years, n (%) 18 (14.6) 14 (11.4) 0.45
Females, n (%) 49 (39.8) 43 (5.0) 0.48
Medical history, n (%)
Prior cardiac surgery 5 (4.1) 6 (4.9) 1.0
Hypertension 84 (68.3) 78 (63.4) 0.53
Diabetes 40 (32.5) 43 (35.0) 0.78
Atrial fibrillation 17 (13.8) 15 (12.2) 0.84
Preoperative hospitalisation for CHF 20 (16.3) 11 (8.9) 0.12
Preoperative renal dysfunction 9 (7.3) 10 (8.1) 1.0
Preoperative dialysis 1 (0) 0 (0.8) 0.32
LVEF < 30% 25 (20.3) 25 (20.3) 1.0
Mean LVEF (± SD) 50.6 (13.9) 45.8 (15.2) 0.02
Procedure-related variables
Mean Parsonnet score (± SD) 14.1 (8.7) 14.4 (9.2) 0.78
CABG only, n (%) 82 (66.7) 85 (69.1) 0.80
Other procedure, n (%) 41 (33.3) 38 (30.9) 0.80
Perioperative IABP, n (%) 3 (2.4) 4 (3.2) 1.0
Emergency operation, n (%) 14 (11.4) 17 (13.8) 0.70
Reoperation, n (%) 4 (3.3) 9 (7.9) 0.23
Mean bypass time, minutes (± SD) 99.3 (28.7) 98.6 (32.9) 0.83
Severe postoperative bleeding, n(%) 2 (1.6) 6 (4.9) 0.29
Outcomes
Died, n (%) 1 (0.8) 10 (8.1) 0.01
Renal dysfunction, n (%) 2 (1.6) 12 (9.8) 0.01
Median ICU length of stay (IQR) 1.0 (0.9,1.8) 1.8 (0.9,3.2) < 0.0001
Median hospital length of stay (IQR) 8.0 (6.0,11.0) 9.0 (6.0,18.0) 0.03
a
CHF, congestive heart failure; LVEF, left ventricular ejection fraction; CABG, coronary artery bypass graft; IABP, intraaortic balloon pump; SD, standard deviation;
IQR, interquartile range.
Key messages Author details

1
Division of Critical Care, McGill University Health Centre, 687 Pine Avenue
• Postoperative inotrope exposure was independently West, Montreal, QC, H3A 1A1, Canada. 2Division of Cardiac Surgery, McGill
associated with hospital mortality and renal University Health Centre, 687 Pine Avenue West, Montreal, QC, H3A 1A1,
dysfunction. Canada. 3McGill University Faculty of Medicine, 845 Sherbrooke Street West,
Montreal, QC, H3A 2T5, Canada. 4Respiratory Epidemiology and Clinical
• The increased hospital mortality and renal dys- Research Unit, Montreal Chest Institute, McGill University Health Centre, 3650
function in patients exposed to inotropes are St-Urbain, Montreal, QC, H2X 2P4, Canada. 5Department of Critical Care,
observed even when recommended six-hour physio- SMBD-Jewish General Hospital, McGill University, 845 Sherbrooke Street
West, Montreal, QC, H3A 2T5, Canada.
logical variables are achieved.
• Patients unexposed to inotropes with six-hour phy- Authors’ contributions
siological variables below recommended targets had JS and SD contributed to the study design and analysis as well as the
drafting of the manuscript. JS, BD, CWT and AA contributed to the
lower mortality than patients exposed to inotropes acquisition of data. All of the authors approved the final manuscript.
who achieved these targets.
Competing interests
The authors declare that they have no competing interests.
Abbreviations Received: 20 January 2011 Revised: 27 March 2011

CABG: coronary artery bypass and graft; CI: confidence interval; SD: standard Accepted: 7 July 2011 Published: 7 July 2011
deviation.
References
Acknowledgements 1. Ahmed I, House CM, Nelson WB: Predictors of inotrope use in patients
We thank Dr Richard Menzies for his help in the analysis of the data and the undergoing concomitant coronary artery bypass graft (CABG) and aortic
drafting of the manuscript. We also thank Barbara Verhoef for all her work in valve replacement (AVR) surgeries at separation from cardiopulmonary
data acquisition. bypass (CPB). J Cardiothorac Surg 2009, 4:24.
2. Gillies M, Bellomo R, Doolan L, Buxton B: Bench-to-bedside review: 21. Kapoor PM, Kakani M, Chowdhury U, Choudhury M, Kiran U: Early goal-
Inotropic drug therapy after adult cardiac surgery: a systematic literature directed therapy in moderate to high-risk cardiac surgery patients. Ann
review. Crit Care 2005, 9:266-279. Card Anaesth 2008, 11:27-34.
3. St André AC, DelRossi A: Hemodynamic management of patients in the 22. McKendry M, McGloin H, Saberi D, Caudwell L, Brady AR, Singer M:
first 24 hours after cardiac surgery. Crit Care Med 2005, 33:2082-2093. Randomised controlled trial assessing the impact of a nurse delivered,
4. Kastrup M, Markewitz A, Spies C, Carl M, Erb J, Grosse J, Schirmer U: Current flow monitored protocol for optimisation of circulatory status after
practice of hemodynamic monitoring and vasopressor and inotropic cardiac surgery. BMJ 2004, 329:258.
therapy in post-operative cardiac surgery patients in Germany: results 23. Pölönen P, Ruokonen E, Hippeläinen M, Pöyhönen M, Takala J: A
from a postal survey. Acta Anaesthesiol Scand 2007, 51:347-358. prospective, randomized study of goal-oriented hemodynamic therapy
5. Butterworth JF: Dobutamine: Too dangerous for routine administration? in cardiac surgical patients. Anesth Analg 2000, 90:1052-1059.
Anesthesiology 2008, 108:973-974. 24. Goepfert MS, Reuter DA, Akyol D, Lamm P, Kilger E, Goetz AE: Goal-
6. Hernandez AF, Li S, Dokholyan RS, O’Brien SM, Ferguson TB, Peterson ED: directed fluid management reduces vasopressor and catecholamine use
Variation in perioperative vasoactive therapy in cardiovascular surgical in cardiac surgery patients. Intensive Care Med 2007, 33:96-103.
care: data from the Society of Thoracic Surgeons. Am Heart J 2009, 25. Singer M: Catecholamine treatment for shock-equally good or bad?
158:47-52. Lancet 2007, 370:636-637.
7. Parissis JT, Rafouli-Stergiou P, Stasinos V, Psarogiannakopoulos P, 26. Bastien O, Vallet B, French Study Group AGIR: French multicentre survey
Mebazaa A: Inotropes in cardiac patients: update 2011. Curr Opin Crit Care on the use of inotropes after cardiac surgery. Crit Care 2005, 9:241-242.
2010, 16:432-441.
8. Fellahi JL, Parienti JJ, Hanouz JL, Plaud B, Riou B, Ouattara A: Perioperative doi:10.1186/cc10302
use of dobutamine in cardiac surgery and adverse cardiac outcome: Cite this article as: Shahin et al.: The relationship between inotrope
propensity-adjusted analyses. Anesthesiology 2008, 108:979-987. exposure, six-hour postoperative physiological variables, hospital mortality
9. Thackray S, Easthaugh J, Freemantle N, Cleland JG: The effectiveness and and renal dysfunction in patients undergoing cardiac surgery. Critical Care
relative effectiveness of intravenous inotropic drugs acting through the 2011 15:R162.
adrenergic pathway in patients with heart failure: a meta-regression
analysis. Eur J Heart Fail 2002, 4:515-529.
10. Rivers E, Nguyen B, Havstad S, Ressler J, Muzzin A, Knoblich B, Peterson E,
Tomlanovich M, Early Goal-Directed Therapy Collaborative Group: Early
goal-directed therapy in the treatment of severe sepsis and septic
shock. N Engl J Med 2001, 345:1368-1377.
11. Mehta RL, Kellum JA, Shah SV, Molitoris BA, Ronco C, Warnock DG, Levin A,
Acute Kidney Injury Network: Acute Kidney Injury Network: report of an
initiative to improve outcomes in acute kidney injury. Crit Care 2007, 11:
R31.
12. Fleming GA, Murray KT, Yu C, Byrne JG, Greelish JP, Petracek MR, Hoff SJ,
Ball SK, Brown NJ, Pretorius M: Milrinone use is associated with
postoperative atrial fibrillation after cardiac surgery. Circulation 2008,
118:1619-1625.
13. Cohn JN, Goldstein SO, Greenberg BH, Lorell BH, Bourge RC, Jaski BE,
Gottlieb SO, McGrew F, DeMets DL, White BG: A dose-dependent increase
in mortality with vesnarinone among patients with severe heart failure.
Vesnarinone Trial Investigators. N Engl J Med 1998, 339:1810-1816.
14. Cuffe MS, Califf RM, Adams KF Jr, Benza R, Bourge R, Colucci WS,
Massie BM, O’Connor CM, Pina I, Quigg R, Silver MA, Gheorghiade M,
Outcomes of a Prospective Trial of Intravenous Milrinone for Exacerbations
of Chronic Heart Failure (OPTIME-CHF) Investigators: Short-term
intravenous milrinone for acute exacerbation of chronic heart failure: a
randomized controlled trial. JAMA 2002, 287:1541-1547.
15. Packer M, Carver JR, Rodeheffer RJ, Ivanhoe RJ, DiBianco R, Zeldis SM,
Hendrix GH, Bommer WJ, Elkayam U, Kukin ML, Mallis GI, Sollano JA,
Shannon J, Tandon PK, DeMets DL, the PROMISE Study Research Group:
Effect of oral milrinone on mortality in severe chronic heart failure. The
PROMISE Study Research Group. N Engl J Med 1991, 325:1468-1475.
16. Mebazaa A, Nieminen MS, Packer M, Cohen-Solal A, Kleber FX, Pocock SJ,
Thakkar R, Padley RJ, Põder P, Kivikko M, SURVIVE Investigators:
Levosimendan vs dobutamine for patients with acute decompensated
heart failure: the SURVIVE Randomized Trial. JAMA 2007, 297:1883-1891.
17. Mebazaa A, Parissis J, Porcher R, Gayat E, Nikolaou M, Boas FV, Delgado JF,
Follath F: Short-term survival by treatment among patients hospitalized
with acute heart failure: the global ALARM-HF registry using propensity Submit your next manuscript to BioMed Central
scoring methods. Intensive Care Med 2011, 37:290-301. and take full advantage of:
18. Giglio MT, Marucci M, Testini M, Brienza N: Goal-directed haemodynamic
therapy and gastrointestinal complications in major surgery: a meta-
• Convenient online submission
analysis of randomized controlled trials. Br J Anaesth 2009, 103:637-646.
19. Boyd O, Grounds RM, Bennett ED: A randomized clinical trial of the effect • Thorough peer review
of deliberate perioperative increase of oxygen delivery on mortality in • No space constraints or color figure charges
high-risk surgical patients. JAMA 1993, 270:2699-2707.
20. Grocott MPW, Hamilton MA, Bennett ED, Harrison D, Rowan K: • Immediate publication on acceptance
Perioperative increase in global blood flow to explicit defined goals and • Inclusion in PubMed, CAS, Scopus and Google Scholar
outcomes following surgery. Cochrane Database Syst Rev 2006, 2:
• Research which is freely available for redistribution
CD004082.
Submit your manuscript at

www.biomedcentral.com/submit
doi:10.1093/icvts/ivs184 Advance Access publication 24 June 2012
Prognostic value of preoperative quality of life on mortality after

isolated elective myocardial revascularization
Rutger ter Horsta, Athanasios L.P. Markoub and Luc Noyeza,*
a
Department of Cardio-Thoracic Surgery—677, Heart Centre, Radboud University Nijmegen Medical Center, Nijmegen, Netherlands
b
Isala Klinieken, De Weezenlanden, Cardio Thoracic Surgery, Zwolle, Netherlands
* Corresponding author. Department of Cardio-Thoracic Surgery—677, Heart Center, Radboud University Nijmegen Medical Center, PO Box 9101,
6500 HB Nijmegen, Netherlands. Tel: +31-24-3613711; fax: +31-24-3540129; e-mail: l.noyez@ctc.umcn.nl (L. Noyez).
Received 24 January 2012; received in revised form 22 March 2012; accepted 1 April 2012
Abstract
OBJECTIVES: This study evaluates whether a low preoperative quality of life (QoL), measured with the EuroQoL instruments EQ-5D and
EQ-visual analogue scale (VAS) can be used as a predictor of mortality after elective isolated myocardial revascularization.
METHODS: A total of 2501 patients, with a mean age of 65.3 ± 9.4 (range 18–93) years and a mean additive EuroSCORE of 2.7 ± 2.1 (0–12),
undergoing an elective isolated coronary artery bypass graft between January 2002 and June 2011 completed preoperative EQ-5D and EQ-VAS.
RESULTS: Hospital mortality [1.0% (25/2501 patients)] and 30-day mortality [1.2% (29/25 patients)] were the studied outcomes. The
ORIGINAL ARTICLE
EQ-5D was 0.69 ± 0.26 (−0.30 to 1.0) with a median of 0.77 and the EQ-VAS was 59.7 ± 22.4 (0–100) with a median of 60. Regression analysis
showed a significant correlation between hospital mortality and EQ-5D (P = 0.016) and EQ-VAS (P = 0.033). There is a significant correlation
between 30-day mortality and EQ-5D (P = 0.048), but not for EQ-VAS (P = 0.06). The c-statistics (95% confidence interval) for EQ-5D and EQ-
VAS for predicting hospital mortality are 0.36 (0.24–0.46) and 0.33 (0.23–0.42), respectively. The c-statistics for predicting 30-day mortality are
0.39 (0.30–0.49) for EQ-5D and 0.35 (0.26–0.44) for EQ-VAS.
CONCLUSIONS: Based on these results, we conclude that, in isolation, poor low preoperative EQ-5D and EQ-VAS scores do not contribute to
deciding which patients should undergo cardiac surgery.
Keywords: Quality of life • Coronary artery bypass graft • Mortality
INTRODUCTION data plus follow-up data from all adult patients undergoing
cardiac surgery at the Radboud University Nijmegen Medical
A patient’s mortality risk from a procedure is an important param- Centre (UMCN)—we identified 4107 patients who underwent
eter in making a clinical decision on whether to go through with isolated CABG between January 2002 and June 2011. Of the 4107,
the procedure or not. It is known that there is a correlation 3330 surgeries were elective and 2501 of these patients completed
between preoperative quality of life (QoL) and post-surgical recov- our preoperative EuroQoL-questionnaire on the day before
ery in patients undergoing cardiac surgery [1–3]. Some research surgery. The initial EuroSCORE [9] was used for risk stratification.
has shown that preoperative QoL, or an aspect of the QoL, can be
used as an independent risk factor to predict post-surgical mortal-
ity and morbidity, certainly because the correlation between Follow-up
preoperative anginal class and preoperative QoL is low [4–7].
These results raise the question whether a cardiac surgical proced- Patients who were discharged alive from the UMCN were seen
ure can be denied for patients with a low preoperative QoL. at the postoperative outpatient clinic 6 weeks after discharge.
The aim of the present study is to evaluate whether pre- Patients who did not show up were contacted by one of the
operative QoL, using the EuroQoL [8], can be used to predict the authors (L.N.) to complete their 30-day survival data.
risk of mortality for patients undergoing isolated elective coron-
ary artery bypass graft (CABG) surgery.
Quality of life
PATIENTS AND METHODS To assess the QoL, the EQ-5D and EQ-VAS, both components of
the EuroQoL instrument, were used [8]. The EQ-5D consists of
Patients five domains of health (mobility, self-care, usual activities, pain/
discomfort and anxiety/depression), and each domain is divided
From our Coronary Surgery Database Radboud Hospital into three levels: (i) no problems, (ii) some or moderate pro-
(CORRAD)—a database that stores pre-, peri- and postoperative blems and (iii) extreme problems. Based on the response to this
652 R. ter Horst et al. / Interactive CardioVascular and Thoracic Surgery
classification, a single index value is estimated using a general completed with a median and the 25th and 75th percentiles
population-based algorithm [10]. Secondly, patients estimated interquartile range (IQR range). Differences in percentages were
their own health on a visual analogue scale (EQ-VAS) ranging tested using the χ 2 test and numerical variables were tested
from 0 to 100, with 0 being the worst possible health state and using the t-test or Mann–Whitney test when appropriate. The as-
100 being the best possible health state. The EQ-5D index can sociation between the EQ-5D, EQ-VAS and hospital or 30-day
be regarded as a societal-based composite global QoL measure, mortality is evaluated using logistic regression analysis. The
whereas the EQ-VAS is a direct global QoL assessment from the Nagelkerke R 2 (0–100%) was used to evaluate the association
patient’s perspective. between the EQ-5D and EQ-VAS. Receiver operator characteristic
All patients participated on a voluntary basis in this QoL re- curves (ROC curves) were used to assess the value of EQ-5D and
search. Registration of data in the CORRAD database and the use EQ-VAS as a predictor of hospitalisation or 30-day mortality.
of this information for research have been approved by the local Only patients with complete data from all pre- and post-
ethical and research council of the Radboud University, operative questionnaires were included in the analyses. A
Nijmegen [11]. P-value of ≤0.05 was considered significant.
Outcome variables RESULTS

Study endpoints are hospital mortality, defined as death occur-
Patients
ring at any time during hospital admission after CABG surgery,
and 30-day mortality, defined as all hospital mortalities and all
Between January 2002 and June 2011, 4107 isolated myocardial
deaths within 30 days postoperation in the group of patients
revascularizations were performed at the UMCN, with a hospital
who were discharged from the UMCN before the 30th post-
mortality of 70 patients (1.7%). After excluding urgent and emer-
surgical day.
gency surgery, plus patients with a recent myocardial infarction
as defined by the EuroSCORE [9], we identified 3330 patients
who underwent isolated elective CABG (total population). Of
Statistical analyses these patients, 2501 (75%) had complete QoL data, our study
population, 829 patients without or with incomplete QoL data
Statistical analyses were performed using 16.0 SPSS Inc., Chicago, were excluded, the exclusion group.
IL, USA. Baseline characteristics are presented as a percentage Table 1 presents the baseline characteristics of the total group,
for dichotomous variables, as mean ± SD, and as a range for nu- the study population and the exclusion group. Diabetes is
merical variables. Results from the EQ-5D and EQ-VAS are defined as diet-controlled, oral therapy or insulin-dependent
Table 1: Baseline characteristics of patients
Variable Total group [n = 3330 (%)] Study population [n = 2501 (%)] Excluded group [n = 829 (%)] P-value*
Age (years) 65.5 ± 9.7 (18–93) 65.3 ± 9.4 (18–93) 66.4 ± 10.2 (35–91) 0.011
Female 753 (22.6) 525 (21.0) 228 (27.8) 0.001
Diabetes 735 (22.1) 542(21.7) 193 (23.1) 0.333
Vascular disease 525 (15.8) 391 (15.6 134 (16.3) 0.717
Neurological disease 298 (8.9) 226 (9.0) 72 (8.7) 0.759
Renal disease 92 (2.8) 68 (2.7) 24 (2.9) 0.789
Pulmonary disease 380 (11.4) 281 (11.2) 99 (11.9) 0.579
Preoperative myocardial infarction 1101 (33.1) 836 (33.4) 265 (32.0) 0.439
Previous cardiac surgery 116 (3.5) 86 (3.4) 30 (3.6) 0.823
Body mass index (kg/m2) 27.4 ± 4.0 (17.5–48.2) 27.5 ± 4.1 (17.5–48.2) 27.1 ± 4.1 (17.5–47.6) 0.555
Additive EuroSCORE 2.8 ± 2.2 (0–12) 2.7 ± 2.1 (0–12) 3.1 ± 2.3 (0–12) 0.001
Groups 0.002
Low risk 1545 (46.4) 1197 (47.9) 348 (42.0)
Medium risk 1353 (40.6) 1003 (40.1) 350 (42.2)
High risk 432 (13.0) 301 (9.0) 131 (15.8)
Three-vessel disease 2465 (74) 1859 (74.3) 606 (73.1) 0.747
Extracorporeal circulation (min) 100 ± 31 (11–359) 100 ± 31.8 (11–310) 99 ± 33.2 (27–359) 0.965
Aortic cross-clamp time (min) 59 ± 20 (6–187) 60 ± 21.8 (7–187) 59.6 ± 22.8 (6–166) 0.886
Off pump 235 (7.1) 181 (7.2) 54 (6.4) 0.481
Grafts (number) 1.9 ± 0.4 (1–4) 1.9 ± 0.44 (1–4) 1.9 ± 0.46 (1–4) 0.348
Distal anastomoses (number) 3.7 ± 1.2 (1–9) 3.6 ± 1.2 (1–9) 3.7 ± 1.2 (1–8) 0.186
EQ-5D 0.69 ± 0.26 (−0.30 to 1.0); 0.77 IQR: 0.65–0.84
EQ-VAS 59.7 ± 22.4 (−0 to 100); 60 IQR:50–75
Hospital mortality 36 (1.1) 25 (1.0) 11 (1.3) 0.430
30-day mortality 41 (1.2) 29 (1.2) 12 (1.4) 0.515
*P-value study population versus exclusion group.

R. ter Horst et al. / Interactive CardioVascular and Thoracic Surgery 653
Table 2: Age, additive EuroSCORE, EQ-5D and EQ-VAS of hospital deaths versus hospital survivors
Variable Survivors (n = 2476) Hospital deaths (n = 25) P-value
Age 65.3 ± 9.4 72.3 ± 7.9 0.001

EuroSCORE 2.7 ± 2.1 5.2 ± 2.7 0.001
EQ-5D 0.70 ± 0.26 (−0.30 to 1); 0.77 IQR 0.65–0.84 0.57 ± 0.31 (−27 to 0.89); 0.68 IQR 0.25–0.80 0.048
EQ-VAS 59.8 ± 22.5 (0–100); 60 IQR 50–75 50.1 ± 17.4 (20–95); 50 (IQR 39–60) 0.011
Survivors (n = 2472) 30-day deaths (n = 29)
Age 65.3 ± 9.4 72.3 ± 7.9 0.001
EuroSCORE 2.7 ± 2.1 5.2 ± 2.7 0.001
EQ-5D 0.70 ± 0.26 (−30 to 10); 0.77 IQR 0.65–0.84 0.60 ± 0.31 (−27 to 0.89); 0.69 IQR 40–82 0.090
EQ-VAS 59.8 ± 22.5 (0–100); 60 (50–75) 51.1 ± 17.4 (20–95); 50 (40–61) 0.021
diabetes; vascular disease as peripheral, abdominal vascular

pathology or operation; and neurological disease as cerebrovas-
cular accidents and/or transient ischaemic attack. Patients with a
preoperative creatinine ≥150 µmol/l, preoperative dialysis or a
renal transplant were registered under renal disease, and for pul-
monary diseases the same definition has been used as applied
by the EuroSCORE [9].
ORIGINAL ARTICLE
The study population has a mean age of 65.3 ± 9.4 (range
18–93) years. The EQ-5D is 0.69 ± 0.26 (−0.30 to 1.0), median of
0.77 (IQR 0.65–0.84) and the EQ-VAS is 59.7 ± 22.4 (0–100),
median of 60 (IQR 50–75). The mean additive EuroSCORE is
2.7 ± 2.1 (0–12). Hospital mortality was 1.0% (25/2501 patients)
and 30-day mortality was 1.2% (29/25). Statistical analysis shows
that the exclusion group is significantly older (P = 0.0110),
consists of more women, 27.8 versus 21% (P = 0.001), has a
significantly higher percentage of high-risk patients, 15.8 versus
9% (P = 0.002), and has a significantly higher additive EuroSCORE
risk score, 3.1 ± 2.3 (0–12) versus 2.7 ± 2.1 (0–12) (P = 0.001). Other
studied variables show no statistically significant difference. Figure 1: ROC curve for EQ-5D (full line), EQ-VAS (dotted line) and additive
EuroSCORE (dashed line) for hospital mortality.
QoL and mortality DISCUSSION
Table 2 presents the values of EQ-5D, EQ-VAS, the additive Several research groups have focused on various QoL assess-
EuroSCORE and the age of hospital survivors versus hospital ments to test the predictive value for short- and long-term mor-
deaths for both 30-day survivors and 30-day deaths. Patients tality of patients undergoing cardiac surgery [1–7]. In the present
who died either during hospital admission or within 30-day study, we took a closer look at the predictive value of EQ-5D
post-surgery were significantly older than survivors (P = 0.001). In and EQ-VAS, both components of the EuroQoL facility [8]. In our
addition, the additive EuroSCORE is significantly higher for study population, only isolated elective CABG patients were
patients who died (P = 0.001), whereas the value of the EQ-5D is included. Patients with urgent and emergency surgeries were
lower in the group of hospital deaths (P = 0.048). The EQ-VAS is excluded from our analyses as the medical indication for cardiac
significantly lower for both hospital and 30-day deaths (P = 0.011 surgery in these cases is strictly regulated. Patients with a recent
and P = 0.021, respectively). myocardial infarction, as defined by the EuroSCORE [9], were
The association between the EQ-5D, EQ-VAS and hospital excluded. For some of these patients, the medical indication was
mortality is significant but very low; EQ-5D (P = 0.016; R 2 = decisive to perform a CABG and it was impossible to distinguish
0.020), EQ-VAS (P = 0.033; R 2 = 0.016). The association between these specific patients in our database. From the 3330 patients
EQ-5D and EQ-VAS and 30-day mortality is significant for EQ-5D (total population) who could be included in our study, 2501
(P = 0.048; R 2 = 0.012), but not for EQ-VAS (P = 0.06; R 2 = 0.011). (75%) had complete QoL data. The remaining 829 patients (25%)
Both associations have a low correlation. were excluded as their QoL data were incomplete. The issue of
The c-index (95% confidence interval) for EQ-5D, EQ-VAS and incomplete QoL data is a known problem in this kind of research
the additive EuroSCORE for predicting hospital mortality are 0.36 [11]. It is important to notice that the exclusion group is signifi-
(0.24–0.46), 0.33 (0.23–0.42) and 0.80 (0.71–0.87), respectively cantly older, has a higher percentage of women and has a
(Fig. 1). The c-index for predicting 30-day mortality is 0.39 (0.30– higher risk score preoperatively. Table 1 shows that besides age
0.49) for EQ-5D and 0.35 (0.26–0.44) for EQ-VAS. and gender, other risk variables show no significant difference. It
654 R. ter Horst et al. / Interactive CardioVascular and Thoracic Surgery
is important to notice that the operative data and the registered patients with no or incomplete QoL data. It is interesting to see
hospital and 30-day mortality show no differences between the that the exclusion group is older and at a higher risk for mortal-
study population and the exclusion group. ity, and this must be taken into account before generalizing the
Table 2 shows that hospital and 30-day mortality appear, as results.
expected, in patients who are significantly older and at a higher
risk than survivors. The EQ-5D is close to a statistically significant
level (P = 0.048). The EQ-5D is lower for hospital deaths but
shows no significant difference between 30-day deaths and CONCLUSION
survivors. The EQ-VAS is significantly lower for both hospital and
30-day deaths. Logistic regression analysis shows a significant In conclusion, our results support the hypothesis that in isolation,
correlation between hospital mortality and EQ-5D and EQ-VAS, poor preoperative QoL, measured using EQ-5D and EQ-VAS,
as well as between 30-day mortality and EQ-5D, albeit with a cannot contribute in deciding which patients should undergo
poor association. (Nagelkerke R 2 < 0.021 for all correlations). isolated elective CABG surgery.
The c-index for the predictive value of EQ-5D and EQ-VAS for
hospital mortality are 0.36 and 0.33, respectively. The c-index for
the predictive value of EQ-5D and EQ-VAS for 30-day mortality ACKNOWLEDGEMENTS
are 0.39 and 0.35, respectively. These results show that both
EQ-5D and EQ-VAS have no prognostic value for hospital or Elise Noyez is thanked for her correction of the English text.
30-day mortality. The c-index of the additive EuroSCORE for
hospital and 30-day mortality (0.80 for both mortality measure- Conflict of interest: none declared.
ments), however, confirms that the additive EuroSCORE is a
good prognostic tool to predict the mortality rate of patients
undergoing isolated elective CABG surgery.
Our results—no prognostic value of QoL, evaluated by the REFERENCES
EuroQoL facility, for hospital and 30-day mortality—seem to be
in contrast with previous published researches [4–6]. Rumsfeld [1] Duits AA, Boeke S, Taams MA, Passchier J, Erdman RA. Prediction of
quality of life after coronary bypass graft surgery: a review and evalu-
et al. [4] proved that the preoperative physical component
ation of multiple recent studies. Psychosom Med 1997;59:257–68.
summary (PCS) score is an important independent risk factor for [2] Koch CG, Li L, Lauer M, Sabik J, Starr NJ, Blackstone EH. Effect of func-
6-month mortality following CABG surgery [4]. However, emer- tional health-related quality of life on long-term survival after cardiac
gency and acute operations are included in his study population, surgery. Circulation 2007;115:692–9.
and the baseline characteristics of the study population differed [3] Mallik S, Krumholz HM, Lin ZQ, Kasi SV, Mattera JA, Roumains SA et al.
Patients with depressive symptoms have lower health status benefits
greatly from the exclusion group. About 37% of the patients after coronary bypass surgery. Circulation 2005;111:271–7.
(1476/3956) were excluded in the study, which is much higher [4] Rumsfeld JS, MaWhinney S, McCarthy M, Shroyer AL, VillaNueva CB,
than in our research. Another point is that we present a series of O’Brien M et al. Health-related quality of life as a predictor of mortality
2501 patients operated over a period of 9 and a half years, with following coronary artery bypass graft surgery. J Am Med Assoc 1999;
281:1298–303.
a hospital mortality of 25 patients (1%) and a 30-day mortality of
[5] Sündermann S, Dademasch A, Rastan A, Praetorius J, Rodruez H,
29 patients (1.2%). This is in contrast with Rumsfeld et al. [4] who Walther T et al. One-year follow-up of patients undergoing elective
present a series of 2480 patients with a 180-day mortality of 117 cardiac surgery assessed with the comprehensive assessment of frailty
patients (4.7%). So, it can be suggested that our number of test and its simplified form. Interact CardioVasc Thorac Surg 2011;13:
deaths (endpoints) is too small. However, if elective isolated 119–23.
[6] Székely A, Nussmeier NA, Miao Y, Huang K, Levin J, Feierfeil H et al. A
CABG can be performed with a hospital, 30-day mortality <1.5%, multinational study of the influence of health-related quality of life on
the influence of QoL on this mortality is of course much less. in-hospital outcome after coronary artery bypass graft surgery. Am Heart
Sündermann et al. [5] use the comprehensive assessment of J 2011;161:1179–85.
frailty (CAF) to analyse the ability to predict one-year mortality. [7] Rumsfeld JS, Magid DJ, O’Brien M, McCarthy M, MaWhinney S, Shroyer
ALW et al. Changes in health-related quality of life following coronary
For their analyses they used a highly specific group of patients,
bypass graft surgery. Ann Thorac Surg 2001;72:2026–32.
only those with an age of ≥74 years. This is an important differ- [8] Brooks R. EuroQol: the current state of play. Health Policy 1996;37:53–7.
ence in our study, mainly because some studies confirm the in- [9] Nashef SAM, Roques F, Michel P, Gauducheau E, Lemeshow S, Salamon
fluence of age on the QoL [12, 13]. QoL is measured using the R, the EuroSCORE study group. European system for cardiac preoperative
CAF test, which is a new facility specifically for older patients [5]. risk evaluation (EuroSCORE). Eur J Cardiothorac Surg 1999;16:9–13.
[10] Dolan P. Modeling valuations for EuroQol health states. Med Care 1997;
Their results show a promising ability to predict one-year mor- 35:1095–108.
tality, even much better than the EuroSCORE. Székely et al. used [11] Wouters CW, Noyez L. Is no news good news? Organized follow-up, an
both PCS and MCS of the SF-12 to predict a prolonged length of absolute necessity for the evaluation of myocardial revascularization. Eur
in-hospital stay and mortality [6]. The MCS is identified as an J Cardiothorac Surg 2004;26:667–70.
[12] Markou ALP, Windt van der A, Swieten van HA, Noyez L. Changes in
independent predictor of a prolonged length of hospital stay
quality of life, physical activity, and symptomatic status one year after
and mortality. However, the data set used for their research is myocardial revascularization for stable angina. Eur J Cardiothorac Surg
almost 10 years old and was gathered together from a total of 2008;34:1009–15.
72 hospitals in 17 different countries. In addition, they give no [13] Järvinen O, Saarinen T, Julkunen J, Huhtala H, Tarkka MR. Changes in
additional information about the exclusion criteria used. The health-related quality of life and functional capacity following coronary
artery bypass graft surgery. Eur J Cardiothorac Surg 2003;24:750–6.
problem of missing QoL data and information about the [14] L Noyez L, Jager de MJ, Markou APL. Quality of life post cardiac surgery.
excluded patients is mentioned earlier in the discussion and is Under-researched research. Interact CardioVasc Thorac Surg 2011;13:
not a new phenomenon [14]. In our research, we excluded 511–5.
TRANSFUSION PRACTICE
The effect of blood transfusion on pulmonary permeability in

cardiac surgery patients: a prospective multicenter cohort study _3231 82..90
Alexander P.J. Vlaar,* Alexander D. Cornet,* Jorrit J. Hofstra, Leendert Porcelijn, Albertus Beishuizen,
Willem Kulik, Margreeth B. Vroom, Marcus J. Schultz, A.B. Johan Groeneveld, and
Nicole P. Juffermans
[Correction added after online publication 11-Jul-11: The spelling of Alexander D. Cornet has been modified.]
T
here is a clear association between blood trans-
BACKGROUND: There is an association between fusion and pulmonary complications in cardiac
blood transfusion and pulmonary complications in surgery.1,2 The exact mechanism of transfusion-
cardiac surgery. Mediators of increased pulmonary vas- related morbidity and mortality is not fully
cular leakage after transfusion are unknown. We understood. Suggested mechanisms include increased
hypothesized that factors may include antibodies or bio- risk for infections,3 hydrostatic pulmonary edema caused
active lipids, which have been implicated in transfusion- by cardiac overload, or enhanced permeability of the pul-
related acute lung injury. monary microvasculature caused by transfusion-related
STUDY DESIGN AND METHODS: We performed a acute lung injury (TRALI).4-6
prospective cohort study in two university hospital inten- Of note, cardiac surgery was found to be a risk factor
sive care units in the Netherlands. Pulmonary vascular for the development of TRALI.7,8 Therefore, mechanisms
permeability was measured in cardiac surgery patients in onset of TRALI may play an important role in explaining
after receiving no, restrictive (one or two transfusions),
or multiple (five or more transfusions) transfusions (n =
20 per group). The pulmonary leak index (PLI), using
67
Ga-labeled transferrin, was determined within 3 hours ABBREVIATIONS: ALI = acute lung injury; CABG = coronary
postoperatively. Blood products were screened for bioac- artery bypass grafting; ICU(s) = intensive care unit(s);
tive lipid accumulation and the presence of antibodies. lysoPC(s) = lysophosphatidylcholine(s); MS/MS = tandem mass
RESULTS: The PLI was elevated in all groups after spectrometry; PAF = platelet-activating factor; PC(s) =
cardiac surgery. Transfused patients had a higher PLI phosphatidylcholine(s); PEEP = positive end-expiratory
compared to nontransfused patients (33 ¥ 10-3 ⫾ pressure; PLI = pulmonary leak index.
20 ¥ 10-3 vs. 23 ¥ 10-3 ⫾ 11 ¥ 10-3/min, p < 0.01). The
amount of red blood cell (RBC) products, but not of From the Department of Intensive Care Medicine, the Depart-
fresh-frozen plasma or platelets, was associated with an ment of Internal Medicine, the Laboratory of Experimental
increase in PLI (b, 1.6 [0.2-3.0]). Concerning causative Intensive Care and Anesthesiology (L.E.I.C.A.), and the Labora-
factors in the blood product, neither the level of bioactive tory of Genetic Metabolic Diseases, Academic Medical Center;
lipids nor the presence of antibodies was associated with Sanquin Diagnostic Services, Sanquin Region North West; and
an increase in PLI. Patient factors such as surgery risk the Department of Intensive Care Medicine and the Institute for
and time on cardiopulmonary bypass did not influence Cardiovascular Research Vrije Universiteit (ICAR-VU), VU
the risk of pulmonary leakage after blood transfusion. Medical Center Amsterdam, Amsterdam, the Netherlands.
CONCLUSIONS: Transfusion in cardiothoracic surgery Address reprint requests to: A.P.J. Vlaar, Laboratory of
patients is associated with an increase in pulmonary Experimental Intensive Care and Anesthesiology (L.E.I.C.A.),
capillary permeability, an effect that was dose dependent Academic Medical Center, Room M0-228, Meibergdreef 9,
for RBC products. The level of bioactive lipids or the Amsterdam, 1105 AZ, the Netherlands; e-mail: a.p.vlaar@
presence of HLA or HNA antibodies in the transfused amc.uva.nl.
products were not associated with increased pulmonary *These authors contributed equally.
capillary permeability. Received for publication February 1, 2011; revision
received May 2, 2011, and accepted May 2, 2011.
doi: 10.1111/j.1537-2995.2011.03231.x
TRANSFUSION 2012;52:82-90.
82 TRANSFUSION Volume 52, January 2012

TRANSFUSION-RELATED PULMONARY PERMEABILITY
the frequently observed pulmonary dysfunction after car- group); or a minimum of 2 units of red blood cells (RBCs),
diopulmonary bypass.9 The finding of an increased TRALI 2 units of fresh-frozen plasma (FFP), and 1 unit of platelets
incidence after cardiac surgery may be related to the “two- (PLTs) pooled from five donors (n = 20, massive transfu-
hit” pathogenesis of TRALI. The “first hit” is an inflamma- sion group). Transfusions administered in the operation
tory condition of the patient that primes the lung room or within the first 3 hours postoperatively were
neutrophils. In cardiac surgery, use of cardiopulmonary included. All RBCs are leukoreduced (buffy coat removed
bypass as well as deflation of the lung during surgery may and the RBC suspension is filtered to remove the WBC
contribute to priming. The “second hit” is caused by count to less than 1 ¥ 106) and stored for up to 35 days. The
factors in the blood product, resulting in neutrophil acti- hospital blood bank policy of one center holds that
vation and increased vascular permeability, with subse- cardiac surgery patients are provided with fresh RBCs only
quent pulmonary edema. Both white blood cell (WBC) (stored <14 days) during surgery, but not while on the ICU.
antibodies in the blood product that react with pulmonary RBCs were transfused to maintain the hemoglobin (Hb)
neutrophils of the recipient,10-13 as well as bioactive lipids concentration higher than 5.0 mmol/L (8.7 g/dL), FFP
(lysophosphatidylcholines [lysoPCs]) that accumulate and PLTs were transfused in the case of (suspected) bleed-
during blood storage, have been implicated as a second ing. The volume of plasma transfused was calculated as
hit.7,14-19 In line with the latter mechanism, storage time of the sum of 325 mL per unit of FFP and 250 mL per platelet
blood has been associated with pulmonary complications (PLT) concentrate.
in cardiac surgery,2,20,21 although not all studies have con-
firmed this association.22-24 Cardiothoracic surgery and anesthesia procedures
Differentiating between hydrostatic and permeability
Patients were anesthetized according to local institutional
pulmonary edema is limited by the subjective interpreta-
protocol, with lorazepam, etomidate, sufentanil, and
tion of clinical findings.25,26 The pulmonary leak index
rocuronium for induction of anesthesia and facilitation of
(PLI) has been used to differentiate between hydrostatic
intubation. During the surgical procedure, sufentanil was
and permeability edema.27 Also, an elevated PLI is an early
used as analgesic and sevoflurane plus propofol were used
marker of acute lung injury (ALI) in at-risk patients, yet
to maintain anesthesia. Muscle relaxants were not given
before acute respiratory distress syndrome.28
during the surgical procedure. Steroids were given at the
We hypothesized that hypoxia occurring after cardiac
discretion of the cardioanesthesiologist. In all patients,
surgery may be due to mild forms of TRALI not meeting
cardiopulmonary bypass was performed under mild to
the TRALI criteria. To determine if there is a correlation
moderate hypothermia (28-34°C), using a membrane oxy-
between transfusion and pulmonary leakage, we mea-
genator and a nonpulsatile blood flow. During the proce-
sured the PLI in a cohort of cardiac surgery patients after
dure, lungs were deflated. After the procedure, all patients
receiving restrictive or multiple blood transfusions, as well
were transferred to the ICU with mechanical ventilation.
as in nontransfused controls. Known causative factors for
the onset of TRALI were determined, including bioactive
lipids and HLA or HNA antibodies in the transfused blood PLI
products. The PLI was measured within 3 hours postoperatively, as
previously described.29 Transferrin was labeled in vivo,
MATERIALS AND METHODS after intravenous (IV) injection of 67Ga-citrate, 4.5 MBq
(physical half-life 78 hr; Mallinckrodt Diagnostica, Petten,
The study was performed in the mixed medical-surgical
the Netherlands). Patients were in the supine position and
intensive care units (ICUs) of two university hospitals in
two scintillation detection probes (Eurorad C.T.T., Stras-
the Netherlands. Both ICUs are a “closed format” depart-
burg, France) were positioned over the right and left lung
ment in which patients are under the direct care of the
apices. Starting at the time of the IV injection of 67Ga,
ICU team. The study was approved of by the ethical com-
radioactivity was detected for 30 minutes. The 67Ga counts
mittee of both hospitals. Before valvular and/or coronary
are corrected for background radioactivity, physical half-
artery surgery, patients of 18 years or older were asked
life, spillover of 67Ga, obtained by in vitro measurement of
informed consent for participation in the study. Exclusion 67
Ga, and expressed as cpm per lung field. At 0, 5, 8, 12, 16,
criteria were off-pump surgery, emergency surgery, and
20, 25, and 30 minutes after 67Ga injection, blood samples
the use of immunosuppressive drugs.
(2-mL aliquots) were taken. Each blood sample was
weighed and radioactivity was determined with a single-
Design well well-counter (LKB Wallac 1480 WIZARD, Perkin
In a prospective cohort study, cardiac surgery patients Elmer, Life Science, Zaventem, Belgium), taking back-
were consecutively included for analysis after they had ground, spillover of 67Ga, and decay into account. Results
received no transfusion (n = 20, no transfusion group); are expressed as cpm/g. For each blood sample, a time-
one or two transfusions (n = 20, restrictive transfusion matched cpm over each lung was taken. The radioactivity
Volume 52, January 2012 TRANSFUSION 83

VLAAR ET AL.
ratio was calculated as (67Galung)/(67Gablood) and plotted the supernatant was stored at -80°C until further analy-
against time. The PLI was calculated from the slope of sis. WBC-reactive antibodies were examined in the
increase of the radioactivity ratio divided by the intercept, plasma samples of PLT and FFP products. Samples
to correct for physical factors in radioactivity detection. were screened for HLA antibodies using a standard
The PLI represents the transport rate of 67Ga-transferrin complement-dependent cytotoxicity assay with an HLA-
from the intravascular to the extravascular space of the typed donor panel (to detect complement-fixing anti-
lungs and is therefore a measure of pulmonary vascular bodies to HLA Class I and II)35 and a Luminex screening
permeability. The values for both lung fields are averaged. assay for HLA Class I and II (Tepnel Lifecode Luminex
The upper limit normal for the PLI is 14.1 ¥ 10-3/min, and Screen Deluxe, Stamford, CT). HLA Class I and II anti-
the measurement error is approximately 10%.30 bodies were identified using a Luminex single antigen
bead technology (Tepnel Lifecode Luminex SA).
Patient data collection Lymphocyte-reactive antibodies were examined by the
lymphocyte immunofluorescence test.36 WBC agglutinat-
Potential risk factors for an increased PLI were scored,
ing antibodies (HNA-3a) were examined by the WBC
including alcohol abuse, smoking, myocardial infarction,
agglutination technique.37 Granulocyte-reactive anti-
hypertension, diabetes, vascular diseases, hematologic
bodies (HNA-1a, -1b, -1c, -2a) were examined by the
malignancy, solid malignancy, cerebrovascular accident,
granulocyte immunofluorescence test.38 HNA-1a, -1b,
and autoimmune disease. The known risk factors for ALI
-2a, and -3a were typed in the granulocyte immunofluo-
such as sepsis, pneumonia, and trauma were not taken
rescence test.
into account as the included patients were elective
Lipid extraction of supernatant from stored RBC and
surgery patients and for this reason presence of such a risk
PLT supernatant was performed using the method of
factor was a reason to cancel surgery in these patients or
Bligh and Dyer. In short, 3 mL of CHCl3 : MeOH (1:2) was
change to emergency surgery, which was an exclusion cri-
added to 100 mL of sample and 100 mL of internal stan-
teria. Preoperative blood values, type of surgery, operation
dard solution (lysoPC 14:0, 2.5 nmol; and phosphatidyl-
time, and clamp time were extracted from the electronic
choline [PC] 28:0, 10 nmol). A quantity of 700 mL of HAc
patient data system. Hemodynamic monitoring was per-
0.5%, 1 mL of CHCl3, and 800 mL of HAc 0.5% were
formed by indwelling arterial and pulmonary arterial
added. After each step samples were mixed vigorously for
catheters. Cardiac output was measured by triplet injec-
30 seconds. The final mixture was centrifuged for 10
tion of 10 mL of saline at random during the respiratory
minutes at 1892 ¥ g at room temperature. After centrifu-
cycle. The pulmonary artery occlusion pressure was
gation, the lower layer of CHCl3 was separated. This step
obtained after balloon inflation and wedging, from
was repeated two times by adding 1 mL of CHCl3. The
graphic recordings at end-expiration, without discontinu-
separated CHCl3 layers were combined and dried (N2,
ation of positive end-expiratory pressure (PEEP). All pres-
30°C). Samples were dissolved in 150 mL of CHCl3/
sures were obtained after calibration and zeroing to
MeOH/H2O/NH3 25% (50/45/5/0.01 vol/vol/vol/vol) for
atmospheric pressure, at the time of PLI measurement.
further analysis.
TRALI was defined using the consensus definition of
The relative concentrations of lysoPCs and PC spe-
TRALI (new-onset hypoxemia or deterioration demon-
cies in supernatant of RBCs and PLTs were deter-
strated by a PaO2/FiO2 < 300 mmHg, within 6 hr after
mined using high-performance liquid chromatography
transfusion, with bilateral pulmonary changes, in the
(HPLC) tandem mass spectrometry (HPLC-MS/MS). Ten
absence of elevated left atrial pressure defined as a pul-
microliters of the lipid extraction was injected on the
monary arterial occlusion pressure ⱕ18 mmHg).31-33
HPLC-MS/MS system. Chromatographic separation was
Chest radiographs were routinely taken before surgery
achieved on a modular HPLC system (Surveyor, Thermo
and on arrival at the ICU and assessed by two indepen-
Finnigan, San Jose, CA) consisting of a cooled autosampler
dent physicians blinded to the predictor variables. When
(T = 12°C), a low-flow quaternary MS pump, and analyti-
interpretation differed, chest radiograph and the descrip-
cal HPLC column (LichroSpher Si60, 2 ¥ 250-mm column,
tion by the radiologist were reviewed to receive consensus.
5-mm particle diameter, Merck, Darmstadt, Germany).
The inspiratory O2 fraction (FIO2), PEEP, and tidal volume
Samples were eluted with a flow rate of 300 mL/min and
were derived from the ventilator at the time of the PLI
a programmed linear gradient between Solution B
measurement. Lung injury score was calculated.34 Data on
(chloroform : methanol, 97:3, vol/vol) and Solution A
blood group, donor sex, and storage time of the blood
(methanol : water, 85:15, vol/vol); A and B contained 1
products were obtained from the National Blood Bank.
and 0.1 mL of 25% (vol/vol) aqueous ammonia per liter of
eluent, respectively. The gradient was: T = 0 to 10 minutes,
Analysis of transfused blood products 20% A to 100% A; T = 10-12 minutes, 100% A; T = 12 to 12.1
Samples of the blood bags were centrifuged (1500 ¥ g minutes, 100% A to 0% A; and T = 12.1 to 17 minutes,
for 10 minutes at 4°C for RBCs and at 22°C for PLTs) and equilibration with 0% A. Total run time, including the

equilibration, was 17 minutes. A splitter between the RESULTS

HPLC and MS was used for the introduction of the eluent
in the MS by 75 mL/min.
PLI
MS/MS analyses were performed on a triple quadru- The mean PLI was elevated in all cardiac surgery
pole mass spectrometer (TSQ Quantum AM, Thermo patient groups. Transfusion was associated with a
Finnigan, Waltham, MA) operated in the positive ion further increase in PLI compared to nontransfused con-
electrospray ionization mode. The skimmer offset was set trols (33 ¥ 10-3 ⫾ 20 ¥ 10-3 vs. 23 ¥ 10-3 ⫾ 11 ¥ 10-3 min-1,
at 10 V; spray voltage was 3600 V and the capillary tem- p < 0.01). The PLI did not differ between multiply trans-
perature was 300°C. In the optimized MS/MS experi- fused patients and patients in the restrictive transfusion
ments, argon was used as collision gas at a pressure of group (33 ⫾ 24 vs. 33 ⫾ 16, NS). Patients with strongly
0.07 Pa and a collision energy of 40 V. The parent ion elevated PLI (>1.5¥ the upper limit) had a nonsignificant
scan of m/z 184.1 (m/z 400-m/z 1000, 1 sec) was used for decrease in PaO2/FiO2 compared to patients with a lower
the quantization of the following precursor ions: m/z PLI (ⱕ1.5¥ the upper limit; 275 ⫾ 79 vs. 320 ⫾ 101, NS), as
468.3 (lysoPC 14:0, internal standard), m/z 496.3 (lysoPC well as nonsignificant increase in lung injury score (1.5
16:0), m/z 524.3 (lysoPC 18:0/PLT-activating factor [0-3] vs. 2.0 [0-4], NS). Of the transfused patients with
[PAF] 16:0), m/z 522.4 (lysoPC 18:1), m/z 482.4 (lysoPAF strongly elevated PLI (n = 27), one patient met the clinical
16:0), m/z 510.4 (lysoPAF 18:0), m/z 508.4 (lysoPAF 18:1), criteria of TRALI.
m/z 678.4 (PC 28:0, internal standard), m/z 758.4 (PC
34:2), m/z 782.4 (PC 36:2).
Patient and transfusion risk factors
Patient groups did not differ in alcohol abuse, smoking,
Statistical analysis myocardial infarction, hypertension, diabetes, vascular
The sample size of 60 cardiac surgery patients was cal- diseases, hematologic malignancy, solid malignancy, cere-
culated as follows: a difference of 10% between the brovascular accident, and autoimmune disease (data not
experimental groups in pulmonary leakage index with a shown). Preoperative left ventricular and pulmonary
standard deviation (SD) of 10% and an alpha of 0.05 function did not differ between the groups. Patients
would result in greater than 80% power to detect modest receiving multiple transfusions were older and had a
independent increase in risk for pulmonary leakage higher Euroscore compared to nontransfused patients
caused by blood transfusion. Data were checked for dis- (Table 1, p < 0.001). Multiply transfused patients more
tribution. Normal distributed data were analyzed using often had undergone a combination surgery of coronary
analysis of variance and Dunnett posttest. Nonparamet- artery bypass grafting (CABG) and valve replacement,
ric data were analyzed with Kruskal-Wallis or Mann- while the majority of the restrictive and nontransfused
Whitney U test. Categorical data were analyzed with the patients had undergone CABG. Clamp time, time on car-
chi-square test. To evaluate independent causal factors diopulmonary bypass, and surgery time were longer in the
for an increase in PLI, a logistic regression analysis was multiply transfused patients compared to restrictive and
performed. The association between the amount of nontransfused patients (Table 1, p < 0.01).
blood products transfused was determined in the total Blood group did not differ between groups (data not
cohort (n = 60). The association between the number of shown). The multiply transfused group received a larger
positive products for antibodies and PLI as well as the amount of RBC units stored for more than 14 days com-
association between lysoPC concentration and PLI were pared to the restrictive transfused group (Table 2, p < 0.01).
determined in a separate model using the transfused The number of units with plasma derived from female
patients only (n = 40). The influence of confounding donors, as well as the number of units containing antibod-
and effect modification from significant patient-related ies, was higher in the multiply transfused group compared
covariates was investigated. Confounding was defined as to the restrictive transfused group (p < 0.001 and p < 0.05,
at least 10% change in the PLI coefficient as a conse- respectively). The median amount of lysoPCs per unit of
quence of adding a covariate. Effect modification was RBCs or PLTs did not differ between the two groups. The
defined as a significant p value for the interaction term median storage time for RBCs and PLTs did not differ
added to the model. Covariates without a confounding between the two groups. Univariate analysis in the total
effect or effect modification were excluded from the cohort (n = 60) revealed that the number of RBCs, but not
models. Subanalysis was performed on strongly elevated amount of FFP or number PLTs, was associated with an
PLI (>1.5¥ the upper limit) and less elevated PLI (ⱕ1.5¥ increase in PLI (Table 3, p < 0.05). To investigate the
the upper limit). A p value of 0.05 or less was considered mechanism of transfusion-associated increase in PLI we
significant. Statistical analyses were conducted with the performed a univariate analysis in the transfused patients
use of computer software (SPSS 16, SPSS, Inc., Chicago, only (n = 40). This analysis showed no association between
IL). the total amount, the median concentration of lysoPCs per

VLAAR ET AL.
TABLE 1. Demographic and perioperative characteristics in transfused and nontransfused cardiac surgery
patients*
Transfused groups
Patient characteristics Restrictive (n = 20) Massive (n = 20) Nontransfused group (n = 20)
Age (years) 64 (15) 73 (6)† 64 (11)
Male sex, n (%) 13 (65) 13 (65) 18 (90)
Euroscore 4.3 (2.4)‡ 7.7 (3.5)† 3.6 (1.8)
ASA 2.9 (0.5) 2.7 (0.7) 2.7 (0.6)
Left ventricular function
Poor 0 (0) 1 (5) 0 (0)
Moderate 7 (35) 6 (30) 7 (35)
Good 13 (65) 13 (65) 13 (65)
Preoperative
Hb (g/dL) 8.8 (0.8) 8.0 (1.1)‡ 8.8 (0.7)
PLT count (¥109) 236 (201-273) 219 (184-288) 218 (71-256)
WBC count (¥106) 7.4 (6.3-9.0) 8.5 (6.2-9.8) 6.8 (5.9-8.5)
Postoperative
Hb (g/dL) 5.8 (0.9) 5.4 (0.6) 6.1 (0.2)†
PLT count (¥109) 134 (53) 136 (47) 158 (51)
WBC count (¥106) 12.5 (5.1) 12.4 (5.5) 10.9 (4.8)
Type of surgery, n (%)
CABG 11 (55) 5 (25)‡ 14 (70)
Valve replacement 7 (35) 4 (20) 4 (20)
CABG and valve replacement 1 (5) 11 (55)‡ 1 (5)
Other 1 (5) 0 (0) 1 (5)
Perioperative
Clamp time (min) 69 (49-104) 100 (76-136)‡ 62 (44-92)
Pump time (min) 100 (80-145) 145 (107-179)‡ 102 (68-131)
Total OR time (min) 311 (246-371) 340 (316-390)§ 303 (231-346)
Hemodynamic variables
Heart rate (beats/min) 85 (78-100) 80 (70-95) 71 (65-78)
Mean arterial pressure (mmHg) 63 (47-81) 68 (29-79) 79 (68-89)
PAPdia (mmHg) 15 (11-17) 14 (9-19) 13 (7-17)
PAPsys (mmHg) 26 (22-32) 29 (22-33) 23 (15-29)
PAOP (mmHg) 10 (8-12) 11 (7-15) 10 (8-14)
CO 4.1 (3.3-5.5) 4.6 (2.5-5.2) 5.2 (4.3-6.3)
CI 2.2 (1.6-3.9) 2.4 (1.5-2.8) 2.6 (2.2-2.8)
CVP 7 (3-13) 8 (4-10) 9 (6-12)
sVO2 63 (61-73) 59 (55-72) 69 (66-70)
Respiratory variables
PEEP (cmH2O) 5 (5-5) 5 (5-5) 5 (5-5)
Pressure plateau (cmH2O) 10 (9-12)§ 12 (11-15) 13 (11-15)
FiO2 50 (50-50) 49 (40-50) 50 (40-50)
PaO2 (mmHg) 149 (115-178) 145 (94-194) 150 (112-187)
PaO2/FiO2 297 (242-356) 340 (185-403) 344 (254-390)
Compliance (mL/cmH2O) 84 (65-101) 102 (73-121) 72 (52-98)
Chest X-ray consistent with ALI, n (%) 1 (5) 4 (20) 0 (0)
Rethoracotomy, n (%) 2 (10) 4 (20) 1 (5)
PLI (¥10-3/min) 33 (16)|| 33 (24)|| 23 (11)
* Data are reported as mean (SD) or median (IQR) unless otherwise noted.
† p < 0.001 massive transfused versus nontransfused or restrictive transfused controls.
‡ p < 0.01 massive transfused versus nontransfused or restrictive transfused controls.
§ p < 0.05 massive transfused versus nontransfused or restrictive transfused controls.
|| p < 0.05 all transfused patients versus nontransfused controls.
ASA = American Society of Anesthesia classification; CO = cardiac output; CVP = central venous pressure; IQR = interquartile range;
OR = operating room; PAPdia = pulmonary artery pressure diastolic; PAPsys = pulmonary artery pressure systolic.
unit, the storage time of the products and the presence of (110 [82-214] hr in the restrictive transfused group and
HLA or HNA antibodies, and an increase of the PLI. Addi- 175 [112-218] hr in the multiply transfused group com-
tion of patient factors did not change associations. pared to 82 [58-156] hr in the nontransfused group
[p < 0.01]). Patients in the multiply transfused groups
had a prolonged duration of mechanical ventilation
Outcome compared to the nontransfused group (14 [10-20] hr vs.
Transfused patients had an increased median length 9.5 [6.3-12.8] hr, p < 0.001). No difference was found in
of hospital stay compared to nontransfused patients survival or ICU stay between the groups.

TABLE 2. Transfusion descriptives and pulmonary leakage index of transfused cardiac surgery patients*
Transfused groups PLI level
Transfusion characteristics Restrictive (n = 20) Massive (n = 20) PLI < 1.5¥ (n = 13) PLI ⱖ 1.5¥ (n = 27)
Number of transfusions (units)
RBCs 1.1 ⫾ 0.8 5.7 ⫾ 3.8† 2.9 ⫾ 4.3 3.5 ⫾ 3.4
FFP 0.1 ⫾ 0.4 3.5 ⫾ 2.6† 2.3 ⫾ 3.0 1.6 ⫾ 2.4
PLTs 0.2 ⫾ 0.4 1.4 ⫾ 0.6† 0.8 ⫾ 0.6 0.7 ⫾ 0.8
RBCs
Storage time >14 days (units) 1 (0-1) 3 (0-5)‡ 2 (0-3) 1 (0-4)
Storage time (days) 15 (8-19) 15 (11-18) 15 (11-23) 15 (10-17)
PLT
Storage time >5 days (units) 0 (0-1) 0 (0-1) 1 (0-1)§ 0 (0-0)
Storage time (days) 3 (1-6) 4 (0-7) 6 (2-7) 0 (0-4)
Antibodies
Number of units derived from female donors 1 (0-1) 2 (2-4.5)† 2 (0-3) 1 (1-2)
Number of antibody-positive units 0 (0-1) 1 (0-2)§ 1 (0-1) 1 (0-1)
Number of HLA I- or II-positive units 0 (0-0) 0 (0-1) 0 (0-1) 0 (0-1)
Number of HNA-positive units 0 (0-01) 1 (0-1) 1 (0-1) 1 (0-1)
Median concentration of lysoPC in RBCs
LysoPC
16:0 mMol 6.8 (6.0-8.6) 7.3 (6.1-8.8) 7.0 (6.5-8.8) 7.0 (6.0-8.9)
18:0 mMol 3.5 (2.9-4.1) 3.7 (3.3-4.5) 3.8 (3.3-4.5) 3.5 (3.0-4.3)
18:1 mMol 2.0 (2.0-3.0) 2.0 (2.0-2.5) 2.0 (2.0-3.0) 2.0 (2.0-2.5)
LysoPAF
16:0 mMol 0.2 (0.2-0.3) 0.3 (0.2-0.3) 0.2 (0.2-0.3) 0.3 (0.2-0.3)
18:0 mMol 0.3 (0.3-0.4) 0.4 (0.3-0.4) 0.3 (0.3-0.4) 0.4 (0.3-0.4)
LysoPC 18:1/PAF 16:0 mMol 0.1 (0.1-0.2) 0.1 (0.1-0.1) 0.1 (0.1-0.1) 0.1 (0.1-0.2)
Median concentration of lysoPC in PLTs
LysoPC
16:0 mMol 98 (89-100) 97 (87-111) 102 (90-110) 91 (86-100)
18:0 mMol 40 (40-56) 41 (38-45) 43 (41-50) 40 (38-44)
18:1 mMol 18 (17-22) 21 (18-22) 21 (18-21) 21 (18-22)
LysoPAF
16:0 mMol 2.0 (1.7-2.1) 1.8 (1.6-2.0) 1.9 (1.5-2.0) 1.7 (1.7-2.1)
18:0 mMol 2.3 (2.0-2.5) 2.3 (2.2-2.9) 2.5 (2.3-2.8) 2.3 (2.0-2.4)
LysoPC 18:1/PAF 16:0 mMol 0.8 (0.7-1.0) 0.9 (0.8-1.0) 0.9 (0.8-1.0) 0.8 (0.7-0.9)
* Data are presented as mean (SD) or as median (IQR).
† p < 0.001.
‡ p < 0.01.
§ p < 0.05.
DISCUSSION The finding that transfusion results in mild lung

injury is in accord with a previous description of three
This study shows that blood transfusion is associated with cases not meeting the TRALI consensus definition and
increased pulmonary vascular leakage in a cohort of who were not diagnosed as TRALI by the treating clini-
cardiac surgery patients, in a dose-dependent manner for cians.39 Also, in an experimental transfusion model, we
RBCs. Vascular injury after transfusion in this cohort of recently showed that transfusion of stored RBCs results in
cardiac surgery patients not meeting the TRALI criteria mild pulmonary inflammation, including extravasation of
was irrespective of the presence of lysoPCs or antibodies. pulmonary neutrophils and production of proinflamma-
We found in the regression analysis that blood trans- tory cytokines.40 These results indicate that pulmonary
fusion is associated with onset of pulmonary vascular leakage was dose dependent for RBCs, in line with the
leakage with concomitant trend for worsening of lung suggestion that mild effects of a single transfusion may
injury, in patients that do not meet the clinical TRALI cri- accumulate after repeated transfusions, contributing to
teria. The finding of similar filling pressures and cardiac pulmonary leakage and hypoxemia after cardiac surgery.
function in restricted and multiply transfused patients In accordance, observational clinical studies show that the
underlines the fact that pulmonary deterioration was not number of RBCs transfused is associated with adverse
due to hydrostatic pulmonary edema. Also, operative pro- outcome.4,41
cedures and time on cardiopulmonary bypass did not We found no association between the presence of HLA
influence the association between transfusion and pul- or HNA antibodies in the transfused products and pulmo-
monary leakage, in line with a previous study indicating nary leakage. Although antibodies are implicated in the
an independent role for transfusion as mediator of pulmo- onset of TRALI, many antibody-containing blood products
nary dysfunction after cardiac surgery.29 fail to produce TRALI.42-44 A threshold model has been sug-

VLAAR ET AL.
the second hit of TRALI pathogenesis.17,48

TABLE 3. Univariate analysis of transfusion-related risk factors for an Here, we found no association between
increase in PLI in cardiac surgery patients*
bioactive lipids in the transfused prod-
Transfusion-related risk factors
ucts and pulmonary vascular leakage. In
Transfusion characteristics b (95% CI) p value
accordance, a recent study in surgical
Number of transfusions (units), n = 60
RBCs 1.6 (0.2 to 3.0) 0.03† patients showed no association between
FFP 0.4 (-1.8 to 2.6) 0.7 the concentration of bioactive lipids and
PLTs 1.2 (-5.9 to 8.2) 0.7 the occurrence of postoperative pulmo-
RBCs (n = 40)
Storage time >14 days (units) -0.9 (-3.2 to 1.3) 0.4 nary dysfunction.49 Also, we did not find
Storage time (days) -0.8 (-2.1 to 0.5) 0.2 an association between storage time
PLT (n = 40) and pulmonary leakage. This is in con-
Storage time >5 days (units) -8.1 (-27.1 to 10.8) 0.4
Storage time (days) -1.3 (-5.1 to 2.6) 0.5 trast with experimental studies17,40,50,51
Antibodies (n = 40) and with an observational study describ-
Number of female donors -0.7 (-4.7 to 3.3) 0.7 ing an association between RBC storage
Number of antibody-positive units -3.8 (-15.5 to 7.9) 0.5
Number of HLA I- or II-positive units -12.0 (-36.1 to 12.0) 0.3 time and pulmonary complications,2 but
Number of HNA-positive units -5.2 (-21.2 to 10.8) 0.5 not with other studies in cardiac surgery
Median concentration of lysoPC in RBCs (n = 40) patients investigating the role of aged
LysoPC
16:0 mMol -0.6 (-5.4 to 4.2) 0.8 RBCs.22-24 Taken together, we cannot
18:0 mMol 2.1 (-7.3 to 11.4) 0.7 rule out the fact that a larger sample
18:1 mMol 2.5 (-15.2 to 20.3) 0.8 size may have yielded an association
LysoPAF
16:0 mMol 54.3 (-84.4 to 193) 0.4 between lysoPC and pulmonary leakage.
18:0 mMol 80.3 (-37.9 to 198.5) 0.2 However, in this cohort of patients,
LysoPC 18:1/PAF 16:0 mMol 68 (-146.3 to 282.0) 0.5 blood transfusion was clearly associated
Median concentration of lysoPC in PLTs (n = 40)
LysoPC with increased PLI, while the association
16:0 mMol -0.3 (-1.0 to 0.5) 0.5 between lysoPCs and PLI was absent.
18:0 mMol -0.4 (-1.8 to 1.0) 0.5 Possibly, the RBC, and not soluble
18:1 mMol 0.6 (-2.7 to 4.0) 0.7
LysoPAF factors in the supernatant, is a key player
16:0 mMol 6.6 (-28.5 to 41.7) 0.7 in onset of lung injury, as suggested by
18:0 mMol 1.9 (-11.6 to 15.4) 0.8 recent experimental studies.40,52
LysoPC 18:1/PAF 16:0 mMol -27.1 (-90.2 to 35.8) 0.4
The design of this study has limita-
* Data are presented as mean (SD) or as median (IQR).
† Significant difference. tions. Sample size is small, which may
have precluded positive associations.
Also, although we used a regression
gested,45 in which a threshold must be overcome to induce analysis with known risk factors for transfusion and onset
a TRALI reaction. Factors that determine the threshold are of ALI to correct for confounding, it is possible that factors
the predisposition of the patient that determines priming not included in the model have contributed to PLI
of the lung neutrophils and the ability of the mediators in increase. Therefore, results on causative factors in the
the transfusion to cause activation of primed neutrophils. increase in PLI should be interpreted within these limits.
An explanation for the absence of increase of pulmonary We cannot exclude that antibodies or lysoPCs are caus-
leakage in the presence of HLA or HNA antibodies may be ative when a larger study sample is investigated. However,
twofold. First, the neutrophil-priming status of the patient given the clear association between transfusion and
may have been too low, thereby not allowing the threshold elevated PLI, the investigated factors may not be strong
for onset of lung injury to be overcome. Second, we only predictors.
determined HLA or HNA antibodies in the blood products, In conclusion, transfusion in cardiothoracic surgery
not the presence of an antibody-antigen match. Third, a patients is associated with an increase in pulmonary
mild increase in pulmonary permeability in cardiac capillary permeability, an effect that was dose dependent
surgery patients may be caused by other pathways than for RBC products. In this study, we found no associa-
those involved in the onset of TRALI. Although recent tion between bioactive lipids or the presence of HLA or
studies suggest that exclusion of female donors for produc- HNA antibodies and increase in pulmonary capillary
tion of plasma reduces pulmonary complications and permeability.
TRALI,46,47 current results do not support a male-only
donor policy with the aim to reduce pulmonary dysfunc-
tion in cardiac surgery patients in general.
CONFLICT OF INTEREST
Bioactive lipids, which accumulate during storage of
cell-containing blood products, have been implicated in None.

REFERENCES 16. Silliman CC, Clay KL, Thurman GW, Johnson CA, Ambruso
DR. Partial characterization of lipids that develop during
1. Koch CG, Li L, Duncan AI, Mihaljevic T, Cosgrove DM, the routine storage of blood and prime the neutrophil
Loop FD, Starr NJ, Blackstone EH. Morbidity and mortality NADPH oxidase. J Lab Clin Med 1994;124:684-94.
risk associated with red blood cell and blood-component 17. Silliman CC, Voelkel NF, Allard JD, Elzi DJ, Tuder RM,
transfusion in isolated coronary artery bypass grafting. Crit Johnson JL, Ambruso DR. Plasma and lipids from stored
Care Med 2006;34:1608-16. packed red blood cells cause acute lung injury in an
2. Koch CG, Li L, Sessler DI, Figueroa P, Hoeltge GA, Mihal- animal model. J Clin Invest 1998;101:1458-67.
jevic T, Blackstone EH. Duration of red-cell storage and 18. Silliman CC, Ambruso DR, Boshkov LK. Transfusion-
complications after cardiac surgery. N Engl J Med 2008; related acute lung injury. Blood 2005;105:2266-73.
358:1229-39. 19. Zallen G, Offner PJ, Moore EE, Blackwell J, Ciesla DJ,
3. Blumberg N. Deleterious clinical effects of transfusion Gabriel J, Denny C, Silliman CC. Age of transfused blood is
immunomodulation: proven beyond a reasonable doubt. an independent risk factor for postinjury multiple organ
Transfusion 2005;45(Suppl):33S-9S. failure. Am J Surg 1999;178:570-2.
4. Gajic O, Rana R, Mendez JL, Rickman OB, Lymp JF, 20. Leal-Noval SR, Jara-Lopez I, Garcia-Garmendia JL, Marin-
Hubmayr RD, Moore SB. Acute lung injury after blood Niebla A, Herruzo-Aviles A, Camacho-Larana P, Loscertales
transfusion in mechanically ventilated patients. Transfu- J. Influence of erythrocyte concentrate storage time on
sion 2004;44:1468-74. postsurgical morbidity in cardiac surgery patients. Anes-
5. Gajic O, Dzik WH, Toy P. Fresh frozen plasma and platelet thesiology 2003;98:815-22.
transfusion for nonbleeding patients in the intensive care 21. Vamvakas EC, Carven JH. Transfusion and postoperative
unit: benefit or harm? Crit Care Med 2006;34(Suppl):S170- pneumonia in coronary artery bypass graft surgery: effect
S173. of the length of storage of transfused red cells. Transfusion
6. Ho J, Sibbald WJ, Chin-Yee IH. Effects of storage on efficacy 1999;39:701-10.
of red cell transfusion: when is it not safe? Crit Care Med 22. van Straten AH, Soliman Hamad MA, Zundert AA, Martens
2003;31(Suppl):S687-S697. EJ, Woorst JF, Wolf AM, Scharnhorst V. Effect of duration of
7. Silliman CC, Boshkov LK, Mehdizadehkashi Z, Elzi DJ, red blood cell storage on early and late mortality after
Dickey WO, Podlosky L, Clarke G, Ambruso DR. coronary artery bypass grafting. J Thorac Cardiovasc Surg
Transfusion-related acute lung injury: epidemiology and a 2011;141:231-7.
prospective analysis of etiologic factors. Blood 2003;101: 23. Yap CH, Lau L, Krishnaswamy M, Gaskell M, Yii M. Age
454-62. of transfused red cells and early outcomes after cardiac
8. Vlaar AP, Binnekade JM, Prins D, van Stein D, Hofstra JJ, surgery. Ann Thorac Surg 2008;86:554-9.
Schultz MJ, Juffermans NP. Risk factors and outcome of 24. van de Watering L, Lorinser J, Versteegh M, Westendord R,
transfusion-related acute lung injury in the critically ill: a Brand A. Effects of storage time of red blood cell transfu-
nested case-control study. Crit Care Med 2010;38:771-8. sions on the prognosis of coronary artery bypass graft
9. Ng CS, Wan S, Yim AP, Arifi AA. Pulmonary dysfunction patients. Transfusion 2006;46:1712-8.
after cardiac surgery. Chest 2002;121:1269-77. 25. Abraham E, Matthay MA, Dinarello CA, Vincent JL, Cohen
10. Kopko PM, Marshall CS, MacKenzie MR, Holland PV, Pop- J, Opal SM, Glauser M, Parsons P, Fisher CJ Jr, Repine JE.
ovsky MA. Transfusion-related acute lung injury: report of Consensus conference definitions for sepsis, septic shock,
a clinical look-back investigation. JAMA 2002;287:1968-71. acute lung injury, and acute respiratory distress syndrome:
11. Palfi M, Berg S, Ernerudh J, Berlin G. A randomized con- time for a reevaluation. Crit Care Med 2000;28:232-5.
trolled trial of transfusion-related acute lung injury: is 26. Li G, Daniels CE, Kojicic M, Krpata T, Wilson GA, Winters
plasma from multiparous blood donors dangerous? Trans- JL, Moore SB, Gajic O. The accuracy of natriuretic peptides
fusion 2001;41:317-22. (brain natriuretic peptide and N-terminal pro-brain natri-
12. Popovsky MA, Moore SB. Diagnostic and pathogenetic uretic) in the differentiation between transfusion-related
considerations in transfusion-related acute lung injury. acute lung injury and transfusion-related circulatory over-
Transfusion 1985;25:573-7. load in the critically ill. Transfusion 2009;49:13-20.
13. Seeger W, Schneider U, Kreusler B, von Witzleben E, 27. Verheij J, van Lingen A, Raijmakers PG, Spijkstra JJ, Girbes
Walmrath D, Grimminger F, Neppert J. Reproduction of AR, Jansen EK, van den Berg FG, Groeneveld AB. Pulmo-
transfusion-related acute lung injury in an ex vivo lung nary abnormalities after cardiac surgery are better
model. Blood 1990;76:1438-44. explained by atelectasis than by increased permeability
14. Geelhoed GW, Bennett SH. “Shock lung” resulting from oedema. Acta Anaesthesiol Scand 2005;49:1302-10.
perfusion of canine lungs with stored bank blood. Am Surg 28. Groeneveld AB, Raijmakers PG. The 67gallium-transferrin
1975;41:661-82. pulmonary leak index in patients at risk for the acute
15. Popovsky MA. Transfusion and lung injury. Transfus Clin respiratory distress syndrome. Crit Care Med 1998;26:685-
Biol 2001;8:272-7. 91.

VLAAR ET AL.
29. Groeneveld AB, Jansen EK, Verheij J. Mechanisms of pul- anti-human leucocyte antigen class II antibodies: a look-
monary dysfunction after on-pump and off-pump cardiac back investigation. Transfus Med 2004;14:225-30.
surgery: a prospective cohort study. J Cardiothorac Surg 43. Toy P, Hollis-Perry KM, Jun J, Nakagawa M. Recipients of
2007;2:11. blood from a donor with multiple HLA antibodies: a look-
30. Raijmakers PG, Groeneveld AB, Schneider AJ, Teule GJ, van back study of transfusion-related acute lung injury. Trans-
Lingen A, Eijsman L, Thijs LG. Transvascular transport of fusion 2004;44:1683-8.
67Ga in the lungs after cardiopulmonary bypass surgery. 44. Van Buren NL, Stroncek DF, Clay ME, McCullough J, Dal-
Chest 1993;104:1825-32. masso AP. Transfusion-related acute lung injury caused by
31. Goldman M, Webert KE, Arnold DM, Freedman J, Hannon an NB2 granulocyte-specific antibody in a patient with
J, Blajchman MA. Proceedings of a consensus conference: thrombotic thrombocytopenic purpura. Transfusion 1990;
towards an understanding of TRALI. Transfus Med Rev 30:42-5.
2005;19:2-31. 45. Bux J, Sachs UJ. The pathogenesis of transfusion-related
32. Kleinman S, Caulfield T, Chan P, Davenport R, McFarland acute lung injury (TRALI). Br J Haematol 2007;136:788-
J, McPhedran S, Meade M, Morrison D, Pinsent T, Robil- 99.
lard P, Slinger P. Toward an understanding of transfusion- 46. Vlaar AP, Binnekade JM, Schultz MJ, Juffermans NP,
related acute lung injury: statement of a consensus panel. Koopman MM. Preventing TRALI: ladies first, what follows?
Transfusion 2004;44:1774-89. Crit Care Med 2008;36:3283-4.
33. Toy P, Popovsky MA, Abraham E, Ambruso DR, Holness 47. Wright SE, Snowden CP, Athey SC, Leaver AA, Clarkson JM,
LG, Kopko PM, McFarland JG, Nathens AB, Silliman CC, Chapman CE, Roberts DR, Wallis JP. Acute lung injury
Stroncek D. Transfusion-related acute lung injury: defini- after ruptured abdominal aortic aneurysm repair: the
tion and review. Crit Care Med 2005;33:721-6. effect of excluding donations from females from the pro-
34. Murray JF, Matthay MA, Luce JM, Flick MR. An expanded duction of fresh frozen plasma. Crit Care Med 2008;36:
definition of the adult respiratory distress syndrome. Am 1796-802.
Rev Respir Dis 1988;138:720-3. 48. Silliman CC, Paterson AJ, Dickey WO, Stroneck DF, Pop-
35. Patel R, Terasaki PI. Significance of the positive crossmatch ovsky MA, Caldwell SA, Ambruso DR. The association
test in kidney transplantation. N Engl J Med 1969;280:735-9. of biologically active lipids with the development of
36. Décary F, Vermeulen A, Engielfriet C. A look at HLA anti- transfusion-related acute lung injury: a retrospective study.
sera in the indirect immunofluorescence technique (LIFT). Transfusion 1997;37:719-26.
In: Kissmeyer-Nielsen F, editor. Histocompatibility testing. 49. Nakazawa H, Ohnishi H, Okazaki H, Hashimoto S, Hotta H,
Copenhagen: Munksgaard; 1975. p. 380-90. Watanabe T, Ohkawa R, Yatomi Y, Nakajima K, Iwao Y,
37. Jiang AF, Lalezari P. A micro-technique for detection of Takamoto S, Shimizu M, Iijima T. Impact of fresh-frozen
leukocyte agglutinins. J Immunol Methods 1975;7:103-8. plasma from male-only donors versus mixed-sex donors
38. Verheugt FW, von dem Borne AE, Decary F, Engelfriet CP. on postoperative respiratory function in surgical patients:
The detection of granulocyte alloantibodies with an indi- a prospective case-controlled study. Transfusion 2009;49:
rect immunofluorescence test. Br J Haematol 1977;36:533- 2434-41.
44. 50. Silliman CC, Bjornsen AJ, Wyman TH, Kelher M, Allard J,
39. Davis A, Mandal R, Johnson M, Makar R, Stowell C, Dzik S. Bieber S, Voelkel NF. Plasma and lipids from stored plate-
A touch of TRALI. Transfusion 2008;48:541-5. lets cause acute lung injury in an animal model. Transfu-
40. Vlaar AP, Hofstra JJ, Levi M, Kulik W, Nieuwland R, Tool sion 2003;43:633-40.
AT, Schultz MJ, de Korte D, Juffermans NP. Supernatant of 51. Vlaar AP, Hofstra JJ, Kulik W, van Lenthe H, Nieuwland R,
aged erythrocytes causes lung inflammation and coagul- Schultz MJ, Levi MM, Roelofs JJ, Tool AT, de Korte D, Juf-
opathy in a “two-hit” in vivo syngeneic transfusion model. fermans NP. Supernatant of stored platelets causes lung
Anesthesiology 2010;113:92-103. inflammation and coagulopathy in a novel in vivo transfu-
41. Gong MN, Thompson BT, Williams P, Pothier L, Boyce PD, sion model. Blood 2010;116:1360-8.
Christiani DC. Clinical predictors of and mortality in acute 52. Mangalmurti NS, Xiong Z, Hulver M, Ranganathan M, Liu
respiratory distress syndrome: potential role of red cell XH, Oriss T, Fitzpatrick M, Rubin M, Triulzi D, Choi A, Lee
transfusion. Crit Care Med 2005;33:1191-8. JS. Loss of red cell chemokine scavenging promotes
42. Nicolle AL, Chapman CE, Carter V, Wallis JP. Transfusion- transfusion-related lung inflammation. Blood 2009;113:
related acute lung injury caused by two donors with 1158-66.

Milrinone and Mortality in Adult Cardiac Surgery: A Meta-analysis
Alberto Zangrillo, MD,* Giuseppe Biondi-Zoccai, MD,† Martin Ponschab, MD,‡ Massimiliano Greco, MD,*
Laura Corno, MD,* Remo Daniel Covello, MD,* Luca Cabrini, MD,* Elena Bignami, MD,*
Giulio Melisurgo, MD,* and Giovanni Landoni, MD*
Objective: The authors conducted a review of randomized Subanalyses confirmed increased mortality with milrinone
studies to show whether there are any increases or de- (9/84 deaths [10.7%] v 3/105 deaths [2.9%] with other drugs
creases in survival when using milrinone in patients under- as control, OR ⴝ 4.19 [1.27-13.84], p ⴝ 0.02) with 189 pa-
going cardiac surgery. tients and 5 studies included) but did not confirm a differ-
Design: A meta-analysis. ence in mortality (4/165 [2.4%] in the milrinone group v
Setting: Hospitals. 3/164 [1.8%] with placebo or nothing as control, OR ⴝ 1.27
Participants: Five hundred eighteen patients from 13 ran- [0.28-5.84], p ⴝ 0.76 with 329 patients and 8 studies in-
domized trials. cluded).
Interventions: None. Conclusions: This analysis suggests that milrinone might
Measurements and Main Results: BioMedCentral, PubMed increase mortality in adult patients undergoing cardiac sur-
gery. The effect was seen only in patients having an active
EMBASE, the Cochrane central register of clinical trials, and
inotropic drug for comparison and not in the placebo sub-
conference proceedings were searched for randomized trials
group. Therefore, the question remains whether milrinone
that compared milrinone versus placebo or any other con-
increased mortality or if the control inotropic drugs were
trol in the setting of cardiac surgery that reported data on
more protective.
mortality. Overall analysis showed that milrinone increased
© 2012 Elsevier Inc. All rights reserved.
perioperative mortality (13/249 [5.2%] in the milrinone
group v 6/269 [2.2%] in the control arm, odds ratio [OR] ⴝ KEY WORDS: anesthesia, cardiac surgery, complications,
2.67 [1.05-6.79], p for effect ⴝ 0.04, p for heterogeneity ⴝ outcomes, heart failure, milrinone, mortality, intensive care,
0.23, I2 ⴝ 25% with 518 patients and 13 studies included). inotropic agents
M YOCARDIAL DYSFUNCTION requiring inotropic

support is a complication of cardiac surgery and can
increase the risk of early postoperative death.1,2 Treatment
patients suggest that the use of epinephrine is associated with
worse clinical outcomes, including a higher incidence of renal
dysfunction.6
includes optimization of loading conditions and myocardial PDE III inhibition by agents such as milrinone and enoxi-
contractility through appropriate fluid and pharmacologic man- mone provides an alternative means of inotropic support.5 In
agement2 and mechanical support. Extensive use is made of addition to their positive inotropic effects, PDE inhibitors also
inotropes in this situation, but the optimal pharmacologic man- have vasodilatory effects because of PDE inhibition in vascular
agement of myocardial dysfunction in cardiac surgery is a smooth muscle cells.4 No adverse metabolic effects have been
matter of ongoing debate.3 reported during treatment with PDE III inhibitors, and their
Inotropic agents increase myocardial contractility through pre-emptive use has been shown to exert beneficial effects on
different pathways that in most cases lead ultimately to an markers of renal tubular injury.6 Milrinone does not produce as
increase in intracellular cyclic adenylate monophosphate great an increase in heart rate or myocardial oxygen consump-
(cAMP) levels which, in turn, induces an increase in calcium tion as dobutamine.7 Moreover, milrinone may reduce the
release from the sarcoplasmic reticulum, enhancing contractile incidence of postoperative myocardial ischemia and infarction
force. Increases in cAMP levels may be achieved either via in patients with impaired left ventricular function undergoing
␤-adrenoceptor–mediated stimulation of adenylate cyclase or coronary artery bypass graft (CABG) surgery.1
by the selective inhibition of phosphodiesterase (PDE) III, It is clearly desirable to know if these encouraging indica-
which is the enzyme that catalyzes the elimination of cAMP.4 tions could be translated into an effect of PDE inhibitors on
Adrenoceptor-based inotropy faces various substantial limi- major clinical outcomes in cardiac surgery patients. Therefore,
tations. Myocardial ␤-adrenergic receptor desensitization oc- the authors conducted a meta-analysis of randomized con-
curs acutely after cardiopulmonary bypass (CPB), thereby lim- trolled trials to investigate the effects of milrinone on survival
iting the efficacy of ␤-adrenergic stimulants for postbypass in this setting.
cardiac failure.5 Moreover, observational data in critically ill
METHODS
From the *Department of Anesthesia and Intensive Care, Universita Search Strategy
Vita-Salute San Raffaele, Milan, Italy; †Division of Cardiology, Uni-
versity of Modena and Reggio Emilia, Modena, Italy; and ‡Department Pertinent studies were independently searched in BioMedCentral and
of Anaesthesia and Intensive Care, Trauma Hospital Linz, Linz, PubMed EMBASE and the Cochrane central register of clinical trials
Austria. (updated November 1, 2010) by 4 trained investigators. The full
Address reprint requests to Giovanni Landoni, MD, Department of PubMed search strategy developed according to Biondi-Zoccai et al8 is
Cardiothoracic Anesthesia and Intensive Care, Istituto Scientifico San available in Appendix 1 and aimed to include any randomized study
Raffaele, Via Olgettina 60, Milan 20132, Italy. E-mail: landoni. ever performed with milrinone in adult cardiac surgery. In addition, the
giovanni@hsr.it authors used backward snowballing (ie, scanning of references cited in
© 2012 Elsevier Inc. All rights reserved. retrieved articles and pertinent reviews) and contacted international
1053-0770/2601-0012$36.00/0 experts for further studies. No language restriction was enforced and
doi:10.1053/j.jvca.2011.06.022 non–English-language articles were translated before further analysis.
70 Journal of Cardiothoracic and Vascular Anesthesia, Vol 26, No 1 (February), 2012: pp 70-77
MILRINONE INCREASES MORTALITY 71
Study Selection inconsistency were measured using Cochran Q tests and I2, respec-
tively. Binary outcomes from individual studies were analyzed in order
References obtained from database and literature searches first were
to compute individual and pooled odds ratios (ORs) with pertinent 95%
examined independently at a title/abstract level by 4 investigators, with
confidence intervals (with equivalence set at 1, OR ⬍1 favoring the
divergences resolved by consensus and then, if potentially pertinent,
first treatment, and OR ⬎1 favoring the second treatment) by means of
retrieved as complete articles. The following inclusion criteria were
the Peto fixed-effect method in case of low statistical inconsistency
used for potentially relevant studies: random allocation to treatment,
(I2 ⱕ 5%) and by means of a random-effect method (which better
comparison of milrinone versus control with no restriction in dose and
accommodates clinical and statistical variations) in case of moderate or
time of administration, information on the primary outcome (mortal-
high statistical inconsistency (I2 ⬎ 25%). Weighted mean differences
ity), and cardiac surgery performed in adult patients. The exclusion
and 95% CIs were computed for continuous variables, again by means
criteria were as follows: duplicate publications (in this case only the
of a fixed-effect method in case of low statistical inconsistency (I2 ⱕ
article reporting the longest follow-up was considered), nonhuman
25%) and by means of a random-effect method in case of moderate or
experimental studies, and a lack of mortality data. Two investigators
high statistical inconsistency (I2 ⬎ 25%).9 The risk of small study bias
independently assessed compliance to selection criteria and selected
(including publication bias, ie, the risk of small nonsignificant studies
studies for the final analysis.
being selectively rejected by medical journals) was assessed by visual
inspection of funnel plots and by analytic appraisal based on the Peters’
Data Extraction and Study Characteristics
regression asymmetry test.
Baseline, procedural, and outcome data were extracted indepen- Statistical significance was set at the 2-tailed 0.05 level for hypoth-
dently by 4 trained investigators, with divergences resolved by con- esis testing and at 0.10 for heterogeneity testing, and unadjusted
sensus. Specifically, the authors extracted study endpoints and main p values are reported throughout. This study was performed in com-
outcomes, study design, population, clinical setting, milrinone dosage, pliance with The Cochrane Collaboration and the Preferred Reporting
and treatment duration. At least 2 separate attempts at contacting the Items for Systematic Reviews and Meta-Analyses guidelines.9
original authors were made in case of missing data. The primary
endpoint of the present review was mortality. Secondary endpoints RESULTS
were myocardial infarction (as per author definition), acute renal failure
(as per author definition), arrhythmias, mechanical ventilation, and Database searches, snowballing, and contacts with experts
length of intensive care unit and hospital stay. yielded a total of 345 articles. Excluding 304 nonpertinent
titles or abstracts, 41 studies were retrieved in complete form
Internal Validity and Risk of Bias Assessment and assessed according to the selection criteria (Fig 1). Of
The internal validity and risk of bias of the included trials were these, 28 studies were excluded because there were no
appraised according to Cochrane Collaboration methods by 2 indepen- outcome data and further details could not be obtained by the
dent reviewers. authors (n ⫽ 23)6,10-31 because of inhaled administration of
milrinone (n ⫽ 2),32,33 administration to children (n ⫽ 1),34
Data Analysis and Synthesis or duplicate publication (n ⫽ 2).5,35 Thirteen eligible ran-
Computations were performed with RevMan 5.0 (a freeware avail- domized clinical trials were included in the final analysis
able from The Cochrane Collaboration). Statistical heterogeneity and (Table 1).1,2,36-46
Fig 1. A review profile.

72 ZANGRILLO ET AL
Table 1. Description of the Studies Included in the Meta-analysis
First Author Journal Year Cardiac Surgery Procedures Control Multicentric
Al-Shawaf2 J Cardiothorac Vasc Anesth 2006 CABG Levosimendan No

Arbeus36 J Cardiothorac Vasc Anesth 2009 CABG Placebo No
Brackbill37 Ann Pharmacother 2007 CABG and valvular surgery Nesiritide No
Couture38 Can J Anaesth 2007 CABG Placebo No
De Hert39 Anesth Analg 2007 CABG and valvular surgery Levosimendan No
De Hert40 J Cardiothorac Vasc Anesth 2008 CABG and valvular surgery Levosimendan No
Doolan41 J Cardiothorac Vasc Anesth 1997 CABG and valvular surgery Placebo No
Hayashida42 Ann Thorac Surg 1999 CABG Nothing No
Jebeli1 Cardiol J 2010 CABG Placebo No
Jo43 Korean J Anesthesiol 2010 CABG Placebo No
Mollhoff 44 Anesthesiology 1999 CABG Placebo No
Mollhoff45 Eur J Anaesthesiol 2002 CABG Nifedipine No
Shi46 J Thorac Cardiovasc Surg 2006 CABG Placebo No
Study Characteristics Quantitative Data Synthesis

The 13 included trials randomized 518 patients (249 receiving The overall analysis showed that the use of milrinone was
milrinone and 269 receiving control) (Tables 2 and 3). Nine of associated with a significant increase in mortality (13/249
these studies included patients who underwent isolated CABG [5.2%] in the milrinone group v 6/269 [2.2%] in the control
surgery,1,2,36,38,42-46 and 4 studies included patients undergoing arm, OR ⫽ 2.67 [1.05-6.79], p for effect ⫽ 0.04, p for heter-
CABG or valve surgery.37,39-41 In 8 studies, milrinone was ogeneity ⫽ 0.23, I2 ⫽ 25% with 518 patients and 13 studies
administered as a bolus.1,2,36-38,41,44,46 In 12 studies, milrinone included) (Fig 2). Sensitivity analyses and funnel plot inspec-
was used as a continuous infusion,1,2,37-46 preceded in 7 studies tion confirmed the overall robustness of the present findings
by an initial bolus.1,2,37,38,41,45,46 The dose varied between 30 and and the lack of evidence for small study bias, respectively
50 ␮g/kg (as an intravenous bolus) and between 0.3 and 0.5 (Fig 3).3 Quantitative evaluation did not suggest publication
␮g/kg/min (as a continuous infusion). Eight studies compared bias as measured by the Peters’ test (p ⫽ 0.75).
milrinone with placebo (n ⫽ 7)1,36,38,41,43,44,46 or nothing,42 in 3 A subanalysis of studies with placebo or nothing as the
studies levosimendan was administered,2,39,40 and 1 study used control showed no difference in the risk of mortality (4/165
either nesiritide37 or nifedipine45 as active controls. [2.4%] in the milrinone group v 3/164 [1.8%] in the control
Studies appeared to be of variable quality. Specifically, alarm, OR ⫽ 1.27 [0.28-5.84], p for effect ⫽ 0.76, p for heter-
though 4 of the randomized controlled trials were regarded to ogeneity ⫽ 0.45, I2 ⫽ 0% with 329 patients and 8 studies
be high quality, many other studies lacked important details to included). A similar subanalysis of studies that compared mil-
appraise the risk of selection, performance, attrition, or detec- rinone with other drugs revealed an increased risk of mortality
tion biases (Table 4). with milrinone (9/84 deaths [10.7%] v 3/105 deaths [2.9%] in
Table 2. Number of Patients and Interventions of Included Studies
No. of Patients No. of Patients

Receiving in Control
First Author Milrinone Group Control Time of Administration Bolus Milrinone Dose Length of Infusion
Al-Shawaf2 14 16 Levosimendan Post operative LCOS within 50 ␮g/kg 0.3-0.5 ␮g/kg/min 24 h

12 hours from CPB
Arbeus36 22 22 Placebo After release of aortic clamp 50 ␮g/kg No
Brackbill37 20 20 Nesiritide After release of aortic clamp 50 ␮g/kg 0.375 ␮g/kg/min 24 h or more
Couture38 25 25 Placebo After anesthesia induction 50 ␮g/kg 0.5 ␮g/kg/min Until skin closure
De Hert39 15 15 Levosimendan After release of aortic clamp No 0.5 ␮g/kg/min 19 ⫾ 4 hours
De Hert40 20 40 Levosimendan After release of aortic clamp No 0.5 ␮g/kg/min 22-23 hours
Doolan41 15 15 Placebo 15 minutes before weaning 50 ␮g/kg 0.5 ␮g/kg/min 4 h or longer
from CPB
Hayashida42 12 12 Nothing After anesthesia induction No 0.5 ␮g/kg/min 24 h
Jebeli1 35 35 Placebo After release of aortic clamp 50 ␮g/kg 0.5 ␮g/kg/min 24 h
Jo43 20 20 Placebo After sternotomy No 0.5 ␮g/kg/min Until skin closure
Mollhoff44 11 11 Placebo After anesthesia induction 50 ␮g/kg 0.5 ␮g/kg/min Unspecified
Mollhoff45 15 14 Nifedipine After anesthesia induction No 0.375 ␮g/kg/min 24 h
Shi46 25 24 Placebo After anesthesia induction 50 ␮g/kg 0.5 ␮g/kg/min Until skin closure
Abbreviation: LCOS, low-cardiac-output syndrome.

Table 3. Preoperative Ejection Fraction and Postoperative Causes of Death in the 2 Groups
Preoperative Ejection Preoperative Ejection Causes of Death Causes of Death

First Author Fraction (Milrinone Group) Fraction (Control Group) Follow-up (Milrinone Group) (Control Group)
Al-Shawaf2 31 ⫾ 6 29 ⫾ 6 Hospital stay Severe congestive heart Sepsis-induced multiple

failure organ failure
Arbeus36 59 ⫾ 12 63 ⫾ 9 Hospital stay Not specified
Brackbill37 ⬍35% ⬍35% 30 days
Couture38 51 ⫾ 15 50 ⫾ 13 Hospital stay Multiple organ failure (2)
De Hert39 27 ⫾ 3 24 ⫾ 6 Hospital stay Multiple organ failure (2) and
cardiac failure
De Hert40 25 ⫾ 3 22 ⫾ 5 30 days Cardiac failure (4) Multiple organ failure
Doolan41 Not specified Not specified 72 hours
Hayashida42 40%-60% (1 patient); ⬎60% (12 patients) Hospital stay
⬎60% (11 patients)
Jebeli1 32 ⫾ 3 34 ⫾ 1 Hospital stay Cardiogenic shock (2)
Jo43 45 ⫾ 14 51 ⫾ 13 Hospital stay
Mollhoff44 Not specified Not specified 1 year
Mollhoff45 37 ⫾ 6 36 ⫾ 6 24 hours Myocardial infarction Ventricular arrhythmia
Shi46 Not specified Not specified 6 months Not specified Not specified
the control arm, OR ⫽ 4.19 [1.27-13.84], p for effect ⫽ 0.02, adult cardiac surgery patients. The central conclusion from the
p for heterogeneity ⫽ 0.40, I2 ⫽ 0% with 189 patients and 5 analysis is that the use of milrinone in these patients is asso-
studies included). A further subanalysis of studies that used ciated with a statistically significant increase in mortality. The
levosimedan as the control similarly indicated an increased effect was confirmed in a subgroup of studies comparing mil-
mortality with milrinone (8/49 deaths [16.3%] v 2/71 deaths rinone with other drugs but not in a subgroup of studies
[2.8%] with levosimendan, OR ⫽ 5.89 [1.56-21.72], p for comparing milrinone with placebo or nothing. Inotropic drugs
effect ⫽ 0.009, p for heterogeneity ⫽ 0.44, I2 ⫽ 0% with 120 repeatedly have been suggested to increase mortality,47-51 but
patients and 3 studies included). No significant differences this is the first time that such evidence has been brought
were found between milrinone and control in the other out- forward for a single drug in randomized trials in an acute
comes studied as shown in Table 5. setting.
No significant effect of milrinone on survival in comparison
DISCUSSION with placebo (or nothing) was shown, but there was a substan-
The authors performed a meta-analysis of all identified, tial detrimental effect of milrinone compared with active con-
published, randomized studies using intravenous milrinone in trols. Several aspects of these findings deserve comment. First,
Table 4. Risk of Bias Assessment of Included Studies
Incomplete Free of
Adequate Concurrent Outcome Selective Free of
Domain/ Sequence Allocation Concealment Therapies Data Outcome Other Overall Risk
Question Generation Used? Blinding? Similar? Addressed? Reporting? Bias? of Bias?
Al-Shawaf2 Unclear Yes (sealed envelopes) No Yes Unclear Yes Yes Moderate
Arbeus36 Unclear Yes (sealed envelopes Yes Yes Unclear Yes Yes Low
Brackbill37 Unclear No No Yes Yes Yes Yes Moderate
Couture38 Yes (computer Yes (sealed envelopes) Yes Yes Yes Yes Yes Low
generated)
De Hert39 Yes (computer Yes (sealed envelopes) Yes (adjudicators only) Yes Unclear Yes Yes Low
generated)
De Hert40 Yes (computer Yes (sealed envelopes) Yes (adjudicators only) Yes Unclear Yes Yes Low
generated)
Doolan41 Unclear Unclear Yes, but blinding was Unclear Yes Yes No High
broken in 10
controls
Hayashida42 Yes (computer Unclear No Unclear Unclear Yes Yes Moderate
generated)
Jebeli1 Unclear Unclear Yes Unclear Unclear Yes Yes Moderate
Jo43 Unclear Unclear Yes Yes Unclear No Yes Moderate
Mollhoff44 Unclear Unclear No Yes Unclear Yes No Moderate
Mollhoff45 Unclear Unclear Yes (patients, Unclear Unclear Yes Yes Moderate
physicians)
Shi46 Unclear Unclear No (adjudicators only) Unclear Yes Yes Yes Moderate
74 ZANGRILLO ET AL
Fig 2. A forest plot for the risk of mortality. CI, confidence interval; df, degrees of freedom.
the authors found that almost a decade after Thackray et al47 The number of deaths in the studies comparing milrinone
identified a dearth of large, well-configured trials of intrave- with active control was higher (9 v 3), and the overall compar-
nous inotropes, these deficits are still present. The present ison was unfavorable to milrinone (p ⫽ 0.02) (Fig 2). The event
study’s calculations on mortality in milrinone-treated cardiac rate in the milrinone group of this subanalysis was 10.7%,
surgery patients represent the most complete analysis of this almost 4-fold higher than that in the equivalent group of the
sort; nevertheless, they are based on a total of 19 deaths in 518 comparison milrinone versus placebo, making it appropriate to
patients. Particularly, when comparing milrinone with placebo, consider if this higher rate is real. According to the authors’
which resulted in 4 deaths versus 3, it has to be recognized that point of view, it probably is; 7 of the 9 deaths recorded in
the introduction of few additional events could transform (in milrinone-treated patients in this analysis occurred in the stud-
either direction) the point estimate and any conclusions drawn ies of De Hert et al,39,40 which were among the highest rated for
from it. In principle, the absence of a clear demonstration of an a lack of potential for bias (Table 4).
adverse effect on survival in that analysis might be used to The majority of the active-controlled trials of milrinone used
justify the administration of milrinone to secure effects on the calcium sensitizer levosimendan as the control. A recent
hemodynamics and similar targets. However, given the small meta-analysis of randomized controlled studies estimated that
number of events contributing to the survival analysis and in levosimendan was associated with a substantial reduction in
the absence of evidence of benefit on any of the other indices mortality risk (OR ⫽ 0.35 [0.18-0.71], p ⫽ 0.003) in cardiac
inspected (Table 5), the authors would be reluctant to support surgery.52 With respect to these findings, it is plausible that the
such an interpretation of these data. present study’s result of an apparently increased mortality risk
Fig 3. A funnel plot for the risk of mortality.

SE, standard error.
Table 5. A Summary of the Global Effect of Different Outcomes
Dichotomous Outcomes
Patients (Studies) Milrinone: Events Control: Events p for I2 (p for

Outcome Included (%) (%) OR 95% CI Effect heterogeneity)
Myocardial infarction 315 (8) 6/147 (0.6%) 9/148 (3.9%) 1.10 0.04–27.45 0.95 75% (0.02)
Arrhythmias 321 (9) 35/150 (23.3%) 28/171 (16.4%) 1.84 0.61–5.54 0.28 63% (0.02)
Acute renal failure 230 (5) 8/114 (7%) 6/116 (5.2%) 1.5 0.49–4.58 0.4 11% (0.3)
Continuous Outcomes
Patients (Studies) I2 (p for

Outcome Included WMD 95% CI p for Effect heterogeneity)
Troponin I (ng/mL) 149 (4) ⫺0.77 min ⫺3.72–⫺2.17 0.6 83% (⬍0.001)
Ventilation time (h) 264 (6) 2.39 h ⫺2.16–6.95 0.3 85% (⬍0.001)
Intensive care unit stay(h) 304 (7) 0.70 h 13.9–12.47 0.9 78% (⬍0.001)
Hospitalization (d) 209 (5) 0.78 d ⫺0.65–2.2 0.29 43% (0.13)
Abbreviations: CI, confidence interval; WMD, weighted mean difference.
with milrinone is in fact an artifact reflecting benefits from Limitations

levosimendan. Indeed, it seems likely to the present authors that
The authors acknowledge that only 4 of the 13 studies
this partially explains the finding. However, the discrepancy
included in the meta-analysis were of high quality. Further-
between event rates with milrinone in the 2 principal subsets of
more, the authors acknowledge that most of the evidence comes
this analysis together with the results of the present authors’
from 2 randomized studies from the same author that compared
analysis on milrinone versus levosimendan only (p ⫽ 0.009 in
milrinone with levosimendan.39,40
favor of levosimendan) suggest that the net difference between
milrinone and active controls is likely to be the product of both
a survival benefit with levosimendan and an increased mortality CONCLUSIONS
risk from milrinone. This meta-analysis suggests that milrinone is associated with a
In accordance with these conclusions, some differences be- significantly increased risk of dying in adult patients undergoing
tween the methodologies used in the included studies should be cardiac surgery when compared with other drugs. Caution and
noted. Six of the 8 studies that compared milrinone with placebo consideration of alternative inotropes are recommended before
(or nothing) used a bolus plus infusion protocol. By contrast, 3 of using milrinone in this setting. There was no difference in mor-
the 5 trials on milrinone versus active controls, including both of tality when milrinone was compared with placebo or nothing.
the studies by De Hert et al,39,40 administered milrinone by infu-
sion only. Similarly, most studies on milrinone versus placebo/
APPENDIX
nothing were undertaken in isolated CABG surgery, whereas the
studies of De Hert et al,39,40 Brackhill et al,37 and Doolan et al41 Search strategy for PubMed, developed according to Biondi-
were undertaken in CABG surgery with simultaneous valve sur- Zoccai et al8: (heart OR cardiac OR myocard* OR coronary)
gery. It is beyond the scope of the present analysis to determine if AND (operatin* OR operationⴱ OR surger*) AND (((phospho-
these differences might have contributed to the findings in sub- diesterase OR pde) AND inhibitor*) OR amrinone OR milri-
groups although it is worth noting that simultaneous valve proce- none OR enoximone OR toborinone) AND (randomized con-
dures are associated with an important increase in the risk of death trolled trial[pt] OR controlled clinical trial[pt] OR randomized
in patients with pre-CABG cardiogenic shock.53 The resolution of controlled trials[mh] OR random allocation[mh] OR double-
these matters would require 1 or more prospective trials, perhaps blind method[mh] OR single-blind method[mh] OR clinical
of factorial design and likely requiring multicenter participation to trial[pt] OR clinical trials[mh] OR (clinical trial[tw] OR
ensure an adequately large total patient cohort. Reflections on ((singl*[tw] OR doubl*[tw] OR trebl*[tw] OR tripl*[tw]) AND
dosage and the need for a precise classification of hemodynamic (mask*[tw] OR blind[tw])) OR (latin square[tw]) OR place-
severity also should be made,3 but it is beyond the power of a bos[mh] OR placebo*[tw] OR random*[tw] OR research
numeric analysis to resolve them in the current state of knowledge. design[mh:noexp] OR comparative study[tw] OR follow-up
The present data do not provide a basis for excluding the use of studies[mh] OR prospective studies[mh] OR cross-over
milrinone as a pharmacologic bridging to transplantation (al- studies[mh] OR control*[tw] OR prospectiv*[tw] OR
though other limits to such use may be adduced)54 or in palliative volunteer*[tw]) NOT (animal[mh] NOT human[mh]) NOT
or semipalliative situations in which iatrogenic effects on patient (comment[pt] OR editorial[pt] OR meta-analysis[pt] OR prac-
survival are not a central consideration. The present findings are in tice-guideline[pt] OR review[pt])).
accordance with those of a large randomized controlled study that
compared milrinone and placebo in patients with severe chronic ACKNOWLEDGMENT
heart failure and found a 28% increase in mortality in those The authors thank Rosalba Lembo, Teresa Greco, and Gaia Bertarelli
patients receiving milrinone, with the greatest effects in patients (all biostatisticians) for the careful revision of the manuscript and the
with the most severe symptoms, who had a 53% mortality.55 support in data analysis.
76 ZANGRILLO ET AL
REFERENCES
1. Jebeli M, Ghazinoor M, Mandegar MH, et al: Effect of milrinone 19. Lee JH, Oh YJ, Shim YH, et al: The effect of milrinone on the
on short-term outcome of patients with myocardial dysfunction under- right ventricular function in patients with reduced right ventricular
going coronary artery bypass graft: A randomized controlled trial. function undergoing off-pump coronary artery bypass graft surgery. J
Cardiol J 17:73-78, 2010 Korean Med Sci 21:854-858, 2006
2. Al-Shawaf E, Ayed A, Vislocky I, et al: Levosimendan or mil- 20. Lobato EB, Janelle GM, Urdaneta F, et al: Comparison of
rinone in the type 2 diabetic patient with low ejection fraction under- milrinone versus nitroglycerin, alone and in combination, on grafted
going elective coronary artery surgery. J Cardiothorac Vasc Anesth internal mammary artery flow after cardiopulmonary bypass: Effects of
20:353-357, 2006 alpha-adrenergic stimulation. J Cardiothorac Vasc Anesth 15:723-727,
3. Mebazaa A, Pitsis AA, Rudiger A, et al. Clinical review: Practical 2001
recommendations on the management of perioperative heart failure in 21. Lobato EB, Urdaneta F, Martin TD, et al: Effects of milrinone
cardiac surgery. Crit Care 14:201, 2010 versus epinephrine on grafted internal mammary artery flow after
4. Parissis JT, Farmakis D, Nieminen M: Classical inotropes and cardiopulmonary bypass. J Cardiothorac Vasc Anesth 14:9-11, 2000
new cardiac enhancers. Heart Fail Rev 12:149-156, 2007 22. Maslow AD, Regan MM, Schwartz C, et al: Inotropes improve
5. Kikura M, Sato S: Effects of preemptive therapy with milrinone right heart function in patients undergoing aortic valve replacement for
or amrinone on perioperative platelet function and haemostasis in aortic stenosis. Anesth Analg 98:891-902, 2004
patients undergoing coronary bypass grafting. Platelets 14:277-282, 23. McNicol L, Andersen LW, Liu G, et al: Markers of splanchnic
2003 perfusion and intestinal translocation of endotoxins during cardiopul-
6. Heringlake M, Wernerus M, Grunefeld J, et al: The metabolic and monary bypass: Effects of dopamine and milrinone. J Cardiothorac
renal effects of adrenaline and milrinone in patients with myocardial Vasc Anesth 13:292-298, 1999
dysfunction after coronary artery bypass grafting. Crit Care 11:R51, 24. Oztekin I, Yazici S, Oztekin DS, et al: Effects of low-dose
2007 milrinone on weaning from cardiopulmonary bypass and after in pa-
7. Movsesian M, Stehlik J, Vandeput F, et al: Phosphodiesterase tients with mitral stenosis and pulmonary hypertension. Yakugaku
inhibition in heart failure. Heart Fail Rev 14:255-263, 2009 Zasshi 127:375-383, 2007
25. Rathmell JP, Prielipp RC, Butterworth JF, et al: A multicenter,
8. Biondi-Zoccai GG, Agostoni P, Abbate A, et al: A simple hint to
randomized, blind comparison of amrinone with milrinone after elec-
improve Robinson and Dickersin’s highly sensitive PubMed search
tive cardiac surgery. Anesth Analg 86:683-690, 1998
strategy for controlled clinical trials. Int J Epidemiol 34:224-225, 2005
26. Yamada T, Takeda J, Katori N, et al: Hemodynamic effects of
9. Liberati A, Altman DG, Tetzlaff J, et al: The PRISMA statement
milrinone during weaning from cardiopulmonary bypass: Comparison
for reporting systematic reviews and meta-analyses of studies that
of patients with a low and high prebypass cardiac index. J Cardiothorac
evaluate healthcare interventions: Explanation and elaboration. BMJ
Vasc Anesth 14:367-373, 2000
339:b2700, 2009
27. Yamaura K, Okamoto H, Akiyoshi K, et al: Effect of low-dose
10. Axelsson B, Arbeus M, Magnuson A, et al: Milrinone improves
milrinone on gastric intramucosal pH and systemic inflammation after
diastolic function in coronary artery bypass surgery as assessed by
hypothermic cardiopulmonary bypass. J Cardiothorac Vasc Anesth
acoustic quantification and peak filling rate: A prospective randomized
15:197-203, 2001
study. J Cardiothorac Vasc Anesth 24:244-249, 2010
28. Zabeeda D, Medalion B, Jackobshvilli S, et al: Comparison of
11. Carmona MJ, Martins LM, Vane MF, et al: Comparison of the systemic vasodilators: Effects on flow in internal mammary and radial
effects of dobutamine and milrinone on hemodynamic parameters and arteries. Ann Thorac Surg 71:138-141, 2001
oxygen supply in patients undergoing cardiac surgery with low cardiac
29. Yamaguchi A, Tanaka M, Naito K, et al: The efficacy of
output after anesthetic induction. Rev Bras Anestesiol 60:237-246,
intravenous milrinone in left ventricular restoration. Ann Thorac Car-
2010
diovasc Surg 15:233-238, 2009
12. Feneck RO, Sherry KM, Withington PS, et al, European Milri- 30. Sha K, Shimokawa M, Ishimaru K, et al: Differences in hemo-
none Multicenter Trial Group: Comparison of the hemodynamic effects dynamic effects of amrinone, milrinone and olprinone after cardiopul-
of milrinone with dobutamine in patients after cardiac surgery. J Car- monary bypass in valvular cardiac surgery. Masui 49:981-986, 2000
diothorac Vasc Anesth 15:306-315, 2001 31. Shibata T, Suehiro S, Sasaki Y, et al: Slow induction of milri-
13. Iribe G, Yamada H, Matsunaga A, et al: Effects of the phos- none after coronary artery bypass grafting. Ann Thorac Cardiovasc
phodiesterase III inhibitors olprinone, milrinone, and amrinone on Surg 7:23-27, 2001
hepatosplanchnic oxygen metabolism. Crit Care Med 28:743-748, 2000 32. Wang H, Gong M, Zhou B, et al: Comparison of inhaled and
14. Kikura M, Lee MK, Safon RA, et al: The effects of milrinone on intravenous milrinone in patients with pulmonary hypertension under-
platelets in patients undergoing cardiac surgery. Anesth Analg 81:44- going mitral valve surgery. Adv Ther 26:462-468, 2009
48, 1995 33. Solina A, Papp D, Ginsberg S, et al: A comparison of inhaled
15. Kikura M, Levy JH, Michelsen LG, et al: The effect of milri- nitric oxide and milrinone for the treatment of pulmonary hypertension
none on hemodynamics and left ventricular function after emergence in adult cardiac surgery patients. J Cardiothorac Vasc Anesth 14:12-17,
from cardiopulmonary bypass. Anesth Analg 85:16-22, 1997 2000
16. Kikura M, Sato S: The efficacy of preemptive milrinone or 34. Momeni M, Rubay J, Matta A, et al: Levosimendan in congen-
amrinone therapy in patients undergoing coronary artery bypass graft- ital cardiac surgery: A randomized, double-blind clinical trial. J Car-
ing. Anesth Analg 94:22, 2002 diothorac Vasc Anesth 25:419-424, 2011
17. Kim JH, Ham BM, Kim YL, et al: Prophylactic milrinone during 35. Lobato EB, Gravenstein N, Martin TD: Milrinone, not epineph-
OPCAB of posterior vessels: Implication in angina patients taking rine, improves left ventricular compliance after cardiopulmonary by-
beta-blockers. Eur J Cardiothorac Surg 24:770-776, 2003 pass. J Cardiothorac Vasc Anesth 14:374-377, 2000
18. Kwak YL, Oh YJ, Kim SH, et al: Efficacy of pre-emptive 36. Arbeus M, Axelsson B, Friberg O, et al: Milrinone increases
milrinone in off-pump coronary artery bypass surgery: Comparison flow in coronary artery bypass grafts after cardiopulmonary bypass: A
between patients with a low and normal pre-graft cardiac index. Eur prospective, randomized, double-blind, placebo-controlled study.
J Cardiothorac Surg 26:687-693, 2004 J Cardiothorac Vasc Anesth 23:48-53, 2009
37. Brackbill ML, Stam MD, Schuller-Williams RV, et al: Periop- the adrenergic pathway in patients with heart failure-a meta-regression
erative nesiritide versus milrinone in high-risk coronary artery bypass analysis. Eur J Heart Fail 4:515-529, 2002
graft patients. Ann Pharmacother 41:427-432, 2007 48. Bayram M, De Luca L, Massie MB, et al: Reassessment of
38. Couture P, Denault AY, Pellerin M, et al: Milrinone enhances dobutamine, dopamine, and milrinone in the management of acute
systolic, but not diastolic function during coronary artery bypass graft- heart failure syndromes. Am J Cardiol 96:47G-58G, 2005
ing surgery. Can J Anaesth 54:509-522, 2007 49. Metra M, Eichhorn E, Abraham WT, et al: Effects of low-dose
39. De Hert SG, Lorsomradee S, Cromheecke S, et al: The effects of
oral enoximone administration on mortality, morbidity, and exercise
levosimendan in cardiac surgery patients with poor left ventricular
capacity in patients with advanced heart failure: The randomized,
function. Anesth Analg 104:766-773, 2007
double-blind, placebo-controlled, parallel group ESSENTIAL trials.
40. De Hert SG, Lorsomradee S, Vanden Eede H, et al: A random-
Eur J Heart 30:3015-3026, 2009
ized trial evaluating different modalities of levosimendan administra-
tion in cardiac surgery patients with myocardial dysfunction. J Cardio- 50. Abraham WT, Adams KF, Fonarow GC, et al: In-hospital mor-
thorac Vasc Anesth 22:699-705, 2008 tality in patients with acute decompensated heart failure requiring
41. Doolan LA, Jones EF, Kalman J, et al: A placebo-controlled trial intravenous vasoactive medications: An analysis from the acute de-
verifying the efficacy of milrinone in weaning high-risk patients from compensated heart failure national registry (ADHERE). J Am Coll
cardiopulmonary bypass. J Cardiothorac Vasc Anesth 11:37-41, 1997 Cardiol 46:57-64, 2005
42. Hayashida N, Tomoeda H, Oda T, et al: Inhibitory effect of 51. Mebazaa A, Parissis J, Porcher R, et al: Short-term survival by
milrinone on cytokine production after cardiopulmonary bypass. Ann treatment among patients hospitalized with acute heart failure: The
Thorac Surg 68:1661-1667, 1999 global ALARM-HF registry using propensity scoring methods. Inten-
43. Jo HR, Lee WK, Kim YH, et al: The effect of milrinone infusion sive Care Med 37:290-301, 2011
on right ventricular function during coronary anastomosis and early 52. Landoni G, Mizzi A, Biondi-Zoccai G, et al: Reducing mortality
outcomes in patients undergoing off-pump coronary artery bypass in cardiac surgery with levosimendan: A meta-analysis of randomized
surgery. Korean J Anesthesiol 59:92-98, 2010 controlled trials. J Cardiothorac Vasc Anesth 24:51-57, 2010
44. Mollhoff T, Loick HM, Van Aken H, et al: Milrinone modulates
53. Mehta RH, Grab JD, O’Brien SM, et al: Clinical characteristics
endotoxemia, systemic inflammation, and subsequent acute phase re-
and in-hospital outcomes of patients with cardiogenic shock undergo-
sponse after cardiopulmonary bypass (CPB). Anesthesiology 90:72-80,
ing coronary artery bypass surgery: Insights from the Society of Tho-
1999
racic Surgeons National Cardiac Database. Circulation 117:876-885,
45. Mollhoff T, Schmidt C, Van Aken H, et al: Myocardial isch-
2008
aemia in patients with impaired left ventricular function undergoing
coronary artery bypass grafting—Milrinone versus nifedipine. Eur J 54. Aranda JM Jr, Schofield RS, Pauly DF, et al: Comparison of
Anaesthesiol 19:796-802, 2002 dobutamine versus milrinone therapy in hospitalized patients awaiting
46. Shi Y, Denault AY, Couture P, et al: Biventricular diastolic cardiac transplantation: A prospective, randomized trial. Am Heart J
filling patterns after coronary artery bypass graft surgery. J Thorac 145:324-329, 2003
Cardiovasc Surg 131:1080-1086, 2006 55. Packer M, Carver JR, Rodeheffer RJ, et al: Effect of oral
47. Thackray S, Easthaugh J, Freemantle N, et al: The effectiveness milrinone on mortality in severe chronic heart failure. The PROMISE
and relative effectiveness of intravenous inotropic drugs acting through Study Research Group. N Engl J Med 325:1468-1475, 1991

ProgramadeGestionHCR 2013-2019

Transféré par

Informations du document

Titre original

Copyright

Formats disponibles

Partager ce document

Partager ou intégrer le document

Options de partage

Avez-vous trouvé ce document utile ?

Ce contenu est-il inapproprié ?

Droits d'auteur :

Formats disponibles

ProgramadeGestionHCR 2013-2019

Transféré par

Droits d'auteur :

Formats disponibles

Confeccionado por: Grenedys López Tápanes.

1. Almeida ASd, Picon PD, Wender OCB.

3. Badreldin AMA, Doerr F, Ismail MM, Heldwein MB, Lehmann T, Bayer O,

4. Barros de Oliveira Sá MP, Barros de Oliveira Sá MV, Lopes

5. Bhukal I, Solanki SL, Ramaswamy S, Yaddanapudi LN, Jain A, Kumar P.

6. Cantero MA, Almeida RMS, Galhardo R.

8. Di Leoni Ferrari A, Pelzer Süssenbach C, Vieira da Costa Guaragna JC,

11. Garcia-Hernandez JA, Benitez-Gomez IL, Martinez-Lopez AI, Praena-

12. Holm J, Hakanson E, Vanky F, Svedjeholm R.

14. Kristensen KL, Rauer LJ, Mortensen PE, Kjeldsen BJ.

16. Lopez-Delgado JC, Esteve F, Javierre C, Perez X, Torrado H, Carrio

18. Morlans Hernández K, Santos Gracia J, Cáceres Lóriga FM, Pérez

22. Rahmanian PB, Adams DH, Castillo JG, Carpentier A, Filsoufi F.

23. Ried M, Haneya A, Kolat P, Potzger T, Puehler T, Schmid C, et al.

24. Schloss B, Gulati P, Yu L, Abdel-Rasoul M, O'Brien W, Von Visger J, et

26. Shahin J, DeVarennes B, Tse CW, Amarica DA, Dial S.

29. Zangrillo A, Biondi-Zoccai G, Ponschab M, Greco M, Corno L, Covello

Nombre de la biblioteca virtual: mortalidad intrahospitalaria cir. cardiovasc.enl

Las palabras claves por las que hicimos la búsqueda: octogenarians,

Fuente Núcleo: En este paquete informativo sobre la Mortalidad

Dr. Giovanni Landoni,

Dr. Alexander P J. Vlaar,

Dr. Luc Noyez

Dr. Juan Carlos López Delgado,

Dr. Katrine Lawaetz Kristensen,

Dr. J.A. García-Hernández

Dr. Antônio Sérgio Cordeiro da Rocha

Marcos Antonio Cantero. Rua Delmar de Oliveira, 1725

Dr. Michel Pompeu Barros de Oliveira Sá

Outcomes of patients subjected to aortic valve

DOI: 10.5935/1678-9741.20110006 RBCCV 44205-1287

Abstract (P=0.057). Factors associated with reoperation were: renal

Resumo reoperação foram: insuficiência renal, endocardite de prótese

INTRODUCTION influence of comorbidities on the outcome in a period of 20

Table 1. Characteristics of the sample

Table 2. Hospital outcomes

(CI 95% = 62.20% -78.60%) and 58.40% (CI 95% = 48.40% -

Table 3. Independent predictors of death by Cox regression

Age> 70 years 2.48 (1.51-4.08) <0.001

Table 4. Incidence of deaths by period

Table 5. Causes of death

Table 6. Outcomes in the cohort during the follow-up period

(95% CI = 86, 80% -96.60%) and 91.70% (CI 95% = 86.80% -

Fig. 3 - Kaplan-Meier time to assess event-free bleeding Mortality

events in patients with mechanical valve substitutes in Limitations of the Study

Predictors of mortality in Patients over 70 years-

Abstract death rates: mechanical Ventilation ³ 12 hours (P <0.001),

INTRODUCTION individuals to offer an optimal treatment, which determines

Table 1. Preoperative variables. N = 265.

The preoperative variables analyzed did not increase

The ischemia in survivors was 55.5 ± 20 minutes and in the

11. Almeida AS, Manfroi WC. Peculiaridades no tratamento da

12. Takiuti ME, Hueb W, Hiscock SB, Nogueira CR, Girardi P,

13. Edwards FH, Clark RE, Schwartz M. Coronary artery bypass

Comparison between Sequential Organ Failure

Downloaded by: World Health Organization ( WHO). Copyrighted material.

Introduction The “Sequential Organ Failure Assessment” score (SOFA)

received Copyright © 2012 by Thieme Medical DOI http://dx.doi.org/

Table 1 The Sequential Organ Failure Assessment (SOFA) scoring system.

Organ System Descriptor Score points

Downloaded by: World Health Organization ( WHO). Copyrighted material.

Table 2 The Cardiac Surgery Score (CASUS).

Organ System Descriptor Score points