Demographic predictors of resource utilization for bleeding peptic ulcer disease: the Ontario GI Bleed Study.

Abstract

Peptic ulcer disease (PUD) and its complications impose a substantial burden on health care resources. To help identify subpopulations in which preventative measures might achieve maximal cost savings, the authors studied the variation in resource utilization and cost for management of bleeding PUD among demographic subgroups. Resource utilization profiles and direct medical cost estimates were generated for consecutive admissions for bleeding PUD at four hospitals in southern Ontario via chart review and adaptation of an administrative cost database. Multiple linear regression models were developed to identify independent demographic predictors of direct medical case cost and hospital length of stay (LOS). Among 158 admissions, the average LOS and case costs were 5.73 days and $2,953 (Canadian) respectively. Age, comorbid illness, nonsteroidal anti-inflammatory drug use, and the absence of prior PUD or upper gastrointestinal hemorrhage were associated with higher cost in univariate analysis, whereas increasing age and comorbidity predicted LOS. Only age and absence of prior PUD persisted as independent predictors of direct medical cost and LOS in a stepwise multiple linear regression. Costs for managing bleeding PUD vary substantially among demographic subgroups. More careful stratification of treatment costs is needed when economic models of interventions to prevent or to treat PUD are applied to specific subpopulations.