Frail patients who fall and their risk on major bleeding and intracranial haemorrhage. Outcomes from the Fall and Syncope Registry

Background: Major bleeding, and intracranial bleeding specifically, are severe complications related to the use of anticoagulation. To what extent the risk for major bleeding is elevated among frail older people is not well known because they are underrepresented in the randomized clinical trials (RCTs). This study investigates the risk for major bleeding (MB) and intra cranial haemorrhage (ICH) in frail older people who fall. Methods: All patients 65 years and older visiting the Fall and Syncope Clinic, between November 2011 and January 2020, and underwent a MRI of the brain were eligible. Frailty was assessed with a Frailty Index, based on the accumulation of deficits model. Cerebral small vessel disease was described and evaluated as proposed in the position paper of Wardlaw and colleagues in 2013. Results: 479 patients were included in this analysis. Mean follow-up was 7 years per patient (ranging from 1 month to 8 years and 5 months). 368 patients (77%) were frail. A total of 81 patients used oral anticoagulation (OAC). 17 extracranial MB of which 3 were traumatic and 14 gastrointestinal, and 16 ICH occurred. There was a total of 603.4 treatment years with OAC, and 8 MBs occurred among patients on OAC (bleeding rate 1.32 per 100 treatment years), of which 2 ICHs (bleeding rate 0.33 per 100 treatment years). The risk for extracranial MB was increased by the use of antiplatelet agents (APA) (adjusted OR 6.9, CI 95% 1.2–38.3), and by the use of OAC (adjusted OR 9.8, CI 95% 1.7–56.1). The risk for ICH was only heightened by white matter hyperintensities (WMH) (adjusted OR 3.8, CI 95% 1.0-13.4). The use of APA (adjusted OR 0.9, CI 95% 0.3–3.3) or OAC (adjusted OR 0.6, CI 95% 0.1–3.3) did not elevate the risk for ICH. Conclusion: In contrast to common belief, frail patients on OAC with repeated falls show a comparable bleeding rate as in the large RCTs, and the use of OAC did not increase the risk for ICH. However, the number of MBs was low, and of ICHs very low, despite extensive follow-up in this registry.