Predictors of bleeding in patients receiving direct oral anticoagulants.

Background: The predictors of clinically significant bleeding events (CSBE) associated with direct oral anticoagulants (DOAC) are poorly characterised in the literature.

Aim: To determine the incidence and predictors of CSBE in patients receiving DOAC.

Methods: Patients who received DOAC during admission to a general medical unit over a 2-year period were retrospectively studied. Following the index admission, patients were followed for 12 months or for the duration of treatment (if the latter was less than 12 months). The relevant data were obtained by review of medical records.

Results: A total of 203 patients was studied over a mean follow-up period of 293 (±81) days. The incidence of CSBE was 13.7 (95% confidence interval (CI): 9.5-21.1) per 100 person-years. Age ≥ 75 years (P = 0.01), concurrent use of antiplatelet medications (P = 0.02) and lower estimated creatinine clearance (CrCl) (P = 0.03) had a significant univariate association with CSBE. However, in the multivariate logistic regression, only concurrent use of antiplatelet medications remained significantly associated with CSBE (adjusted odds ratio (OR) 3.6; 95% CI: 1.4-9.6; P = 0.01). Concurrent use of antiplatelet medications was also independently associated with major bleeding events (MBE) (adjusted OR 4.9; 95% CI: 1.1-21.4; P = 0.04). Although 39 (19.2%) patients received antiplatelet medications, the indications for concurrent antiplatelet use complied with current guidelines in only 3 (7.7%) patients.

Conclusion: Caution should be exercised when prescribing antiplatelet medications with DOAC as this combination is a potential risk factor for both major and non-major clinically significant bleeding events. In most patients, the concurrent use of antiplatelet medications was discordant with the current consensus guidelines.