Abstract WMP84: High Morbidity And Mortality Associated With Anticoagulation Prior To Intracerebral Hemorrhage: Propensity Score Analysis Of 10-year Real-world Data

Abstract

Introduction: Anticoagulation for prevention of cardioembolic stroke is safe and effective. However, there is a paucity of data on morbidity and mortality among non-traumatic intracerebral hemorrhage (ICH) patients with history of anticoagulant (AC) use. Methods: Using ICD-10 diagnosis/procedure, procedural terminology codes and medications, we identified index ICH events with and without prior AC use from deidentified pooled data; in a network of 50 healthcare organizations (Aug 2011-July 2021). Non adult (<18 years) and the presence of a prosthetic heart valve were excluded. AC and no-AC patients were propensity score (PS) matched for demographic, comorbidity and clinical variables. Match adequacy was assessed by standardized mean difference (SMD). Absolute Risk Difference (RD) and Risk Ratios (RR) with 95% Confidence Intervals (CI) are reported for morbidity [seizure, sepsis, intraventricular hemorrhage or external ventricular drain(IVH/EVD)] and mortality at 5 years post index ICH event in the PS-matched sample. Results: Among 193,600 patients with an index ICH event, 171,561 met the inclusion criteria, of whom, 62,578 (36.5%) had prior AC use. Significant pre-match differences in demographic and clinical parameters were observed between the AC and no AC groups (table). PS algorithm generated a 1:1 optimally matched sample (95% SMD reduction) of 21,808 AC and no AC ICH patients each, without significant covariate differences. In the matched sample, the risk of post-ICH seizure, sepsis, IVH/EVD, and mortality were significantly higher in the AC group. RR (CI) for seizure: 1.27(1.22-1.32), sepsis: 1.56(1.46-1.67), IVH/EVD: 1.78(1.66-1.92), mortality: RR (CI) 1.05 (1.02 - 1.09). RD of 5-year death between AC and no-AC groups was 1.3%, compared to a 4.1% RD of seizure, 3.2% RD of sepsis and a 3.7% RD of IVH/EVD. Conclusion: In Real World multicenter data, utilization of AC prior to ICH demonstrates significantly high long-term morbidity and mortality.