Article Text

Download PDFPDF

Cohort study
Restarting oral anticoagulation among patients with atrial fibrillation with gastrointestinal bleeding was associated with lower risk of all-cause mortality and thromboembolism
  1. Waqas T Qureshi1,
  2. Usama Nasir2
  1. 1Heart and Vascular Center of Excellence, Wake Forest School of Medicine, Medical Center Blvd, Winston-Salem, North Carolina, USA
  2. 2Department of Internal Medicine, University of Connecticut, Farmington, Connecticut, USA
  1. Correspondence to : Dr Waqas T Qureshi, Heart and Vascular Center of Excellence, Wake Forest School of Medicine, Medical Center Blvd, NC 27157, USA; wqureshi{at}

Statistics from

Commentary on: OpenUrlAbstract/FREE Full Text


Atrial fibrillation (AF) is associated with higher risk of thromboembolism and thus warrants the use of anticoagulants for prevention of adverse outcomes.1 However, anticoagulation comes at the price of a higher risk of gastrointestinal bleeding (GIB). Current guidelines recommend interrupting anticoagulation in the case of GIB but do not inform on whether it should be resumed. This study assesses risks of all-cause mortality, thromboembolism, major bleeding and recurrent GIB associated with resuming antithrombotic treatment after GIB among patients with AF.


This prospective cohort study included non-valvular patients with AF who experienced GIB while receiving single or combined antithrombotic treatment with an antiplatelet and/or anticoagulant in a Danish cohort from 1996 to 2012. Patients with deep vein thrombosis or pulmonary embolism up to 6 months before the inclusion event were excluded. Study outcomes were all-cause mortality or death due to thromboembolism, major bleeding, or recurrent GIB. Thromboembolism was defined by diagnosis codes for ischaemic stroke, transient ischaemic attack or systemic thromboembolism. Major bleeding was defined by diagnosis codes of intracranial bleeding, or severe bleeding from the respiratory, gastrointestinal, or urinary tract. Patients were followed up until 31 December 2012. Multivariable Cox regression models were used to examine the risk of outcomes for the groups restarting anticoagulation versus group not resuming anticoagulation.


Overall, 3409 patients (mean age 77.9±9.3 years, 44.6% females, mean CHA2DS2-VASc score 3.6±1.5 and mean HAS-BLED score was 3±1) were included. Antithrombotic therapy was restarted in 72.9%; single agent in 21.3% (aspirin 35.5%, adenosine receptor antagonist 3.0%); dual treatment with anticoagulants plus aspirin in 10.7%; anticoagulants plus adenosine receptor antagonist in 0.6%, aspirin plus adenosine receptor antagonist in 1.5% and triple treatment with oral anticoagulants plus aspirin and adenosine receptor antagonist in 0.3% of patients. Over a median period of 2 years, restarting single anticoagulant was associated with the lowest rate of all-cause mortality (HR 0.39, 95% CI 0.34 to 0.46) and thromboembolism (HR 0.41, CI 0.31 to 0.54). On the other hand, risk of major bleeding was significantly increased only in patients who restarted single treatment with oral anticoagulants (HR 1.37, 95% CI 1.06 to 1.77). Recurrent GIB was insignificant in all the other groups. Dual treatment with oral anticoagulants and platelets also resulted in decreased mortality (HR 0.41, 95% CI 0.32 to 0.52) and thromboembolism (HR 0.54, 95% CI 0.36 to 0.82).


This study highlights the importance of restarting anticoagulation among patients with AF who experience GIB. Furthermore, resuming single oral antithrombotic treatment appeared to be better in terms of net benefit as compared with resuming dual or triple antithrombotic treatment. In addition, use of oral anticoagulant versus use of antiplatelet was better in terms of mortality and thromboembolic outcomes without significant increased risk in recurrent GIB. These results are reassuring and have laid down the need for an randomised controlled trial.

Once we are sure that restarting anticoagulation is surely beneficial for our patients who present with GIB, the next step would be to address the timing of restarting anticoagulation. In addition, the risks of recurrence of upper and lower GIB differ significantly and have not been well addressed in this study. It is also unclear if HAS-BLED and CHA2DS2-VASC scores hold similarly in these high-risk individuals and if they could be used to inform decision-making for restarting anticoagulation in these patients. A common rule used is if CHA2DS2-VASC - HAS-BLED ≥2, restarting anticoagulation is safe. However, such practice is neither evidence-based nor widely accepted. Furthermore, racial and gender disparities exist among restarting anticoagulation, which should be further studied.2 It is interesting to note that the resumption of anticoagulation among this cohort was much higher than US-based studies.3 In addition, several of these patients might have been started on antacids, proton pump inhibitors and H2 receptor antagonists. The impact of these drugs on modifying the risk of recurrent GIB is still unknown.

Clinical implications

Although the study is not a major clinical trial that could change clinical practice, the study does underscore the need for aggressive reassessment of resuming anticoagulation in patients with AF GIB who present with a first episode of GIB. In addition, resuming single antithrombotic treatment appears to be the best approach in such patients.


View Abstract


  • Competing interests None declared.

  • Provenance and peer review Commissioned; internally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.