All-cause mortality and use of antithrombotics within 90 days of discharge in acutely ill medical patients
Conflicting evidence exists regarding predictors of and antithrombotic benefit on mortality in hospitalised acutely-ill medical patients. We compared mortality risk within 90 days post-discharge among medically ill patients who did and did not receive antithrombotics. This retrospective claims analysis included patients >/= 40 years with nonsurgical hospitalisation >/= 2 days between 2005 and 2009 using the HealthCore Integrated Research Database. Antithrombotic use (i.e. anticoagulants and antiplatelets) post-discharge was captured from pharmacy claims. All-cause mortality was determined from Social Security Death Index; cause of death was identified from National Death Index database. Kaplan-Meier survival curves were generated and hazard ratios (HR) for mortality risk were estimated using Cox proportional hazards models. Patients prescribed anticoagulants or antiplatelets post-discharge had lower risk of short-term mortality. For the anticoagulant model, the most significant predictors of mortality were malignant/benign neoplasms (hazard ratio [HR] 1.6, 95 % confidence interval [CI] 1.5-1.7), liver disease (HR 1.6, 95 % CI 1.5-1.7), anticoagulant omission (HR 1.6, 95 % CI 1.4-1.8), gastrointestinal or respiratory tract intubations (HR 1.5, 95 % CI 1.3-1.7), and blood dyscrasias (HR 1.4, 95 % CI 1.4-1.5). For the antiplatelet model, the most significant predictors of mortality were antiplatelet omission (HR 3.7, 95 % CI 3.3-4.1), liver disease (HR 1.6, 95 % CI 1.4-1.7), malignant/benign neoplasms (HR 1.6, 95 % CI 1.5-1.6), gastrointestinal or respiratory tract intubations (HR 1.5, 95 % CI 1.3-1.7), and blood dyscrasias (HR 1.4, 95 % CI 1.4-1.5). These mortality risk factors may guide future studies assessing potential benefits of antithrombotics in specific subsets of patients.
School of Medicine
General Internal Medicine