Continued use of low-dose aspirin may increase risk of bleeding after gastrointestinal endoscopic submucosal dissection: A meta-analysis.

Background/aims Endoscopic submucosal dissection has been widely accepted. At present, the number of antiplatelet (APT) users has been growing. Moreover, because of high risks of thromboembolism, some patients need to continuously receive APT agents. The relationship between hemorrhage and continuous therapy with low-dose aspirin (LDA) remains controversial. Materials and methods A systematic search was conducted; studies were screened out- if data of no-anticoagulant/APT drugs use and interrupted and continued-LDA use were reported separately. The Newcastle-scale was chosen to assess the quality of the included studies. Review Manager 5.2 was used for quality assessment statistical analysis, and the odd ratio (OR) and 95% confidence interval (CI) were calculated. Results Pooled data suggested a significantly higher bleeding ratio in the LDA-continued group compared to both the LDA-interrupted group (OR=2.05, 95% CI=1.05-3.99) and no-anticoagulant/APT group (OR=2.89, 95% CI=1.86-4.47). However, the LDA-interrupted group did not differ significantly from the no-anticoagulant/APT group. The en bloc resection rates of the LDA-continued group versus the LDA-interrupted group, the LDA-continued group versus no-anticoagulant/APT group, and the LDA-interrupted group versus the no-anticoagulant/APT group were similar (OR=0.82, 95% CI=0.21-3.24, p=0.78; OR=0.80, 95% CI=0.24-2.65, p=0.71; OR=1.41, 95% CI=0.38-5.24, p=0.60, respectively). Conclusion There is an extremely high ratio of bleeding in the LDA-continued group compared to both the LDA-interrupted group and no-anticoagulant/APT group. All groups had similar ratios of en bloc resection.