AIMS The evidence of a protective effect of proton-pump inhibitor (PPI) in oral anticoagulant (OAC)-treated patients against gastrointestinal bleeding (GIB) is still lacking. We conducted a meta-analysis to estimate the risk of GIB in patients with OAC and PPI cotherapy.

METHODS A systematic search of PubMed, EMBASE, Cochrane and Scopus databases was performed for studies reporting GIB risk in OAC and PPI cotherapy. Primary outcomes were total GIB and major GIB events. Pooled estimates of GIB risk were calculated by a random-effect meta-analysis and reported as odds ratios and 95% confidence interval.

RESULTS A total of 10 studies and 1 970 931 patients were included. OAC and PPI cotherapy were associated with a lower odds of total and major GIB; odds ratio (95% confidence interval) was 0.67 (0.62-0.74) for total and 0.68 (0.63-0.75) for major GIB, respectively. No differences in the GIB of PPI cotherapy were observed between Asians and non-Asians (P-for-difference, total GIB = .70, major GIB = .75, respectively). For all kinds of OAC except for edoxaban, PPI cotreatment was related to lower odds of GIB by 24-44%. The protective effect of PPI on total GIB was more significant in concurrent antiplatelets or nonsteroidal anti-inflammatory drug users and those with high bleeding risks: patients with previous GIB history, HAS-BLED ≥3 or underlying gastrointestinal diseases.

CONCLUSION In patients who receive OAC, PPI cotherapy is associated with a lower total and major GIB irrespective of ethnic group and OAC type, except for edoxaban. PPI cotherapy can be considered particularly in patients with high risk of GIB.