Académique Documents
Professionnel Documents
Culture Documents
OBJECTIVE: To describe the implementation of an emergency depart- of Medicine, Houston, Texas; and dEmergency Department, Texas
Children’s Hospital, Houston, Texas
ment (ED) protocol for the recognition of septic shock and facilitate
adherence to national treatment guidelines. KEY WORDS
goal-directed therapy, pediatric, sepsis
PATIENTS AND METHODS: Root-cause analyses and morbidity and ABBREVIATIONS
mortality conferences identified system problems with sepsis recog- ED—emergency department
nition and management. A group of ED and critical care physicians met QI—quality improvement
to identify barriers and create solutions. Drs Cruz and Perry contributed equally to this work.
e758 CRUZ et al
Downloaded from pediatrics.aappublications.org by guest on September 12, 2015
QUALITY-IMPROVEMENT REPORTS
Between 20 000 and 40 000 US children prioritization; and barriers to patient secutive patients from 2009 to the
develop septic shock annually1,2; flow through the institution. study period (February to August
chronically ill children are overrepre- Subsequently, a prospective QI project 2010). Control limits define 3 SDs
sented in incidence and mortality was designed to measure the impact around the mean; outliers signify
rates.2,3 Despite evidence-based guide- of early recognition and intensive causes not inherent to the process.
lines,4 a minority of children receive nursing resources on the ability to de- Data points within the control limits
the standard of care.5 Although delays liver fluids and antibiotics more rap- signify variability in the process that
in fluid resuscitation have been idly to children in shock; the project requires system changes. Retrospec-
associated with increased mortality was termed the “shock protocol.” This tive anecdotal discussions helped us
rates,5,6 many barriers to timely resus- project was conducted in the Texas to document institutional protocol-
citation exist in busy emergency de- Children’s Hospital ED, where ⬎200 related cultural changes. Institutional
partments (EDs).7,8 practitioners and 170 nurses care for review board approval was obtained.
We hypothesized that process barriers ⬃85 000 children annually. All children
for whom the shock protocol was im- RESULTS
resulted in delays in shock recognition
and management in our ED and that plemented were included. Children Our first priority was improving shock
system changes would result in im- were clinically diagnosed as septic; recognition, specifically easily identify-
proved outcomes. We designed a case definitions9 were not used as in- ing patients with abnormal vital signs.
quality-improvement (QI) intervention clusion criteria. Patients included in We needed to create a system that
the protocol were identified from a would minimize variation in ED pro-
to maximize the recognition of patients
shock protocol order set. Patients in vider experience and the fluctuations
at risk for septic shock and facilitate
shock who did not receive the protocol in ED patient arrivals that contributed
the ED implementation of preexisting
(missed patients) were identified from to delayed recognition of abnormal
national guidelines in 3 ways. First, we
several sources: records of triage vital signs. Information technology
created an automated triage tool to
tool alerts; admission/discharge diag- helped create a computerized triage
recognize vital-sign abnormalities.
noses of PICU admissions; and inpa- tool that corrected heart rate for pyr-
Second, we harnessed resources for
tient rapid-response team calls ⬍24 exia.10 If vital signs were outside of age-
the more intensive nursing care
hours after admission. For patients in appropriate norms,4,9,11 an electronic
needed to implement national guide-
shock for whom the triage tool was alert forced the triage nurse to con-
lines. Third, we designed a physiologic
triggered but the protocol was not sider the shock protocol. If the patient
flow sheet to characterize temporal was at high risk (Table 1) or appeared
used, chart review and discussion with
changes in vital signs that could assist the individual clinician(s) were per- ill, the triage nurse was empowered to
in patient handoffs and maintain formed to determine obstacles to pro- call the charge nurse and activate the
awareness of the guideline. tocol use. protocol. With activation, the transport
Resuscitation timeliness was mea- team and PICU charge nurse were also
METHODS alerted of a potential admission. The
sured by documentation of time from
Root-cause analyses and morbidity triage to initiation of first fluid and sub- patient was immediately taken to a
and mortality conferences revealed sequent boluses, bolus volume/dura- designated room, and an attending
areas for improvements in sepsis tion, and timing of vasoactive agent physician was called to evaluate the
management. With support from hos- use. Antibiotic use and timing were patient and initiate treatment appro-
pital leadership, a multidisciplinary measured. Both included and missed priately. Although intended to be acti-
team (ED and ICU physicians and patients were evaluated for ED and vated from triage, the shock protocol
nurses and ancillary services) identi- hospital length of stay, and rapid- could be initiated by nurses or physi-
fied several obstacles including varia- response teams called within 24 hours cians for any patient at any point in the
tion in experience of staff in perform- of admission to a non-PICU bed. These ED stay, and children could be taken off
ing initial evaluations; lack of adequate admissions were compared with PICU the protocol by an attending physician
nursing staff for resource-intensive admissions for sepsis in 2009. Data at any time.
patients; difficulty obtaining frequent were analyzed by using MiniTab (State In addition, nurses on the QI team
vital-sign measurements; lack of stan- College, PA). Statistical process con- pointed out that the existing nurse/pa-
dardization of empiric antibiotics and trol charts were used to compare out- tient ratio was impractical given the
diagnostic tests; lack of medication come measures in time order in con- urgency and resource intensity of
e760 CRUZ et al
Downloaded from pediatrics.aappublications.org by guest on September 12, 2015
QUALITY-IMPROVEMENT REPORTS
FIGURE 1
Shock flow sheet.
patients for whom the protocol was developed with information technol- cates. Within 1 month, nurses were
not used. Of these children, 12% had ogy as electronic medical records asking if fluid boluses for children who
high-risk conditions; however, none were introduced. Second, project suc- were not in shock could be given more
was tachycardic at triage. The nonen- cess led to the protocol being priori- rapidly. The nurses were instrumental
rolled patients were more acutely ill tized for integration into electronic in redesigning order sets and the flow
and required immediate interventions algorithms. Third, the protocol was sheet. There was concern that there
and initial airway management (Table integrated into hospital evidence- would be territorial issues or blurring
4). Statistical process control charts based guidelines for shock. Fourth, of responsibilities between the ED
that presented time to interventions in the decision of where to admit poten- nurses and transport team. The ED
chronological order showed that chil- tially septic children was critically nurses appreciated the skills and
dren on the shock protocol received evaluated. workforce provided by the transport
interventions more rapidly and with Changes also occurred at the nursing team.
less variation than the patients with level. The most significant change was Before project roll-out, we recognized
sepsis in 2009 (Fig 2). There were no that nurses were empowered to initi- that frequent measurement of vital
rapid-response teams called for pa- ate the protocol from triage, and the signs was of minimal utility without
tients admitted to non-PICU settings af- team encouraged physicians to com- creating a documentation system that
ter the shock protocol. municate with nurses when the deci- was easy to complete and read. Multi-
Several changes occurred at a sys- sion was made to stop the protocol. tasking ED physicians were given in-
tems level. First, the triage tool was The nurses became protocol advo- stantly readable graphical vital-sign
e762 CRUZ et al
Downloaded from pediatrics.aappublications.org by guest on September 12, 2015
QUALITY-IMPROVEMENT REPORTS
TABLE 3 Demographic Features of the Study Population The main changes for physicians were
Characteristic Enrolled at Enrolled After Not Used No. of Patient the ability to intervene earlier and har-
Triage Triage Encounters ness resources for very ill children. ED
No. of patients 139 28 25 192 physicians and nurses immediately
No. (%) of patient encountersa 158 (73) 33 (15) 25 (12) 216
Age group, %b,c recognized that this ability would be
0–23 mo 70 12 19 43 beneficial even for patients who were
2–5 y 71 20 9 55 not in shock but needed more intense
6–12 y 78 15 6 65
⬎12 y 72 13 15 53 evaluation and treatment. In addition,
Triage acuity, median, 5-point scaled 2 3 2 — earlier communication with the PICU
Male gender, %b,c 72 14 14 115 led to shorter ED stays. Physicians who
Medical comorbidities, %b,c
Hematologic malignancy 91 7 1 70
treated children off-protocol were ap-
Nonhematologic malignancy 85 7 7 27 proached by the team to explore the
Bone marrow transplantation 94 6 0 16 reasons why; data were shared with
Solid-organ transplantation 90 10 0 21
Asplenia 100 0 0 6
the group on outcomes of children
Short gut 100 0 0 4 treated on and off the protocol. Like-
Other 76 24 0 21 wise, subspecialty services shared
None 27 29 43 51
Indwelling vascular catheter, %b,c 94 4 3 109
concerns or suggestions. The protocol
Presence of neutropenia (absolute neutrophil 76 16 8 49 did not decrease physician autonomy
count ⬍ 500), %b,d but did standardize and facilitate care.
Disease location, %b,d,e
There was collaboration between ED
Pneumonia 54 17 29 35
Skin/soft-tissue infection 33 22 44 9 and PICU physicians with a similar
Bacteremia 65 17 17 23 goal, which was attained by education
Fungemia 100 0 0 2 and reeducation of caregivers across
Urinary tract infection 55 27 18 11
Meningitis 33 0 67 3 the hospital by the team.
Otherf 81 14 5 139 Pharmacy culture changed by priori-
Etiologies, %b,d,e
Streptococcus pneumoniae 25 0 75 4 tizing and streamlining antibiotics. On
Staphylococcus aureus 43 14 43 7 the basis of the medications pre-
Other Gram-positives 100 0 0 9 scribed that were not on the initial or-
Escherichia coli 33 33 33 12
Other Gram-negatives 50 29 21 14
der sheet, preprinted orders were re-
Parainfluenza viruses 60 30 10 10 vised. A bundled laboratory package
H1N1 influenza 83 17 0 6 was created by the team in conjunc-
Other viruses 50 25 25 12
Candida species 100 0 0 3
tion with the laboratory to decrease
Anaerobes 0 50 50 2 variation in diagnostic evaluation.
No organism identified 82 12 6 142
Disposition from ED, %b,d DISCUSSION
ICU 54 23 24 106
Intermediate care floor 80 20 0 10 The Surviving Sepsis Campaign12 has
Inpatient floor 93 7 0 87 made great inroads in delivery of
Discharge home 92 8 0 12
Died in ED 100 0 0 1 timely care for septic adults, and
a Eighteen patients had 2 encounters; 3 patients had 3 encounters. “shock” teams exist in many adult EDs.
b Percentages within rows may not sum to 100% as a result of rounding.
c Numbers given for unique patients.
These teams have led to decreased
d Numbers given for individual patient encounters. mortality rates,13 a decreased need for
e Seven children had ⬎1 site of disease, and 11 had ⬎1 infectious agent identified.
invasive monitoring in the ICU,14 and an
f “Other” includes no identifiable disease location and viral respiratory infections.
increased proportion of patients who
receive goal-directed therapy per na-
trends with specific times and quanti- created. It should be noted that al- tional guidelines.15,16 The few published
ties of fluid and medications given; though the triage alert did require an pediatric series have focused on the
thus, flow-sheet results could help electronic medical record (EMR), use logistic difficulties with meeting Pedi-
guide therapy. A graphical, imminently of the flow sheet, although it could be atric Advanced Life Support/American
useful handoff tool for the receiving incorporated into an EMR, was not con- College of Critical Care Medicine
PICU or other admitting physician was tingent on it. (PALS/ACCCM) guidelines. These barri-
e764 CRUZ et al
Downloaded from pediatrics.aappublications.org by guest on September 12, 2015
QUALITY-IMPROVEMENT REPORTS
REFERENCES
1. Watson RS, Carcillo JA, Linde-Zwirble WT, inquiry. Pediatr Crit Care Med. 2010;11(4): 1-year experience with implementing early
Clermont G, Lidicker J, Angus DC. The epide- 469 – 474 goal-directed therapy for septic shock in
miology of severe sepsis in children in the 8. Inwald DP, Peters MJ, Nadel S; Paediatric the emergency department. Chest. 2006;
United States. Am J Respir Crit Care Med. Intensive Care Society Study Group. Emer- 129(2):225–232
2003;167(5):695–701 gency management of children with severe 15. Crowe CA, Mistry CD, Rzechula K, Kulstand
2. Odetola FO, Gebremariam A, Freed GL. Pa- sepsis in the United Kingdom: the results of CE. Evaluation of a modified early goal-
tient and hospital correlates of clinical sep- the Paediatric Intensive Care Society sepsis directed therapy protocol. Am J Emerg Med.
sis and resource utilization in severe pedi- audit. Arch Dis Child. 2009;94(5):348 –353 2010;28(6):689 – 693
atric sepsis. Pediatrics. 2007;119(3): 9. Goldstein B, Giroir B, Randolph A. Interna- 16. Jones AE, Focht A, Horton JM, Kline JA. Pro-
487– 494 tional pediatric sepsis consensus confer- spective external validation of the clinical
3. Fiser RT, West NK, Bush AJ, Sillos EM, ence: definitions for sepsis and organ sys-
effectiveness of an emergency department-
tem dysfunction in pediatrics. Pediatr Crit
Schmidt JE, Tamburro RF. Outcome of se- based early goal-directed therapy protocol
Care Med. 2005;6(1):2– 8
vere sepsis in pediatric oncology patients. for severe sepsis and septic shock. Chest.
Pediatr Crit Care Med. 2005;6(5):531–536 10. Hanna CM, Greenes DS. How much tachycar- 2007;132(2):425– 432
dia in infants can be attributed to fever? Ann
4. Brierley J, Carcillo JA, Choong K, et al. Clin- 17. Oliveira CF, Nogueira de Sá FR, Oliveira DSF,
Emerg Med. 2004;43(6):699 –705
ical practice parameters for hemodynamic et al. Time- and fluid-sensitive resuscitation
support of pediatric and neonatal septic 11. Custer JW, Rau RE. The Harriet Lane Hand-
for hemodynamic support of children in
book. 18th ed. Philadelphia, PA: Elsevier
shock: 2007 update from the American Col- septic shock: barriers to the implementa-
Mosby; 2009
lege of Critical Care Medicine [published tion of the American College of Critical Care
correction appears in Crit Care Med. 2009; 12. Dellinger RP, Levy MM, Carlet JM, et al. Sur-
Medicine/Pediatric Advanced Life Support
37(4):1536]. Crit Care Med. 2009;37(2): viving sepsis campaign: international
Guidelines in a pediatric intensive care unit
666 – 688 guidelines for the management of severe
in the developing world. Pediatr Emerg
sepsis and septic shock: 2008 [published
5. Han YY, Carcillo JA, Dragotta MA, et al. Early Care. 2008;24(12):810 – 815
correction appears in Crit Care Med. 2008;
reversal of pediatric-neonatal septic shock 36(4):1394 –1396]. Crit Care Med. 2008; 18. Institute of Medicine, Committee on Health
by community physicians is associated with 36(1):296 –327 Care Quality in America. Crossing the Qual-
improved outcome. Pediatrics. 2003;112(4): ity Chasm: A New Health System for the 21st
13. Puskarich MA, Marchick MR, Kline JA,
793–799 Century. Washington, DC: National Academy
Steuerwald MT, Jones AE. One year mortal-
6. Carcillo JA, Davis AL, Zaritsky A. Role of early ity of patients treated with an emergency Press; 2001. Available at: www.nap.edu/
fluid resuscitation in pediatric septic department based early goal-directed ther- html/quality_chasm/reportbrief.pdf. Ac-
shock. JAMA. 1991;266(9):1242–1245 apy protocol for severe sepsis and septic cessed July 21, 2010
7. Launay E, Gras-Le Guen C, Martinot A, et al. shock: a before and after study. Crit Care. 19. Brierley J, Peters MJ. Distinct hemodynamic
Suboptimal care in the initial management 2009;13(5):R167 patterns of septic shock at presentation to
of children who died from severe bacterial 14. Trzeciak S, Dellinger RP, Abate NL, et al. pediatric intensive care. Pediatrics. 2008;
infection: a population-based confidential Translating research to clinical practice: a 122(4):752–759
e766 CRUZ et al
Downloaded from pediatrics.aappublications.org by guest on September 12, 2015
Implementation of Goal-Directed Therapy for Children With Suspected Sepsis in
the Emergency Department
Andrea T. Cruz, Andrew M. Perry, Eric A. Williams, Jeanine M. Graf, Elizabeth R.
Wuestner and Binita Patel
Pediatrics 2011;127;e758; originally published online February 21, 2011;
DOI: 10.1542/peds.2010-2895
Updated Information & including high resolution figures, can be found at:
Services http://pediatrics.aappublications.org/content/127/3/e758.full.h
tml
The online version of this article, along with updated information and services, is
located on the World Wide Web at:
http://pediatrics.aappublications.org/content/127/3/e758.full.html