Abstract
Abstract: :
Purpose: This study evaluated cost–effectiveness (CE) among ocular prostaglandins (PGs) from a health plan perspective using a decision–analytic framework. Methods: Data inputs were derived from a population–based, retrospective cohort analysis using the Constella managed care database. Patients ≥40 years of age who initiated therapy with bimatoprost 2.5 ml (BIM_2.5) or 5 ml (BIM_5), latanoprost 2.5 ml (LAT), or travoprost 2.5 ml (TRA) between 4/01 and 12/02 were included. Eligible patients had no ocular hypotensive therapy and were continuously enrolled 180 days prior to the start. Success was defined in terms of persistency, continuous time on initial therapy for at least 180 days. Patients were deemed persistent until they either discontinued or changed the initial therapy and were censored upon termination of insurance coverage or reaching the study end (12/30/02). Cox regression analyses were adjusted for sex and age. First–year treatment (medical and pharmacy) costs were assigned based on whether patients were persistent, discontinued, or changed therapy. The CE ratio for each therapy was the mean first–year treatment cost divided by the persistency rate and provides an estimate of the average cost per treatment success. Results: In all, 3096 patients met the inclusion criteria (BIM_2.5, n=236; BIM_5, n=398; LAT, n=1791; TRA, n=671). Persistency rates were: BIM_2.5, 25%; BIM_5.0, 30%; LAT, 36%; TRA, 25%. First–year treatment costs were: BIM_2.5, $791; BIM_5.0, $958; LAT, $742; TRA, $707. CE ratios were: BIM_2.5, $3126; BIM_5.0, $3183; LAT, $2078; TRA, $2862. For each additional treatment success, LAT had the lowest additional cost ($318). A sensitivity analysis revealed that these findings were generally stable to changes in assumptions, including requirements for defining persistency success. Conclusions: This is the first study to evaluate CE among ocular PGs using persistency to define treatment success. Latanoprost demonstrated greater CE than either bimatoprost or travoprost.
Keywords: clinical (human) or epidemiologic studies: health care delivery/economics/manpower • clinical (human) or epidemiologic studies: biostatistics/epidemiology methodology