Abstract 226: Impact of Gastrointestinal Disorders among Nonvalvular Atrial Fibrillation Patients on Healthcare Resource Usage and Costs

Abstract

Objective: This study evaluated the incremental healthcare burden associated with GI disorders among nonvalvular atrial fibrillation (NVAF) patients. Methods: NVAF patients ≥18 years of age with continuous medical and prescription coverage were identified (1/1/2007-12/31/2009) from the MarketScan ® Commercial and Medicare Research Databases. Patients with at least 1 inpatient or 2 outpatient GI diagnoses within 12 months following any NVAF diagnosis were grouped into the GI cohort, with patients without any GI diagnosis during the entire study period grouped into the non-GI cohort. The date of first GI diagnosis after NVAF diagnosis and a random date within 12 months after NVAF diagnosis date were selected as the index dates for GI and non-GI patients, respectively. Baseline and follow-up periods were each 12 months. Of the overall cohorts, patients were matched (1:1) by key patient characteristics. The GI cohort was further categorized as having a prior history of GI disorders (chronic) or no GI disorders (non-chronic) during the baseline period. Healthcare usage and costs were evaluated and compared during the follow-up period for the matched cohorts. Results: Among NVAF patients 38% were diagnosed with GI disorders, of which 52% had no baseline history of GI disorders. Patient characteristics, including age (non-GI: 71.16 vs. GI: 74.1 years), gender (male: 60.22% vs. 51.9%), Charlson Comorbidity Index (1.34 vs. 2.35) and CHADS 2 score (1.45 vs. 1.76) significantly differed among cohorts. When patients were matched for key patient characteristics including comorbidity, NVAF patients with a GI disorder used healthcare resources (inpatient, outpatient and pharmacy) to a much greater extent than those who did not have a GI disorder, which was reflected in their significantly higher healthcare costs (Table). These results were independent of a prior history of GI disorders in the baseline period. NVAF patients with chronic GI disorders were the least likely to have used warfarin in the follow-up period (non-GI: 59.8%, no history of GI: 53.4%, chronic GI: 48.5%, p<0.001). Conclusions: Greater than a third of NVAF patients have a GI disorder, which is associated with much greater healthcare resource utilization. Warfarin usage appears to be lower among NVAF patients with GI disorders, although further study is needed.