PerspectiveIn the Journals Plus

Dabigatran associated with higher 30-day readmission rate for bleeding than warfarin

Show Citation

September 21, 2017

Patients treated with dabigatran for nonvalvular atrial fibrillation appeared to be at higher risk for 30-day hospital readmission due to bleeding, according to a population-based, retrospective cohort study.

Use of dabigatran — the first non-vitamin-K antagonist oral anticoagulant to be approved as an alternative treatment to warfarin for non-valvular atrial fibrillation — has increased among hospitals. However, cases of serious bleeds associated with dabigatran therapy have been reported.

“Dabigatran works rapidly following its initiation; hence it may lead to more early-onset bleeds,” Esther W. Chan, BPharm(Hons), MClinPharm, PhD, GradCertPharmEc, PhD, FSHP, associate professor in the department of pharmacology and pharmacy at The University of Hong Kong, and colleagues wrote. “In contrast, warfarin may take weeks to achieve anticoagulation stability following its initiation, resulting in less bleeding.”

Limited information is available regarding the 30-day readmission rates for patients with nonvalvular atrial fibrillation treated with dabigatran or warfarin. Chan and colleagues aimed to address this knowledge gap.

The researchers reviewed population-wide electronic medical records from the Clinical Data Analysis and Reporting System of the Hong Kong Hospital Authority to identify 51,946 patients newly diagnosed with nonvascular atrial fibrillation from 2010 to 2014. Investigators matched 5,160 of these individuals by propensity score (n = 2,580 each for dabigatran and warfarin users).

Chan and colleagues assessed rate of hospital admission with bleeding (mean follow-up, 425 ± 434 days) in association with dabigatran or warfarin therapy.

hospitalization with bleeding occurred in 151 patients who received dabigatran and 172 patients who received warfarin; incidence rates appeared comparable between groups (5 vs. 5.8 per 100 patient-years; incidence rate ratio (IRR): 0.92; 95% CI, 0.66-1.28).

Subgroup analyses showed dabigatran use was associated with higher admission rate with gastrointestinal bleeding (2.9 vs. 2.1 per 100 patient-years; IRR = 2.21; 95% CI, 1.28-3.83) and a lower rate of intracranial hemorrhage (0.5 vs. 1.4 per 100 patient-years; IRR = 0.26; 95% CI, 0.12-0.55).

Post hoc analysis showed an association between dabigatran use and a lower rate of bleeding among patients aged younger than 75 years (IRR = 0.59; 95% CI, 0.35-0.97) compared with those aged 75 years or older (IRR = 1.29; 95% CI, 0.83–2.01).

Results showed 13.5% of dabigatran-treated patients and 5.1% of warfarin-treated patients were readmitted to the hospital with bleeding within 30 days of discharge.

Among patients who were continuously prescribed their initial anticoagulants upon discharge, dabigatran use appeared associated with a higher risk for 30-day readmission with bleeding (HR = 2.87; 95% CI, 1.10-7.43).


Chan and colleagues attributed these associations to dabigatran’s ability to achieve full anticoagulation effect more quickly than warfarin. They also cited limited guidance on prevention of recurrent bleeding with dabigatran.

“Considering that dabigatran achieves full anticoagulation more rapidly compared with warfarin, close early monitoring of patients initiated on anticoagulation following hospital discharge and strategies to reduce recurrence of bleeding with dabigatran are warranted,” Chan and colleagues wrote. – by Kristie L. Kahl

Disclosures: Chan reports research funding from Bristol-Myers Squibb, Hong Kong Research Grants Council, Hong Kong Health and Medical Research Fund, Janssen and Pfizer. Please see the full study for a list of all other researchers’ relevant financial disclosures.

itj+ Infographic

itj+ Perspective

Author Name

This study’s findings are consistent with a recent U.S. survey — conducted by Nadine Shehab, PharmD, MPH, and colleagues — of ED admissions for drug adverse events that showed oral anticoagulants have the highest risk for acute injury of any class of therapeutic drugs. This new study’s annualized hospitalization rates of 5% and 5.8% for dabigatran and warfarin for bleeding in patients with atrial fibrillation also are consistent with previously published results. However, had this Hong Kong study been conducted in the United States, one would expect even higher hospital admission rates for dabigatran because the lower 110-mg dose — taken by 75% in this population — was not available in the United States until recently, and is not recommended for use in atrial fibrillation.

The study also identifies an important new anticoagulant safety issue — high rates of rehospitalization for bleeding within 30 days affecting 13.5% of dabigatran patients compared with 5.1% of warfarin patients. After a bleeding event, the warfarin dose can be adjusted downward, a flexibility not available for dabigatran except for the untested 75-mg dose. However, one safety limitation of dabigatran — lack of reversal agent — has been addressed in the United States with the approval of idarucizumab (Praxbind, Boehringer Ingelheim).

Although the benefit of oral anticoagulants in preventing ischemic stroke is well established, the growing evidence of harms illustrates the need to improve the safety of this high-risk treatment. Among key issues are patient selection to exclude lower-risk patients, and appropriate use of concomitant therapy with antiplatelet agents. When my colleagues and I analyzed the large dabigatran trial, we observed that concomitant use of antiplatelet agents, which were used in more than 70% of the bleeding cases in this new study, and we found it doubled the risk of a bleeding event. Regulators and manufacturers also need to reexamine whether dabigatran’s safety profile can be improved with plasma level monitoring, which my previous analysis suggests might lower bleeding rates by 20%.


Moore TJ, et al. BMJ. 2014;doi:10.1136/bmj.g4517.

Shehab N, et al. JAMA. 2016;doi:10.1001/jama.2016.16201.

Thomas J. Moore

Institute for Safe Medication Practices

Milken Institute School of Public Health at The George Washington University

Disclosure: Moore reports no relevant financial disclosures.