Resuming antiplatelet or anticoagulant therapy after gastrointestinal bleeding was associated with fewer major vascular events and lower all-cause mortality, despite a higher risk for recurrent bleeding, according to a systematic review and meta-analysis. The analysis also found that restarting treatment within seven days appeared to offer a favorable benefit-risk balance, driven by fewer major vascular events despite more recurrent bleeding.
“Because of the lack of randomized controlled studies, current guidelines offer limited specific recommendations on optimal resumption timing,” corresponding authors Kelvin K.F. Tsoi, PhD, and Joseph J.Y. Sung, MD, PhD, and colleagues wrote in the analysis, published in Clinical Gastroenterology and Hepatology. Earlier cohort studies, the authors noted, helped establish the widely adopted practice of restarting antithrombotic therapy within one to two weeks after gastrointestinal bleeding, and previous meta-analyses have shown that resumption reduces thromboembolic events and mortality. “However, findings regarding recurrent GIB risk have been inconsistent, and these prior meta-analyses lacked a systematic evaluation of resumption timing strategies.”
The current analysis included 21 studies involving 18,622 patients who experienced gastrointestinal bleeding (GIB) while receiving antithrombotic therapy, including antiplatelet therapy (APT), anticoagulant therapy (ACT), or both. The investigators compared outcomes between patients who resumed therapy and those who did not, and also compared restarting treatment within seven days of GIB vs after seven days.
The review included one randomized trial, one prospective cohort study, one combined retrospective-prospective cohort study, and 18 retrospective cohort studies from North America, Europe, and Asia. Researchers searched five databases for studies published through September 2025. The primary outcomes were recurrent GIB, major vascular events, and all-cause mortality.
Among eight studies of antiplatelet therapy, patients who resumed treatment had a 42% higher risk for recurrent GIB than those who did not. However, restarting therapy reduced major vascular events by 30% and lowered all-cause mortality by 56%. In absolute terms, recurrent bleeding increased from 6% to 8%, while major vascular events fell from 13% to 9% and all-cause mortality dropped from 19% to 9% within one year.
The findings were similar for anticoagulant therapy. Across 15 studies with 17 treatment groups, restarting anticoagulation increased the risk for recurrent GIB by 59% but reduced major vascular events by 55% and all-cause mortality by 47%. At one year, recurrent bleeding increased from 5% to 7%, while major vascular events fell from 13% to 6% and all-cause mortality declined from 26% to 15%.
The investigators also examined when therapy should be restarted, an area where current guideline recommendations remain limited, the paper noted. Six studies compared restarting treatment within seven days of GIB with restarting after seven days.
Restarting therapy within seven days increased the risk for recurrent bleeding by 49%, but it also reduced major vascular events by 61% without significantly affecting all-cause mortality. In absolute terms, recurrent bleeding increased from 6% to 8%, while major vascular events fell from 20% to 8%. Although mortality was lower with earlier resumption, the difference was not statistically significant.
Taken together, the findings suggest that antithrombotic therapy resumption after GIB may offer a favorable benefit-risk trade-off for many patients once the acute bleeding episode is controlled, but the authors emphasized that timing should be individualized by bleeding severity, thrombotic risk, indication for therapy, and drug class. Although decisions should be individualized based on bleeding severity, thrombotic risk, and the reason for antithrombotic therapy, the pooled data support restarting treatment within one week for many patients rather than delaying it longer.
Subgroup analyses showed similar results regardless of the bleeding source. Among patients with upper GIB, resuming therapy was associated with more than a twofold increase in recurrent bleeding but reduced major vascular events by about two-thirds and lowered all-cause mortality by 44%. Similar reductions in major vascular events were seen after lower GIB, although the reduction in mortality did not reach statistical significance because fewer studies were available for analysis.
Risk patterns also changed over time. The increased risk for recurrent bleeding was greatest during the first year after the initial bleeding event and diminished with longer follow-up. In contrast, the reductions in major vascular events and all-cause mortality persisted for up to five years, suggesting that the long-term benefits of resuming antithrombotic therapy may outweigh the early increase in bleeding risk.
Several sensitivity analyses supported the findings. Restricting the analysis to studies that reported hazard ratios produced similar results, and no single study had a substantial effect on the overall estimates. Visual inspection of funnel plots also suggested a low risk for publication bias.
The investigators acknowledged several limitations. Most of the evidence came from retrospective observational studies, and only one randomized trial was available, limiting the ability to draw causal conclusions. Definitions of major vascular events varied across studies, and relatively few studies examined when therapy should be restarted. The analysis also could not determine how the reason for antithrombotic therapy, the specific drug used, bleeding severity or source, or patient characteristics affected the best timing for resuming treatment.
The study received no external funding, and the authors reported no conflicts of interest.
Expert Insight
GI & Hepatology News spoke with first author Yihang Liu, a PhD student at Lee Kong Chian School of Medicine, Nanyang Technological University, Singapore, about the findings.
Your meta-analysis suggests that resuming antiplatelet or anticoagulant therapy within seven days offers the best overall benefit-risk balance. In everyday practice, which patients do you think should not have therapy resumed this early, despite the overall findings?
Liu: While early resumption (within 7 days) showed a favorable benefit-risk balance in our pooled analysis, this should not be applied uniformly.
First, patients with severe index bleeding, such as those with hemodynamic instability or ongoing transfusion requirements, likely represent a population where the ulcer or bleeding site has not yet stabilized. Resuming antithrombotic therapy too early in this group likely carries a disproportionately higher rebleeding risk than our overall estimate suggests, and warrants a more cautious, individualized approach.
Second, patients with a weaker indication for antithrombotic therapy may also be better served by waiting. For example, someone on aspirin for primary cardiovascular prevention, or on a P2Y12 inhibitor for a remote coronary stent placed many years ago, has comparatively low thrombotic risk if therapy is held a bit longer. For these patients, the calculus tips toward delay, since they have less to lose from waiting and more to lose from early rebleeding.
Current guidelines provide limited recommendations on the optimal timing of antithrombotic resumption after gastrointestinal bleeding. Based on your findings, do you believe these data are strong enough to influence future guideline recommendations, and what additional evidence is still needed before practice should change?
Liu: I think our findings can meaningfully influence how clinicians and future guidelines think about resumption timing, since they provide one of the most comprehensive quantitative estimates available to date on this question. However, I don’t think they’re strong enough, on their own, to justify a specific practice change or a fixed timing rule. The evidence base remains dominated by retrospective observational studies, with only one RCT available in this space.
What’s needed now are prospective randomized trials that test specific resumption timepoints. Ideally, these would be stratified by bleeding severity and thrombotic risk, and conducted separately for antiplatelet versus anticoagulant therapy, or even by specific agent, such as warfarin versus DOACs, or aspirin versus P2.