J Thromb Thrombolysis
DOI 10.1007/s11239-016-1417-5
Evaluation of anticoagulation selection for acute venous
thromboembolism
HishamBadreldin1 HunterNichols1 JessicaRimsans1 DanielleCarter1
Springer Science+Business Media New York 2016
Abstract Treatment of venous thromboembolism (VTE)
has been confined to parenteral agents and oral vitamin K
antagonists for decades; however, with the approval of the
direct oral anticoagulants (DOACs), clinicians now have
more options. This study aims to evaluate the real world
prescribing practices of all oral anticoagulants for VTE
at a single center. A retrospective cohort analysis of all
adult patients diagnosed with acute onset VTE was con-
ducted. Of the 105 patients included in the analysis, 45
(43%) patients received warfarin and 60 (57%) patients
received a DOAC. Rivaroxaban and apixaban were the
most common DOACs initiated. There were significantly
more patients in the warfarin group with an eCrCl of
<60ml/min compared to patients who received a DOAC
(77.8% vs. 15%; P<0.05). There were significantly less in the management of VTE. Oral vitamin K antagonists
patients in the warfarin group with serum aminotranferase
concentrations three times the upper limit of normal com-
pared to those who received a DOAC (15.6% vs. 55%;
P<0.05). Patients who received a DOAC had less days
on parenteral anticoagulation compared to patients who
received warfarin (median 2.5days [IQR 04] vs. 6days sive drug and food interactions [57].
[IQR 57], p<0.05). Patients who received a DOAC had
a shorter hospital length of stay compared to patients who
received warfarin (median 3 days [IQR 24] vs. 8 days
[IQR 610], p<0.05). This analysis showed that DOACs tion effect by directly inhibiting thrombin formation. Riva-
are being prescribed more than warfarin for treatment of
new onset VTE. Renal and liver function may influence
the agent prescribed. Utilization of DOACs may decrease
the hospital length of stay.
Hisham Badreldin
hbadreldin@partners.org
1
Department of Pharmacy Services, Brigham and Womens
Hospital, 75 Francis Street, Boston, MA 02120, USA
Keywords Venous thromboembolism
Direct oral anticoagulants (DOACs) Warfarin
Pulmonary embolism Deep vein thrombosis
Introduction
In North America, Venous thromboembolism (VTE), is
the third most common cardiovascular disorder after coro-
nary artery disease and stroke [1]. VTE affects more than
500,000 people in the general population annually. Addi-
tionally, it has a substantial impact on morbidity and mortal-
ity in both the general population and hospitalized patients
[24]. Anticoagulants are considered the mainstay treatment
(VKA), like warfarin, have been the only oral anticoagulant
option for clinicians to manage VTE for decades. Warfarin
has several limitations, including interindividual variability
in response, slow onset to reach its full effect, narrow thera-
peutic index, frequent coagulation monitoring, and exten-
Recently, four direct oral anticoagulants (DOACs), dabi-
gatran, rivaroxaban, apixaban and edoxaban, were approved
for the treatment of VTE. Dabigatran exerts its anticoagula-
roxaban, apixaban, and edoxaban work by inhibiting factor
Xa. DOACs have several advantages over VKA, including
less variability in response, rapid onset of action, predict-
able pharmacokinetic profiles, decreased coagulation moni-
toring, and less interactions [810]. Multiple trials have
been published and demonstrated DOACs to be non inferior
to warfarin for the treatment of VTE with either similar or
less major bleeding events [1115]. As a result, the most
current guidelines for Antithrombotic Therapy for VTE Dis-
ease recommends using a DOAC over warfarin in patients
13
2 H. Badreldin et al.
with deep vein thrombosis (DVT) or pulmonary embolism
(PE) without cancer [16].
The purpose of this study is to describe the real-world
anticoagulation selection and to evaluate the prescribing
practices of all oral anticoagulants used to treat new onset
VTE at a large, tertiary, academic medical center.
Materials and methods
Study design
Investigators conducted a single-center, retrospective cohort
analysis of all adult patients admitted to Brigham and Wom-
ens Hospital from August 2015 through February 2016 with
an objectively confirmed diagnosis of acute onset DVT, PE,
or both. An electronic hospital database using ICD-9 and
ICD-10 codes was used to identify patients. Chart review
was conducted to collect patient demographics and agent
prescribed. Patients were followed until discharge and eval-
uated for study endpoints. Institutional review board (IRB)
approval was obtained before the initiation of this study.
Patient population
For the purpose of this evaluation, patients 18years or older
were included if they presented to the emergency depart-
ment or were admitted to an inpatient service with an
objectively confirmed diagnosis of new onset DVT, PE, or
both, and initiated on warfarin or a DOAC. Patients were
excluded if they had a contraindication to oral anticoagula-
tion, received enoxaparin monotherapy for long term man-
agement, received anticoagulation for another indication,
received systemic fibrinolysis, or were enrolled in any clini-
cal trial that might impact the oral anticoagulant selection.
Endpoints
The major endpoints of this study were to identify the pro-
portion of patients who were initiated on a DOAC versus
warfarin and the most common DOAC initiated. Patients
demographics, major risk factors for developing VTE,
and the type and location of DVT and/or PE were evalu-
ated to examine differences between the prescribing prac-
tices of warfarin versus DOACs. Other endpoints examined
included average days on parenteral anticoagulation and
average hospital length of stay (LOS).
Statistical analyses
Continuous variables were presented as means with stan-
dard deviations or medians with interquartile ranges (IQR).
Dichotomous variables were presented as numbers and
13
proportions. Endpoints were analyzed using chi-squared test
to compare categorical data and Students t-test or Mann
Whitney U test to compare continuous data. A P value of
less than 0.05 was considered to be statistically significant.
Results
Patients and follow-up
During the study period, 246 patients were identified that
met the inclusion criteria. A total of 141 patients were
excluded; 79 patients received enoxaparin monotherapy,
45 patients received anticoagulation for another indica-
tion, 14 patients received systemic fibrinolysis and three
patients were enrolled in a clinical trial. One hundred and
five patients were included in the final analysis.
Oral anticoagulant initiated
Of the 105 patients included, more patients were initiated
on a DOAC versus warfarin, 60 (57%) versus 45 (43%),
respectively. Of those initiated on a DOAC, rivaroxa-
ban was the most common agent prescribed. A total of 32
(53%) patients received rivaroxaban, followed by 24 (40%)
patients received apixaban, 3 (5%) patients received edoxa-
ban, and 1 (2%) patient received dabigatran (Fig.1).
Patient demographics
K ey baseline characteristics of the patients who received
warfarin or a DOAC are shown in Table1. Overall, there
were no statistical differences in age, sex, body mass index,
average hemoglobin/hematocrit, VTE index events, or doc-
umented risk factors for the development of VTE between
2%
n=1
5%
n=3
Rivaroxaban
Apixaban
40% 53%
Edoxaban
n = 24 n = 32
Dabigatran
Fig. 1 Selection of direct oral anticoagulant
Evaluation of anticoagulation selection for acute venous thromboembolism 3

Table 1 Patient baseline demographics those receiving warfarin compared to 33 (55%) receiving a
Warfarin DOAC P-value DOAC (P<0.05).
(N=45) (N=60)
Initiating services
Patient demographics
Age, yearsa 61.6 (15.2) 62.3 (18.1) 0.84
Primary initiating clinical services for oral anticoagulants
Femaleb 22.0 (48.9) 35.0 (58.3) 0.34
were cardiology, general medicine, intensive care, and
BMI, kg/m2a 30.5 (8.9) 28.1 (7.2) 0.29 emergency medicine services (Table2). Clinicians in both
eCrCl<60 mL/minbc 35.0 (77.8) 9.0 (15.0) <0.05 cardiology and general medicine services had no prefer-
Serum aminotranferase 7.0 (15.6) 33.0 (55.0) <0.05 ence in terms of initiating one agent over the other. Patients
concentrations >3 ULNb
who were admitted to the intensive care services were ini-
Hemoglobin, g/dla 10.4 (2.4) 11.1 (2.1) 0.12
tiated on warfarin more commonly than DOACs with 6
Hematocrit, %a 33.9 (2.4) 34.1 (6.0) 0.09
(13.3%) patients compared to 1 (1.7%) patient respectively,
Received ultrasound- 2.0 (4.4) 1.0 (1.7) 0.40
(P=0.018). Patients who presented to the emergency depart-
enhanced thrombolysisb
ment were initiated on DOACs more commonly than warfa-
Index event
rin, with 13 (21.7%) patients initiated on a DOAC compared
DVTb 25 (55.6) 35 (58.3) 0.78
b to 2 (4.5%) patients initiated on warfarin (P=0.012).
PE 10 (22.2) 10 (16.7) 0.47
DVT and PEb 10 (22.2) 15 (25.0) 0.74
Days on parenteral anticoagulation
Type of DVT
Femoralb 16 (45.7) 19 (38.0) 0.48
Patients who received a DOAC had significantly less days
Poplitealb 8 (22.9) 20 (40.0) 0.10
on parenteral anticoagulation compared to patients who
Illiacb 5 (14.3) 3 (6.0) 0.20
received warfarin (median 2.5days [IQR 04] vs. 6days
Saphenousb 5 (14.3) 6 (12.0) 0.76
[IQR 57], P<0.05) (Table2) Of the 60 patients initiated on
Otherbd 7 (20) 6 (12) 0.31
a DOAC, 35 (58.3%) patients were initiated on parenteral
Type of PE
anticoagulation prior to DOAC initiation and 25 (41.7%)
Unilateralb 9 (45.0) 12 (48.0) 0.84 patients were initiated on a DOAC without parenteral anti-
b
Bilateral 11 (55.0) 13 (52.0) 0.84 coagulation (Table3).
Massiveb 8 (40.0) 5 (20.0) 0.14
Submassiveb 12 (60.0) 20 (80.0) 0.14 Hospital length of stay
VTE Risk Factor, n (%)
Recent surgery 15 (33.3) 28 (46.7) 0.17 Patients in the DOAC group had a statistically significant
Previous VTE episode 19 (42.2) 15 (25.0) 0.06 shorter hospital length of stay compared to patients in the
Trauma 7 (15.6) 12 (20.0) 0.55 warfarin group (median 3days, [IQR 14] vs. 8days [IQR
Thromboembolic conditione 7 (15.6) 4 (6.7) 0.14 610], P<0.05) (Table 3).
BMI body mass index, DOAC direct oral anticoagulant, eCrCl esti-
mated creatinine clearance, kg kilogram, m meter, ml milliliter, min
minutes, SD standard deviation, ULN upper limit of normal, VTE Discussion
venous thromboembolism
a
Data presented as meanSD
b
This study was conducted to describe the real-world pre-
Data presented as n, (%)
c
scribing patterns of initial and long term oral anticoagulant
Creatinine clearance was calculated with the use of the Cockcroft
Gault formula
use in the setting of acute onset VTE at a large, tertiary,
d
Other includes: tibial, mesenteric, and subclavian veins
e
Includes antiphospholipid syndrome, protein C and S deficiency and
Table 2 Oral anticoagulant initiating service
factor V Leiden gene mutation
Initiating service, n (%) Warfarin DOAC P-value
patients who received warfarin or a DOAC. There was a (N=45) (N=60)
significant difference in the number of patients with an
Cardiology 27 (60.0) 35 (58.3) 0.87
estimated creatinine clearance (eCrCl) <60mL/min with
General medicine 10 (22.2) 11 (18.3) 0.62
35 (77.8%) patients receiving warfarin compared to 9
Intensive care 6 (13.3) 1 (1.7) 0.01
(15%) patients receiving a DOAC (P<0.05). Proportion of
Emergency care 2 (4.5) 13 (21.7) 0.01
patients with serum aminotranferase concentrations greater
than three times the upper limit of normal was 7 (15.6%) in DOAC direct oral anticoagulant

13
4 H. Badreldin et al.
Table 3 Days on parenteral anticoagulation and hospital length of stay
Warfarin DOAC P-value
(N=45) (N=60) were initiated on LMWH for the initial and long term man-
Duration on parenteral anticoagula- 6 (57) 2.5 (04) <0.05
tion, median days (IQR)
Median hospital length of stay IQR, 8 (610) 3 (34) <0.05
median days (IQR)
DOAC direct oral anticoagulant, IQR interquartile range
academic medical center. After almost five years since the
approval of the first DOAC, utilization of a DOAC repre-
sented about 57% of oral anticoagulant of choice to treat
new onset VTE at the Brigham and Womens hospital. We
believe the increase in DOAC utilization is likely due to the
advantages these agents offer over warfarin. Some advan-
tages includes: no routine monitoring, predictable pharma-
cokinetic and pharmacodynamic profiles, a rapid onset and
offset of action, fewer drug and food interactions and less
laboratory monitoring [810]. Overall, we observed renal
and hepatic function may be major factors considered in
deciding which agent to prescribe. Compared to patients
who received warfarin, patients receiving a DOAC had bet-
ter renal function and poorer hepatic function profiles at
baseline.
Our study also observed the vast majority of patients ini-
tiated on a DOAC were initiated on either rivaroxaban or
apixaban. One of the likely contributing factors for prefer-
ence of these two agents is the lack of requiring intravenous
anticoagulant bridging. Clinicians have the option of admin-
istering rivaroxaban 15mg twice daily for 3weeks followed
by 20mg daily, and initiating apixaban at10mg twice daily
for 7days followed by 5mg twice daily [11, 12]. In our
analysis, 50% of patients who were initiated on rivaroxaban
had received at least one dose of LMWH or UFH infusion
for at least one day before transitioning to rivaroxaban. For
patients initiated on apixaban, 62.5% of patients received at
least one dose of LMWH or UFH infusion for at least 1day
before starting apixaban.
Due to the hypercoagulable and prothrombotic state,
patients with malignancy are more likely to develop VTE
compared to those without malignancy [17, 18]. Despite
the fact that some patients with active malignancy were
included in the major DOAC trials [1115], the numbers
were relatively small and there is a lack of evidence compar-
ing DOACs and LMWH in this patient population and more
future research in this area is encouraged. Two meta-analyses
of patients with a history of cancer or active cancer showed
no difference in terms of safety and efficacy for patients on
DOACs compared to warfarin plus parenteral anticoagula-
tion [19, 20]. LMWHs have been shown to be superior to
warfarin for the prevention of recurrent VTE in the cancer
population [21]. According to the most recent guidelines,
it is still recommended to use LMWH monotherapy over
13
warfarin and DOACs in patients with cancer [16]. All VTE
patients with an active malignancy admitted to our hospital
agement of their VTE episode.
We also observed an increase in DOAC utilization in the
emergency department. This could be due to the fast onset
of peak plasma levels without the need to be bridged with
parenteral anticoagulation in the initial phase, which could
expedite discharging patients. Due to the increase in DOAC
utilization in the emergency department, clinicians who
serve in these settings should have a complete understand-
ing of the pharmacology of these agents in conjunction with
practical guidance on their usage [22, 23].
Treatment of new onset VTE with DOACs at our institu-
tion was associated with shorter hospital length of stay and
less days on parenteral anticoagulation compared to warfa-
rin. As a result, this could reduce the overall medical costs
which have been documented in several other investigations
[24, 25].
Our study has certain limitations including: retrospective
design, single-center, small sample size, and the possibility
of hospital database incompletely capturing eligible patients
which could affect the external validity. We did not assess
for prescription coverage or financial factors that could have
influenced anticoagulant selection. There are also vascular
medicine and cardiology consult services that might have
been utilized for assistance in oral anticoagulant selection,
however, these services are not required for initiation of any
oral anticoagulant. This study highlights the importance of
factors that might contribute to prescribing DOACs over
warfarin and vice versa. In the future, larger investigations
and postmarketing registries are warranted to evaluate and
assess the utilization of these agents in real-world patients
who might be less adherent and have more comorbidities
than trial populations.
In conclusion, clinicians now have multiple anticoagulant
options for the treatment of VTE. This evaluation observed
at a large, tertiary, academic medical center, DOACs are
being prescribed more than warfarin for treatment of new
onset VTE. Renal and liver function may influence the agent
prescribed, and utilization of a DOAC may decrease the
number of days on parenteral anticoagulants and hospital
LOS.
Compliance with ethical standards
Conflict of interest The authors have no conflicts of interest to
declare.
References
1. Piazza G et al (2009) Venous thromboembolic events in hospital-
ised medical patients. Thromb Haemost 102(3):505510
Evaluation of anticoagulation selection for acute venous thromboembolism
2. Heit JA (2006) The epidemiology of venous thromboembolism in
the community: implications for prevention and management. J
Thromb Thrombolysis 21:2329
3. White RH (2003) The epidemiology of venous thromboembo-
lism. Circulation 107(23 Suppl 1):I4I8
4. Goldhaber SZ, Bounameaux H (2012) Pulmonary embolism and
deep vein thrombosis. The Lancet 379(9828):18351846
5. Johnson JA et al (2011) Clinical Pharmacogenetics Implemen-
tation Consortium Guidelines for CYP2C9 and VKORC1 geno-
types and warfarin dosing. Clin Pharmacol Ther 90(4):625629
6. Hawkins D (2004) Limitations of traditional anticoagulants.
Pharmacotherapy 24(7P2):62S65S
7. Weitz JI, Linkins LA (2007) Beyond heparin and warfarin: the
new generation of anticoagulants. Expert Opin Investig Drugs
16(3):271282
8. Bauer KA (2013) Pros and cons of new oral anticoagulants. Hema-
tol Am Soc Hematol Educ Program 2013:464470. doi:10.1182/
asheducation-2013.1.464
9. Finks SW, Trujillo TC, Dobesh PP (2016) Management of venous
thromboembolism: recent advances in oral anticoagulation ther-
apy. Ann Pharmacother 50(6):486501
10. Kubitza D et al (2005) Safety, pharmacodynamics, and pharma-
cokinetics of single doses of BAY 59-7939, an oral, direct factor
Xa inhibitor. Clin Pharmacol Ther 78(4):412421
11. Agnelli G et al (2013) Oral apixaban for the treatment of acute
venous thromboembolism. N Engl J Med 369(9):799808
12. Bauersachs R et al (2010) Oral rivaroxaban for symptomatic
venous thromboembolism. N Engl J Med 363(26):24992510
13. Buller HR et al (2012) Oral rivaroxaban for the treatment of symp-
tomatic pulmonary embolism. N Engl J Med 366(14):12871297
14. Buller HR et al (2013) Edoxaban versus warfarin for the treat-
ment of symptomatic venous thromboembolism. N Engl J Med
369(15):14061415
15. Schulman S et al (2009) Dabigatran versus warfarin in the
treatment of acute venous thromboembolism. N Engl J Med
361(24):23422352
5
16. Kearon C et al (2016) Antithrombotic therapy for VTE disease:
CHEST guideline and expert panel report. Chest 149(2):315352
17. Green KB, Silverstein RL (1996) Hypercoagulability in cancer.
Hematol Oncol Clin North Am 10(2):499530
18. Falanga A et al (1994) The hypercoagulable state in cancer
patients: evidence for impaired thrombin inhibitions. Blood
Coagul Fibrinolysis 5 (Suppl 1):S19S23, (discussion 5964)
19. Larsen TB et al (2014) Non-vitamin K antagonist oral anticoagu-
lants and the treatment of venous thromboembolism in cancer
patients: a semi systematic review and meta-analysis of safety
and efficacy outcomes. PLoS One 9(12):e114445
20. Van der Hulle T et al (2014) Meta-analysis of the efficacy and
safety of new oral anticoagulants in patients with cancer-asso-
ciated acute venous thromboembolism. J Thromb Haemost
12(7):11161120
21. Lee AY et al (2003) Low-molecular-weight heparin versus a cou-
marin for the prevention of recurrent venous thromboembolism in
patients with cancer. N Engl J Med 349(2):146153
22. Cervellin G et al (2015) Quality and safety issues of direct oral
anticoagulants in the emergency department. Semin Thromb
Hemost 41(3):348354
23. Hogg K et al (2016) Direct oral anticoagulants: a practical guide
for the emergency physician. Eur J Emerg Med. doi:10.1097/
mej.0000000000000400
24. Amin A et al (2015) Comparison of differences in medical costs
when new oral anticoagulants are used for the treatment of patients
with non-valvular atrial fibrillation and venous thromboembolism
vs warfarin or placebo in the US. J Med Econ 18(6):399409
25. Deitelzweig S et al (2013) Medical costs in the US of clinical
events associated with oral anticoagulant (OAC) use compared
to warfarin among non-valvular atrial fibrillation patients 75
and <75 years of age, based on the ARISTOTLE, RE-LY, and
ROCKET-AF trials. J Med Econ 16(9):11631168
13