Non-vitamin K antagonist oral anticoagulants, proton pump inhibitors and gastrointestinal bleeds.

Abstract

To evaluate if proton pump inhibitor (PPI) treatment reduces the risk of upper gastrointestinal bleeding (UGIB) in patients with atrial fibrillation (AF) treated with non-vitamin K antagonist oral anticoagulants (NOACs). We used a common protocol, common data model approach to conduct a cohort study including patients with AF initiated on a NOAC in Stockholm, Denmark and the Netherlands from April 2011 until July 2018. The outcome of interest was a UGIB diagnosed in a secondary care inpatient setting. We used an inverse probability weighted (IPW) Poisson regression to calculate incidence rate ratios (IRRs), contrasting PPI use to no PPI use periods. In 164 290 NOAC users with AF, providing 272 570 years of follow-up and 39 938 years of PPI exposure, 806 patients suffered a UGIB. After IPW, PPI use was associated with lower UGIB rates (IRR: 0.75; 95% CI: 0.59 to 0.95). On an absolute scale, the protective effect was modest, and was found to be largest in high-risk patients, classified as age 75-84 years (number needed to treat for 1 year (NNTY): 787), age ≥85 years (NNTY: 667), HAS-BLED score ≥3 (NNTY: 378) or on concomitant antiplatelet therapy (NNTY: 373). Concomitant treatment with a PPI in NOAC-treated patients with AF is associated with a reduced risk of severe UGIB. This indicates that PPI cotreatment can be considered, in particular among the elderly patients, patients with a HAS-BLED score ≥3, and/or in patients on concomitant antiplatelet therapy.