American Journal of Emergency Medicine 34 (2016) 9–13

Contents lists available at ScienceDirect

American Journal of Emergency Medicine

journal homepage: www.elsevier.com/locate/ajem

Review

Safety of direct oral anticoagulants: insights from

postmarketing studies☆,☆☆,★,★★
Todd C. Villines, MD a, W. Frank Peacock, MD b,⁎
a
Department of Medicine, Cardiology Service, Walter Reed National Military Medical Center, Bethesda, MD
b
Department of Emergency, Medicine, Baylor College of Medicine, Houston, TX

a r t i c l e i n f o a b s t r a c t

Direct oral anticoagulants (DOACs) have been marketed in the United States since 2010. While numerous large-
scale prospective phase 3 outcomes studies have documented the effectiveness of DOACs for the prevention of
stroke and systemic embolism in patients with nonvalvular atrial fibrillation, the primary safety concern with
all of these drugs—as it is with the more established oral anticoagulant warfarin—is the risk of major bleeding.
Postmarketing surveillance studies (PMSS) provide the opportunity to evaluate the safety of these recently ap-
proved drugs across a spectrum of patients that may be broader than those included in randomized controlled
trials. This review will summarize the safety findings of numerous recently performed, large-scale PMSS evalua-
tions, and consider the currently available evidence regarding the risks for bleeding in patients treated with
DOACs, in order to give providers and patients additional evidence regarding the safety of DOACs.
© 2016 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).

Direct oral anticoagulants (DOACs) have been marketed in the
United States since 2010, when the U.S. Food and Drug Administration
(FDA) approved the direct thrombin inhibitor dabigatran etexilate for
prevention of stroke and systemic embolism in patients with
nonvalvular atrial fibrillation (NVAF) [1]. The oral direct factor Xa
(FXa) inhibitors rivaroxaban and apixaban were subsequently ap-
proved for treatment and prevention of deep vein thrombosis (DVT)
and pulmonary embolism (PE) (including in patients who have under-
gone hip or knee replacement surgery) and for reduction in risk of
stroke and systemic embolism in patients with NVAF [2,3]. Dabigatran
was also approved for treatment and prevention of DVT and PE and
for the prophylaxis of DVT and PE in patients who have undergone
☆ Funding: This work was supported by Boehringer Ingelheim Pharmaceuticals, Inc
(BIPI). Editorial support was provided by Mark Poirier of Envision Scientific Solutions,
which was contracted and funded by BIPI. The authors received no direct compensation
related to the development of the manuscript.
☆☆ Conflict of Interest: TCV serves on a speaker bureau for BIPI and is compensated for
such activity. WFP has obtained research grants fromAbbott, Alere, Banyan, Cardiorentis,
Portola, Roche, and The Medicines Company; served as a consultant for Alere, BG Medi-
cine, Beckman, Boehringer Ingelheim, Cardiorentis, Instrument Labs, Janssen, Prevencio,
The Medicines Company, and ZS Pharma; and has ownership interests in Comprehensive
Research Associates, LLC, and Emergencies in Medicine, LLC.
★ Authorship: The authors meet criteria for authorship as recommended by the Interna-
tional Committee of Medical Journal Editors (ICMJE). Both authors had access to the data and
participated in writing the manuscript. BIPI was given the opportunity to review the manu-
script for medical and scientific accuracy as well as intellectual property considerations.
★★ The opinions expressed herein are those of the authors only and are not to be con-
strued as representing those of the U.S. Department of Defense or the U.S. Government.
⁎ Corresponding author at: Emergency Medicine, Ben Taub General Hospital, 1504 Taub
Loop, Houston, TX 77030.
E-mail address: Frankpeacock@gmail.com (W.F. Peacock).
hip replacement surgery [1]. Additionally, the FXa inhibitor edoxaban
has been approved in the United States to reduce the risk of stroke
and systemic embolism in patients with NVAF and for the treatment
of DVT and PE [4].
The primary safety concern with all of these drugs—as it is with the
older, more established oral anticoagulant warfarin—is the risk of bleed-
ing as a complication of deliberate anticoagulation aimed at preventing
pathologic thrombosis. Although statistically rare, an intracranial hemor-
rhage (ICH) is the most feared adverse event associated with all oral an-
ticoagulants because of its devastating clinical sequelae and high rate of
mortality. Anticoagulant associated gastrointestinal hemorrhages are
more common, but are less often likely to be fatal adverse events.
In recent years, researchers have reported the findings of
postmarketing surveillance studies (PMSS) of adverse events associated
with the DOACs. These studies followed the publication of randomized
controlled trials (RCTs) that established the foundational evidence for
the comparative safety and efficacy of DOACs vs warfarin and formed
the basis for FDA approval [5-7]. Postmarketing surveillance may take
the form of independent studies, evaluations performed by regulators,
or as part of phase 4 research performed by the drug manufacturers.
These studies are observational in nature. Postmarketing surveillance
is typically conducted in retrospect from large databases (eg, those
maintained by Medicare, the FDA Adverse Event Reporting System
[FAERS], private health-maintenance organizations, health benefits
provider roles, health insurance company), or obtained from ongoing
prospective registries (eg, Global Registry on Long-Term Oral Anti-
thrombotic Treatment in Patients with Atrial Fibrillation [GLORIA-AF]
[8] or Global Anticoagulant Registry in the FIELD-Atrial Fibrillation
[GARFIELD] [9] for patients with NVAF).

http://dx.doi.org/10.1016/j.ajem.2016.09.047
0735-6757/© 2016 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
 10
Table
Bleeding and Myocardial Infarction Risk with DOACs vs Warfarin Given for Stroke Prevention in Atrial Fibrillation in RCTs and Large-Scale Observational Studies (Hazard Ratios and Relative Risks are Adjusted Unless Indicated)

Study Mean Age Follow-Up Adjustment Major Bleeding Risk Intracranial Bleeding Risk GI Bleeding Risk Myocardial Infarction Risk
(95% CI) (95% CI) (95% CI) (95% CI)

Dabigatran etexilate

T.C. Villines, W.F. Peacock / American Journal of Emergency Medicine 34 (2016) 9–13
RE-LY [5]⁎ (150 mg BID) 71 y Median 2.0 y Cox proportional- RR 0.93 (0.81-1.07) RR 0.40 (0.27-0.60) RR 1.50 (1.19-1.89) RR 1.38 (1.00-1.91)
hazards models
U.S. Medicare [1] N65 y Total 37,587 patient-years PSM HR 0.97 (0.88-1.07) HR 0.34 (0.26-0.46) HR 1.28 (1.14-1.44) HR 0.92 (0.78-1.08)
(75 mg and 150 mg)
U.S. Department of Defense [15] 74 y Mean 297 ± 259 d PSM HR 0.87 (0.74-1.03) Unadjusted HR 0.49† (0.30-0.79) Unadjusted HR 1.13† (0.94-1.37) Unadjusted HR 0.65† (0.45-0.95)
(75 mg and 150 mg)
MarketScan/Clinformatics [16] 68 y Mean 5 mo PSM HR 0.75 (0.65-0.87) HR 0.31 (0.17-0.54) HR 0.97 (0.79-1.18) HR 0.89 (0.57-1.38)
(dose not specified)
Danish Registry of Medicinal 71 y Median 10.5 mo‡ PSM HR 0.66 (0.36-1.14) HR 0.08 (0.01; 0.40) HR 1.12 (0.67-1.83) HR 0.40 (0.21-0.70)
Product Statistics [17] (150 mg)

Rivaroxaban 20 mg QD
ROCKET AF [6]⁎ Median 73 y Median 707 d Cox proportional- HR 1.04 (0.90-1.20) HR 0.67 (0.47-0.93) RR 1.46 P b.001ǁ HR 0.81 (0.63-1.06)
hazards models
§
U.S. Department of Defense [18] 78 y Total 455 d None IR 2.86 (2.61-3.13) IR 0.22 (0.15-0.30) IR 2.53 (2.30-2.78) NR
Post-Marketing Safety Surveillance 78 y Total 2 y None IR 2.89 (2.71-3.08)§ Incidence 0.2% (79 of 39,052) Incidence 2.2% NR
(U.S. Department of Defense) [19]§ (846 of 39,052)

Apixaban 5 mg BID
ARISTOTLE [7]⁎ Median 70 y Median 1.8 y Cox proportional- HR 0.69 (0.60-0.80) HR 0.42 (0.30-0.58) HR 0.89 (0.70-1.15) HR 0.88 (0.66-1.17)
hazards models
Humedica [20] NR Total 180 d Cox proportional- HR 0.75 (0.63-0.88) NR NR NR
hazards models

ARISTOTLE = Apixaban for Reduction In STroke and Other ThromboemboLic Events in Atrial Fibrillation; BID = twice daily; CI = confidence interval; GI = gastrointestinal; HR = hazard ratio; IR = incidence rate per 100 person-years; MVA =
multivariate analysis; NR = not reported; NVAF = nonvalvular atrial fibrillation; PSM = propensity score matched; QD = every day; RCT = randomized controlled trial; RE-LY = Randomized Evaluation of Long-term anticoagulation therapY; ROCK-
ET AF = Rivaroxaban Once daily oral direct factor Xa inhibition Compared with vitamin K antagonism for prevention of stroke and Embolism Trial in Atrial Fibrillation; RR = relative risk; SPAF = stroke prevention in atrial fibrillation.
⁎ These studies are phase 3 RCTs in patients with NVAF.
†
Following propensity score matching, the unadjusted and adjusted hazard ratios were almost all identical. For secondary bleeding end points, unadjusted hazard ratios were reported.
‡
Median follow-up includes 110-mg and 150-mg patients.
§
No warfarin group.
ǁ
RR was calculated from data reported in the primary publication [6].
 T.C. Villines, W.F. Peacock / American Journal of Emergency Medicine 34 (2016) 9–13
Postmarketing research, therefore, attempts to assess the effective-
ness and safety of the drug in a “real-world” setting that is representa-
tive of how it is being prescribed and used in clinical practice [10,11].
The most common methodology in these studies utilizes International
Classification of Diseases (ICD) coding data to identify patients, deter-
mine baseline demographics and clinical characteristics, and assess out-
comes or parameters of interest. The positive predictive value of using
ICD-9 and ICD-10 codes for identification of patients with strokes has
been validated at 80% to 97% [12,13]. Validation studies have also dem-
onstrated good positive predictive values with these codes for identify-
ing the presence and location of GI bleeding [14]. Because ICD-10 for
inpatient hospital procedures was recently adopted in the United
States, published studies are based on ICD-9 coding.
Study cohorts in PMSS are not randomized, but researchers may
control for differences in patient characteristics by using multivariable
modeling or propensity score matching. Observational research, such
as PMSS, has inherent limitations because of its uncontrolled,
nonrandomized nature. The compared populations are potentially sub-
ject to confounding factors that may have been excluded in RCTs.
Modeling or propensity score matching can reduce or eliminate these
factors, but some residual confounding variables may remain. As a con-
sequence of potential for bias, the assessment of effectiveness—while
performed and reported—should be interpreted with caution.
These studies may consider treatment in larger and more variable
populations over greater periods of time than would be feasible in a
phase 3 RCT, and thus have the potential to reveal more rare adverse
events or provide more information about anticipated adverse event
rates. Postmarketing research may also provide information on
parameters that RCTs are unable to evaluate due to ethical considerations
(eg, time delay to treatment or the management of rare intentional over-
doses). In addition, PMSS can provide information about the treatment of
patients who would be excluded from RCTs or complex therapeutic sce-
narios (eg, multiple conflicting comorbidities, extremes of age/body hab-
itus, or lifestyle consequences related to complications of the drug in
question).
1. Published studies
1.1. Dabigatran
As the first FDA-approved DOAC, dabigatran has been the most fre-
quent subject of PMSS (a PubMed literature search in October 2015
found 21 completed observational studies with dabigatran and 10
with rivaroxaban) in this therapeutic area. Since the drug was intro-
duced in 2010, several observational studies have provided insights
into the risk of bleeding in patients treated with dabigatran vs the vita-
min K antagonist (VKA) warfarin (Table) [1,5-7,15-20]. The findings of
these studies have been broadly consistent with the results of the Ran-
domized Evaluation of Long-term anticoagulation therapY (RE-LY) trial,
which compared dabigatran with warfarin in patients with NVAF [5].
To date, the U.S. Medicare study reported by Graham et al [1] evalu-
ated the largest cohort of patients taking dabigatran. These researchers
compared bleeding risk in a propensity score matching population of
patients with NVAF who were naïve to anticoagulation and were pre-
scribed either warfarin or dabigatran etexilate (n = 67,207 in each
group). In the Medicare cohort, the risk for major bleeding with
dabigatran was similar to warfarin (adjusted hazard ratio [HR] 0.97;
95% confidence interval [CI], 0.88-1.07). Risk for ICH was significantly
reduced with dabigatran (HR0.34; 95%CI, 0.26-0.46), but risk for major
GI bleeding was increased (HR 1.28; 95% CI, 1.14-1.44). The risk of GI
bleeding was highest in women aged 75-84 years (HR 1.50; 95%CI,
1.20-1.88) and in men and women ≥85 years (HR 1.55; 95% CI, 1.04-
2.32) and (HR 2.18; 95% CI, 1.61-2.97), respectively. There was no differ-
ence in the rate of acute myocardial infarction between the groups (HR
0.92; 95% CI, 0.78-1.08).
11
Several additional studies assessed safety outcomes of dabigatran as
compared with warfarin users among patients with NVAF in the U.S. De-
partment of Defense database [3], in 2 privately administered U.S. pa-
tient databases [16], and in a Danish national database (Table) [17,21].
These studies reported similar findings as compared with the Medicare
analysis, extending PMSS data to non-Medicare patient cohorts. These
authors also found no increased risk of myocardial infarction among
dabigatran users vs patients taking warfarin. In addition to these stud-
ies, researchers for the FDA published a postmarketing bleed compari-
son using data from the FAERS for the first year that dabigatran was
available [22]. Their data showed that despite initial concerns about
bleeding adverse events with dabigatran, incidence rates were not
higher than concurrent incidence rates with warfarin.
1.2. Rivaroxaban
Postmarketing data for the FXa inhibitors (rivaroxaban and
apixaban) have also been published. Two noncomparative studies
have assessed rivaroxaban using the U.S. Department of Defense data-
base (Table) [18,19]. Tamayo et al [18] performed a noncomparative
pharmacovigilance study of records for 27,467 patients with NVAF tak-
ing rivaroxaban. They observed an incidence rate for major bleeding of
2.86 (95% CI, 2.61-3.13) per 100 person-years, and for fatal bleeding
events of 0.08 (95% CI, 0.05-0.14) per 100 person-years. Overall, of the
bleeding population, the vast majority of bleeding events were GI
(88.5%), and fatalities were rare in this cohort. Conversely, the ICH
rate was very low (36 of 478 major bleeding cases), but 50% of all fatal-
ities in the study were recorded in patients with ICH. A subsequent
postmarketing assessment of 39,052 NVAF patients on rivaroxaban con-
firmed these findings, reporting an incidence rate for major bleeding of
2.89 (95% CI, 2.71-3.08) per 100 person-years [19]. The most common
site for major bleeding was GI (87.2%). Intracranial bleeds made up
8.1% of the overall incidence. These findings show that “real-world” rou-
tine clinical care is consistent with the safety profile observed in the
Rivaroxaban Once daily oral direct factor Xa inhibition Compared with
vitamin K antagonism for prevention of stroke and Embolism Trial in
Atrial Fibrillation (ROCKET AF) [6].
1.3. Apixaban
To date, we have one preliminary report of “real-world” use of
apixaban in comparison with warfarin. Using records of patients with
NVAF (2038 on apixaban and 24,872 on warfarin) from a Humedica
(Boston, Mass.) medical record database, the investigators found a
higher rate of bleeds with warfarin than with apixaban (HR 1.34; 95%
CI, 1.13-1.58) [20].
1.4. All DOACs
Recently, a group of researchers presented abstracts at the 2015
European Society of Cardiology meeting from 3 observational studies
comparing bleeding outcomes with apixaban vs dabigatran,
rivaroxaban, or warfarin [23-25]. In the first, using data from the
MarketScan Early View health insurance claims database, they evaluat-
ed major bleeding in 60,277 patients with NVAF (n = 8785 apixaban, n
= 20,963 dabigatran, and n = 30,529 rivaroxaban) [24]. They reported
that rivaroxaban increased the risk for major bleeding (adjusted HR
1.36; 95% CI, 1.23-1.52) compared with apixaban, while dabigatran
had a risk similar to that of apixaban for major bleeding (adjusted HR
0.99; 95% CI, 0.88-1.10). Similar findings were reported in the other 2
abstracts concerning patients with NVAF who newly initiated oral
anticoagulant therapy [20,23]. Comparing Cox proportional-hazards
model adjusted HRs for inpatient and outpatient bleeding events, they
found increased risk with warfarin or rivaroxaban compared with
apixaban (adjusted HRs: warfarin 1.62; 95% CI, 1.20-2.18; rivaroxaban
12 T.C. Villines, W.F. Peacock / American Journal of Emergency Medicine 34 (2016) 9–13
1.70; 95% CI, 1.26-2.29), but not with dabigatran compared with
apixaban (adjusted HR 1.28; 95% CI, 0.92-1.79).
In a study of patients in the Cerner Health Facts hospital database,
Deitelzweig et al [25] compared rates of bleedingrelated hospitalization
readmissions among those previously hospitalized for NVAF and treated
with apixaban, dabigatran, or rivaroxaban. After adjusting for baseline
differences in stroke and bleeding risks, they found increased risk of
bleeding-related 30-day hospitalizations with rivaroxaban vs apixaban
(OR 1.6; P = .04) but no differences with dabigatran vs apixaban
(OR 1.3; P = .30). Analyses of all-cause hospitalization, length of all-
cause hospital stay, and costs related to all-cause hospitalization also
showed apixaban to be better than rivaroxaban and comparable to
dabigatran.
2. Discussion
Randomized and controlled clinical trials with DOACs established
their efficacy and safety, and provided the basis for FDA approval for
the prevention of stroke and systemic embolism in patients with
NVAF and for the prevention or treatment of VTE. In the years since
these drugs became available to health-care providers and their pa-
tients, observational studies with data involving more than 250,000 pa-
tients have provided additional information on safety end points,
particularly bleeding—an adverse effect of all anticoagulation agents.
The primary advantage of postmarketing surveillance is that it pro-
vides information not available in RCTs. Randomized trials generally
represent a small segment of the overall population, one that is selected
based on precise entry and exclusion criteria. These trials are performed
in a cohort that excludes patients likely to be noncompliant, the termi-
nally ill, and those with advanced age or suffering multiple comorbidi-
ties. Furthermore, PMSS analyses allow the evaluation of physician
bias in selecting the intervention population. As physicians commonly
opt to treat only the lowest-risk patients when there are perceived ther-
apeutic risks, the PMSS study methodology thus allows the evaluation of
any potential alterations in a drug’s risk-to-benefit relationship that
may occur as a result.
The primary limitation of all PMSS is their retrospective nature and
their (typically) relatively short duration of follow-up when considering
the currently available data involving DOACs. The currently available
PMSS are limited in that efficacy and safety outcomes have been evaluat-
ed over mean follow-up durations of b1 year and are not designed to se-
lect comparable groups at treatment onset. Warfarin patients, for
example, may have substantial clinical and demographic differences in
comparison with patients treated with the DOACs. Sophisticated statisti-
cal modeling techniques can attempt to adjust for baseline differences in
order to (hopefully) approximate the effects of randomization in cohort
preparation, but the risk of residual confounding always persists. The
strength of PMSS data is that they are based on large cohorts using
well-established coding extraction methods from reliable data sources.
They constitute evidence from “real-world” prescribing and use, and
they may be performed independently of the drug makers (eg, the FDA
dabigatran Medicare study [1]). As such, PMSS data function as a “real-
world” effectiveness translation of the efficacy and safety results of RCTs.
Findings from PMSS with DOACs have, in general, supported the
findings of the phase 3 clinical trials that provided evidence of their ef-
ficacy and tolerability for regulatory agencies [5-7]. The potential for un-
controlled bleeding has long been a concern with warfarin (especially
when taken with certain drugs or food) [26], and similar concerns
arose at the onset of the availability of the first DOAC, dabigatran. How-
ever, surveillance of the FAERS database demonstrated no differences in
reported bleeding during the first year of availability of dabigatran [22].
That assurance was borne out in the Medicare and Department of De-
fense studies that confirmed the initial safety evidence of dabigatran
in the RE-LY trial [1,5,15,27]. More recently, additional PMSS have pro-
vided clarification of bleeding risks for rivaroxaban and apixaban.
Pharmacovigilance studies with rivaroxaban found similar rates of
bleeding to the FDA registration ROCKET AF trial. Importantly, these
studies confirmed that the vast majority of major bleeding with
rivaroxaban was GI rather than intracranial in location. Recently, PMSS
data comparing apixaban with warfarin or other DOACs have provided
further insights.
An interim analysis of the GLORIA-AF patient registry has shown
that the introduction of DOACs has influenced patient treatment pat-
terns, with a substantial increase in appropriate anticoagulation
among patients compared with a prior evaluation [8]. The GLORIA-AF
(N = 56,000) data show that DOACs are already more commonly pre-
scribed than warfarin in newly diagnosed patients with NVAF. In
North America in 2015, overall OAC use was distributed as follows:
VKA 26.1%, dabigatran 25%, rivaroxaban 20.5%, and apixaban 6.6%.
When all DOACs are grouped, they represent more than half of all anti-
thrombotic prescriptions— including antiplatelet agents or aspirin.
Investigators researching the GARFIELD registry database
(N = 17,184) recently confirmed suboptimal thromboprophylaxis with
oral anticoagulants by both female and male patients with newly diag-
nosed NVAF, and observed that perceived bleeding risk was the main rea-
son for not giving VKAs (female, 14.7%; male, 17.8%) [28]. However, just
1.9% of either female or male patients cited a previous bleeding incident.
In addition to GLORIA-AF, GARFIELD, and the RE-LY AF Registry [29],
thromboprophylaxis for patients with NVAF is being prospectively eval-
uated in ongoing registry studies, including the Outcomes Registry for
Better Informed Treatment of Atrial Fibrillation (ORBIT-AF) registry
study [30] and the Euro Heart Survey on Atrial Fibrillation [31]. Findings
from these studies will provide enhanced information on the incorpora-
tion of DOACs into the atrial fibrillation and VTE therapeutic regimens.
Robust PMSS evaluations of DOACs used for the prevention of VTE are
also needed to further understand the safety of these agents.
3. Conclusion
Data from PMSS evaluations of DOACs provide assurance that the
risks associated with their use are manageable and in line with the re-
sults seen in RCTs. As expected, overall risks for major bleeding are sim-
ilar to warfarin, although the risk for intracranial hemorrhage is
significantly reduced, and risk for GI hemorrhage slightly increased.
While all instances of major bleeding are a concern, oral anticoagulation
therapy is aimed at preventing potentially debilitating or fatal thrombo-
embolic events. The risks of bleeding must be balanced against the ben-
efits of reduced risk for stroke or venous thromboembolism.
References
[1] Graham DJ, Reichman ME, Wernecke M, et al. Cardiovascular, bleeding, and mortal-
ity risks in elderly Medicare patients treated with dabigatran or warfarin for
nonvalvular atrial fibrillation. Circulation 2015;131:157–64.
[2] Bristol-Myers Squibb. Eliquis (apixaban) prescribing information. Available at: http://
packageinserts.bms.com/pi/pi_eliquis.pdf. [June 2015. Accessed December 2, 2015].
[3] Villines TC, Schnee J, Fraeman K, et al. A comparison of the safety and effectiveness of
dabigatran and warfarin in non-valvular atrial fibrillation patients in a large
healthcare system. Thromb Haemost 2015;114(6):1290–8.
[4] Daiichi-Sankyo, Inc. Savaysa (edoxaban) prescribing information. Available at:
http://dsi.com/prescribing-information-portlet/getPIContent?productName=
Savaysa&inline=true. [January 2015. Accessed December 2, 2015].
[5] Connolly SJ, Ezekowitz MD, Yusuf S, et al. Dabigatran versus warfarin in patients
with atrial fibrillation. N Engl J Med 2009;361:1139–51.
[6] Patel MR, Mahaffey KW, Garg J, et al. Rivaroxaban versus warfarin in nonvalvular
atrial fibrillation. N Engl J Med 2011;365:883–91.
[7] Granger CB, Alexander JH, McMurray JJ, et al. Apixaban versus warfarin in patients
with atrial fibrillation. N Engl J Med 2011;365:981–92.
[8] Huisman MV, Rothman KJ, Paquette M, et al. Antithrombotic treatment patterns in
patients with newly diagnosed nonvalvular atrial fibrillation: the GLORIA-AF Regis-
try, Phase II. Am J Med 2015;128(12):1306–1313.e1.
[9] Kakkar AK, Mueller I, Bassand JP, et al. International longitudinal registry of patients
with atrial fibrillation at risk of stroke: Global Anticoagulant Registry in the FIELD
(GARFIELD). Am Heart J 2012;163:13–19.e11.
[10] Suvarna V, Phase IV. of drug development. Perspect Clin Res 2010;1(2):57–60.
[11] Glasser SP, Salas M, Delzell E. Importance and challenges of studying marketed
drugs: what is a phase IV study? Common clinical research designs, registries, and
self-reporting systems. J Clin Pharmacol 2007;47:1074–86.
 T.C. Villines, W.F. Peacock / American Journal of Emergency Medicine 34 (2016) 9–13
[12] Kokotailo RA, Hill MD. Coding of stroke and stroke risk factors using international
classification of diseases, revisions 9 and 10. Stroke 2005;36:1776–81.
[13] McCormick N, Bhole V, Lacaille D, Avina-Zubieta JA. Validity of diagnostic codes for
acute stroke in administrative databases: a systematic review. PLoS One 2015;10:
e0135834.
[14] Valkhoff VE, Coloma PM, Masclee GM, et al. Validation study in four health-care da-
tabases: upper gastrointestinal bleeding misclassification affects precision but not
magnitude of drug-related upper gastrointestinal bleeding risk. J Clin Epidemiol
2014;67:921–31.
[15] Villines TC, Schnee J, Fraeman K, et al. The comparative safety and effectiveness of
the oral anticoagulant (OAC) dabigatran versus warfarin utilized in a large
healthcare system in non-valvular atrial fibrillation (NVAF) patients. Circulation
2014;130(Suppl. 2):A18353.
[16] Seeger JD, Bykov K, Bartels DB, Huybrechts K, Zint K, Schneeweiss S. Safety and effec-
tiveness of dabigatran and warfarin in routine care of patients with atrial fibrillation.
Thromb Haemost 2015;114:1277–89.
[17] Larsen TB, Rasmussen LH, Skjøth F, et al. Efficacy and safety of dabigatran etexilate
and warfarin in “real-world” patients with atrial fibrillation: a prospective nation-
wide cohort study. J Am Coll Cardiol 2013;61:2264–73.
[18] Tamayo S, Frank Peacock W, Patel M, et al. Characterizing major bleeding in patients
with nonvalvular atrial fibrillation: a pharmacovigilance study of 27 467 patients
taking rivaroxaban. Clin Cardiol 2015;38:63–8.
[19] Peacock WF, Patel M, Tamayo S, Sicignano N, Hopf K, Yuan Z. Major bleeding in a
post-marketing assessment of 39,052 nonvalvular atrial fibrillation patients on
rivaroxaban. Eur Heart J 2015;36:687 [Suppl.].
[20] Lin I, Masseria C, Mardekian J, et al. Real-world bleeding risk among NVAF patients
prescribed apixaban, dabigatran, rivaroxaban, and warfarin: analysis of electronic
health records. Eur Heart J 2015;36 [Suppl.:Abstract P6216].
[21] Larsen TB, Gorst-Rasmussen A, Rasmussen LH, Skjøth F, Rosenzweig M, Lip GY.
Bleeding events among new starters and switchers to dabigatran compared with
warfarin in atrial fibrillation. Am J Med 2014;127:650–656.e5.
13
[22] Southworth MR, Reichman ME, Unger EF. Dabigatran and postmarketing reports of
bleeding. N Engl J Med 2013;368:1272–4.
[23] Lip GYH, Pan X, Kamble S, et al. Real world comparison of major bleeding risk among
non-valvular atrial fibrillation patients newly initiated on apixaban, dabigatran,
rivaroxaban or warfarin. Eur Heart J 2015;36(Suppl. 1):1085 [Abstract P6217].
[24] Tepper P, Mardekian J, Masseria C, et al. Real-world comparison of bleeding risks
among non-valvular atrial fibrillation patients on apixaban, dabigatran, rivaroxaban:
cohorts comprising new initiators and/or switchers from warfarin. Eur Heart J 2015;
36:339 [Suppl.].
[25] Deitelzweig S, Bruno A, Trocio J, et al. An early evaluation of bleeding-related hospi-
tal readmissions among hospitalized patients with nonvalvular atrial fibrillation
treated with direct oral anticoagulants. Curr Med Res Opin 2016;32:573–82.
[26] Ageno W, Gallus AS, Wittkowsky A, et al. Oral anticoagulant therapy: antithrombotic
therapy and prevention of thrombosis, 9th ed: American College of Chest Physicians
evidence-based clinical practice guidelines. Chest 2012;141:e44S–88S.
[27] Seeger JD, Bartels DB, Huybrechts K, et al. Safety and effectiveness of
dabigatran relative to warfarin in routine care. Circulation 2014;130(Suppl. 2)
[Abstract A16227].
[28] Lip GY, Rushton-Smith SK, Goldhaber SZ, et al. Does sex affect anticoagulant use for
stroke prevention in nonvalvular atrial fibrillation? The prospective global anticoag-
ulant registry in the FIELD-Atrial Fibrillation. Circ Cardiovasc Qual Outcomes 2015;8:
S12–20.
[29] Oldgren J, Healey JS, Ezekowitz M, et al. Variations in cause and management of atri-
al fibrillation in a prospective registry of 15,400 emergency department patients in
46 countries: the RE-LY Atrial Fibrillation Registry. Circulation 2014;129:1568–76.
[30] Piccini JP, Fraulo ES, Ansell JE, et al. Outcomes registry for better informed treatment
of atrial fibrillation: rationale and design of ORBITAF. Am Heart J 2011;162:
606–612.e1.
[31] Nieuwlaat R, Capucci A, Lip GY, et al. Antithrombotic treatment in real-life atrial fi-
brillation patients: a report from the Euro Heart Survey on Atrial Fibrillation. Eur
Heart J 2006;27:3018–26.