OBJECTIVES: Urinary incontinence (UI) is associated with adverse physical, psychological and economic outcomes effecting ~13M Americans. UI is underreported and frequently untreated. First-line therapy includes agents decreasing incontinence frequency. This study characterizes utilization associated with tolterodine immediate release (TOL) versus oxybutynin (OXY). METHODS: Drug markers associated with UI management in the Merck-Medco pharmacy claims database (N>=65M) were used to construct a continuously benefit-eligible, new therapy cohort from 9/1/99-8/31/00. Utilization metrics were evaluated through 8/31/01. All patients were followed for 12 months post-therapy initiation. Chronic disease scores were used to estimate patient chronic disease burden via drug markers during the 6 months preceding UI initiation. Propensity scoring via logistic regression was used to control for possible selection bias, and to create a matched cohort (N=36,142) based on a minimum difference in propensity scores. Persistence was defined as any patient with drug supply on hand at least one day in the 12th month following the index prescription. Compliance measured the percentage of time between the first and last fills that the patient had drug supply "on hand". RESULTS: Mean age was 62 years; 65% were female. Eight percent of the cohort had drug markers for diabetes (OHA or insulin) during the 6 months preceding UI therapy. Twelve-month persistence was higher for TOL patients than OXY patients (24.2% vs 16.8%, p<.0001). TOL patients were more compliant than OXY patients (mean=71.0% VS 64.3%, p<.0001). Compliance and persistence differences were slightly more pronounced in diabetics; TOL compliance mean was 75.9% vs OXY 65.3% (p<.0001), while 30.8% of TOL diabetics were persistent, compared to 19.3% of OXY diabetics (p<.0001). CONCLUSIONS: TOL had higher persistence and compliance rates than OXY. These improvements in compliance and persistence differences were slightly more pronounced in diabetics. This data suggests that differences exist between pharmaceutical agents used to manage UI.