Background: To investigate the clinical relevance of two different preservative formulations, we compared 1-year incidence rates of additional coding of dry eye, ocular infection, or ocular surface disease (either dry eye or ocular infection) in open-angle glaucoma and ocular hypertension patients newly treated with latanoprost with benzalkonium chloride (BAK) or with travoprost-Z with SofZia. Methods. This was a retrospective study of three U.S.-based patient-centric medical/pharmacy claims databases (MedStat, PharMetrics, i3-Ingenix). Patients were eligible if they filled a prescription for latanoprost or travoprost-Z between October 2006 and Q2 2008 (prescription date = index date) AND were continuously enrolled 6 months prior through 12 months after the index date AND had any open-angle glaucoma or ocular hypertension diagnosis within 90 days prior to the index date AND did not have an ocular surface disease diagnosis during the 180 days prior to the index date AND if they had not had a prescription for the index agent in the 180 days prior to the index date. Time to incidence of new coding for ocular surface disease in the first year post-index was estimated with a composite endpoint: diagnosis of dry eye or ocular infection by ICD-9-CM or Current Procedural Terminology code OR by prescription for cyclosporine ophthalmic emulsion or ocular antibiotics. Results: In all, 15,933 patients were treated with latanoprost and 7670 with travoprost-Z. Over 1 year, 4.3% of latanoprost and 4.5% of travoprost-Z patients were identified with dry eye (p = 0.28), and 10.9% and 11.1%, respectively, were identified with an ocular infection (p = 0.79). The 1-year incidence of new coding for ocular surface disease also was similar across treatments (13.9% vs 14.3%, respectively; p = 0.48). Conclusions: The retrospective analysis of three large prescription databases revealed that open-angle glaucoma and ocular hypertension patients newly treated with latanoprost preserved with BAK or travoprost-Z preserved with SofZia did not differ statistically in rates of dry eye, ocular infection, or ocular surface disease (either dry eye or ocular infection) during the first year post-index. Claims-based analyses are limited by nonrandomization and the inability to account for over-the-counter use or samples.
ASJC Scopus subject areas