OBJECTIVES: Non-vitamin K anticoagulants (NVKA) are replacing vitamin K anticoagulants due to simplified dosing and monitoring. However, patients on NVKAs are not exempt from bleeding risk. Potential NVKA drug interactions have been scarcely studied. The purpose of this study was to estimate the risk of bleeding in patients on NVKA and potentially interacting drugs.

METHODS: This was an observational cohort study using Medicare data. Participants receiving an NVKA from 2017—2020 were studied. The outcome was a clinical encounter associated with bleeding. A concomitant overlapping period while on NVKA was assessed for: non-steroidal anti-inflammatory drugs (NSAIDS); selective serotonin reuptake inhibitors (SSRI); proton pump inhibitors (PPI); platelet aggregation inhibitors; and glucocorticosteroids. Logistic regressions predicting either all bleeding or GI bleeding were conducted.

RESULTS: The study cohort of 102,531 had a mean age 77 years (SD=9.8) and 55% was female (N=56,671). There were 2,908 (2.8%) participants with a history of GI bleeding and 3,602 (3.5%) with a history of other bleeding. Risk for any bleeding was shown for clopidogrel (OR: 1.42, 95% CI: 1.34, 1.48), PPIs (OR: 1.60, 95% CI: 1.53, 1.67), prasugrel/ticagrelor (OR: 1.35, 95% CI: 1.16, 1.56), glucocorticosteroids (OR: 1.25, 95% CI: 1.18, 1.33), and SSRIs (OR: 1.10, 95% CI: 1.04, 1.16). GI bleeding risk was shown for clopidogrel (OR: 1.49, 95% CI: 1.39, 1.60), PPIs (OR: 2.53, 95% CI: 2.39, 2.60), prasugrel/ticagrelor (OR: 1.39, 95% CI: 1.14, 1.69), and glucocorticosteroids (OR: 1.31, 95% CI: 1.21, 1.43). PPI use was correlated with both NSAID (r=0.07, p-value<0.0001) and SSRI use (r=0.09, p-value<0.0001).

CONCLUSIONS: Persons on NVKA are at increased risk for overall bleeding and GI bleeding when taking anti-platelets, glucocorticosteroids, and SSRIs. Use of PPIs may be a confounding medication because it does not increase risk of bleeding but may be prescribed to patients with increased bleeding risk.