Potential interactions between SSRIs and DOACs: population-based cohort and case-crossover study.

Abstract

Bleeding is a side effect of direct oral anticoagulants (DOACs) and selective serotonin reuptake inhibitors (SSRIs). However, it is unknown whether their concomitant use would further exacerbate bleeding risk. To compare hazard of bleeding in patients with concomitant use of DOACs and SSRIs versus non-SSRI antidepressants. Population-based cohort and case-crossover study using primary care data from the UK Clinical Practice Research Datalink (CPRD) Aurum between 1/1/2011 and 29/3/2021. We used a cohort design to estimate hazard ratios (HRs) using propensity score weighting, comparing DOAC+SSRI and DOAC+non-SSRI users, and a 6-parameter model case-crossover design comparing odds of exposure to different drug initiation patterns for outcomes in hazard vs referent window within an individual to eliminate time-invariant confounding (confounding that do not change over time). There was no difference in bleeding risk in the cohort design (intracranial bleeding: HR1.16, 99% confidence interval [CI] 0.62-2.20; gastrointestinal bleeding: HR1.09, 99% CI 0.83-1.41; other bleeding: HR1.01, 99% CI 0.78-1.29). In the case-crossover design, we observed higher odds ratio (OR) of 1.64 (99%CI 1.14-2.35) for other bleeding associated with SSRI initiation while taking DOAC than SSRI monotherapy (OR1.06; 99% CI 1.01-1.11; p for Wald test=0.002), but greater odds ratio was not observed in DOAC users initiated non-SSRI (p for Wald test=0.83). We found no evidence of increased risk of intracranial and gastrointestinal bleeding during concomitant use of DOAC+SSRI in the cohort analysis. However, the case-crossover analysis suggested some evidence of a higher risk of other bleeding when initiating SSRIs (but not non-SSRIs) while taking DOACs.