Abstract
Purpose: :
To evaluate the comparative cost-effectiveness of topical prostaglandin agents for maintaining persistent therapy within the first 2 years of initiating therapy.
Methods: :
Data were derived from the Ingenix managed care database. Patients who had latanoprost (LAT), bimatoprost (BIM), or travoprost (TRAV) dispensed between 1/1/04-12/31/04 were screened for inclusion. Index agent=first agent filled; index date=fill date. Patients were excluded if they were <40 years old; were not continuously enrolled for 180 days before the index date; or had any ocular hypotensive dispensed or had no glaucoma diagnosis within 180 days before index date. Patient data were censored at the earliest of end of enrollment; end of study (12/31/2005); or upon adding/switching to a new agent. Cox regression (adjusted for age, gender, and recent diagnosis of preglaucoma/ocular hypertension) was used to compare the relative risk of discontinuation of the initial prostaglandin and to produce a survival [on therapy] plot over the first 720 treatment days for each of the 3 prostaglandins. The area under the survival curve was used to estimate expected days on therapy for each agent.
Results: :
9124 patients met all inclusion criteria (LAT, n=5816; BIM, n=1665; TRAV, n=1643). Compared to LAT, the relative risk of discontinuing the index prostaglandin over the first 2 years was 8.3% higher for BIM (p=0.016) and 24.4% higher for TRAV (p<0.001). Within the first 720 days after starting therapy, expected days of uninterrupted, continuous therapy were estimated as 245 for LAT, 226 for BIM, and 203 for TRAV. The mean drug cost (AWP) of therapy for this same period was estimated from actual claims as $301 (95% CI: $293-$309) for LAT, $364 (95% CI: $344-$384) for BIM, and $278 (95% CI: $263-$294) for TRAV. Compared to TRAV, the incremental cost effectiveness ratio of LAT was $0.56 and of BIM was $3.91 per additional day of uninterrupted therapy.
Conclusions: :
Patients generally do not stay on ocular prostaglandin therapy longer than 6 to 7 months before therapy interruption or discontinuation. Patients using LAT stayed on therapy longer than those using BIM or TRAV and at a substantially lower cost per additional day of therapy than BIM.
Keywords: clinical (human) or epidemiologic studies: health care delivery/economics/manpower • clinical (human) or epidemiologic studies: biostatistics/epidemiology methodology