Some scoring systems may predict major bleeds during DOAC treatment, but more work is needed

Direct oral anticoagulants (DOACs) are increasingly prescribed, in many cases in preference to warfarin treatment. A range of predictive scoring systems have been in use for some time to quantify (to some extent) the risk of serious bleeding associated with warfarin treatment, but evidence for use of this approach has been lacking for DOACs to date. Researchers have used a retrospective analysis of patients presenting to hospital from 2015 to 2018: patients were dichotomized into a bleed group (adults on a DOAC for atrial fibrillation and/or VTE who presented with a major bleed) and a non-bleed group consisting of patients randomly selected from those who presented to the ED with a history of DOAC use without active or prior history of bleeding. Patients in the non-bleed group were matched in a 2:1 fashion with patients in the bleed group based on date of ED presentation, indication, sex, and DOAC used. Patients who were <18 years old, pregnant, or presented with bleeding caused by other anticoagulants or without identifiable anticoagulant exposure were excluded. Over the three years, 4827 patients presented to the ED with a history of DOAC use. Of these patients, 42 (0.87%) met inclusion criteria for the bleed group, and 84 patients were matched into the non-bleed group. Overall, 126 patients were included in the analyses. The HAS-BLED, HEMORR₂HAGES, RIETE, and CHEST scores were found to have sufficient diagnostic accuracy for predicting risk of major bleeding.

Those in the bleed group had statistically significant higher rates of cirrhosis, anemia, and peripheral vascular disease in their past medical history (P = 0.042, 0.03, and 0.02, respectively). There were also higher rates of aspirin use (54.2% vs 32.1%; P = 0.045) and smoking (35.7% vs 16.7%; P = 0.030) in the bleed group patients. In addition to this, patients in the bleeding group had a significantly higher risk of fall per the Johns Hopkins fall risk score (P < 0.001). Other differences in characteristics included an increased rate of Acute Kidney Injury (AKI) in the bleed group (40.5% vs 14.3%; P = 0.002) and an increased number of patients with improperly high DOAC doses (16.7% vs 4.8%; P = 0.035).