Association of Direct Oral Anticoagulant-Proton Pump Inhibitor Co-Therapy with Adverse Outcomes: A Population-Based Cohort Study

BACKGROUND: Direct-acting oral anticoagulants (DOACs) are first-line therapy for atrial fibrillation and for venous thromboembolism prevention and treatment. In addition, proton pump inhibitors (PPIs) are recommended by the ACC Expert Consensus for patients receiving DOAC. OBJECTIVE: To determine the risk of clinically important thromboembolic and bleeding events and death in patients concomitantly prescribed DOACs and PPIs compared with DOAC alone. METHODS: We undertook a population-based cohort study of Ontario residents aged 66 years or older who were newly dispensed a DOAC from 2009 to 2020. The primary outcome was a composite of clinically relevant bleeding, thrombotic events, and all-cause death. We then used a time-dependent Cox regression model to estimate the adjusted hazard of outcomes during time periods where patients were exposed vs not exposed to PPI. RESULTS: We included 283 771 new DOAC users (mean age 78.3 years, 49.4% female). The age/sex-standardized outcome event rate of the composite outcome was higher in the DOAC-PPI cohort [20.2 (95% CI 20.0-20.5) per 100 person-years] than in the DOAC-alone cohort [15.2 (95% CI 15.1-15.4) per 100 person-years]. In time-dependent Cox regression analyses examining time periods during DOAC-PPI co-therapy vs DOAC alone, risks were elevated for all outcomes during periods of PPI co-therapy, suggesting residual confounding. In a secondary analysis of patients receiving DOAC-PPI co-therapy (n = 115 493), periods of DOAC-PPI co-therapy were associated with a greater hazard for death (HR 2.24, 95% CI 2.14-2.35), but a lower risk for both thrombosis (HR 0.93, 95% CI 0.89-0.96) and clinically relevant bleeding (HR 0.79, 95% CI 0.76-0.81). CONCLUSION AND RELEVANCE: Our cohort study of older adults suggests that DOAC-PPI co-therapy is associated with decreased upper GI bleeding but increased mortality. Since the implications are major and may be due to residual confounding, we recommend that the findings be verified in randomized trials before clinical application.