(VP049) EVALUATION OF REAL-WORLD MANAGEMENT STRATEGIES ADOPTED IN PATIENTS WITH ATRIAL FIBRILLATION ON ANTICOAGULANTS FOLLOWING AN EPISODE OF CLINICALLY SIGNIFICANT BLEEDING

Background: Systemic anticoagulation for stroke prevention in patients with atrial fibrillation (AF) carries inherent bleeding risks. As such, determining whether and when to resume anticoagulation after a significant bleed is a common dilemma. This may lead to modification or discontinuation of the thromboembolic prevention strategy (TPS) utilized. Left atrial appendage occlusion (LAAO) has emerged as an alternative TPS in these patients but referral patterns remain underreported. We therefore aimed to describe the clinical characteristics of AF patients discharged after a bleeding event, document real-life TPS management decisions and analyze their associated clinical outcomes.

METHODS AND RESULTS: We retrospectively reviewed the charts of anticoagulated AF patients admitted for bleeding between 2017 and 2019. The type of bleed, CHA2DS2VASC and HASBLED scores, TPS at discharge, readmissions for bleeding and all-cause mortality at 2 years were assessed. A total of 140 patients were included in the analysis with a mean age of 78.6 years. Median CHA2DS2VASC and HASBLED scores were 4 and 3 , respectively. A majority of patients were on DOACs at admission (79.3%) and gastrointestinal bleeding was the most frequent type (44.3%). Four discharge groups were defined: 75 (53.5%) patients had optimal anticoagulation (OA), 37 (26.4%) had a suboptimal anti-thrombotic regimen (SAR: low-dose DOAC without dose-reduction criteria or anti-platelet therapy), 10 (7.1%) were referred for LAAO and 18 (12.9%) left without any TPS. A greater proportion of women (24% vs. men 7%) were discharged with no TPS. All-cause mortality at 2-years was high (28.6%) but not statistically different between groups (p=0.71). Patients discharged with a TPS (OA/SAR/LAAO referral) were more likely to be readmitted for bleeding at 2 years (34% vs 0%, p=0.002), while those discharged without a TPS had higher rates of stroke (16.6% vs 1.4%, p=0.003). Use of a SAR yielded similar readmission rates for bleeding compared to resumption of OA (27% vs 34.7%, p=0.41). Of note, 71.4% of the adverse events in the LAAO group occurred during the waiting period before the procedure.

Conclusion: Bleeding while on anticoagulation for AF has significant prognostic implications. This real-life cohort reveals that clinicians frequently downgrade or discontinue long-term thromboembolic protection after a bleeding event in spite of current guideline recommendations to the contrary, and downgrading resulted in similar bleeding risk compared to OA. Given the risks of recurrent bleeding with pharmacological strategies and the rate of stroke without any protection, LAAO appears to be an underutilized alternative. However, its benefit may depend on timely access.