Physicians treatment decisions, patient persistence, and interruptions in the continuous use of prostaglandin therapy in glaucoma

Background: Uninterrupted use of ocular hypotensive medication by glaucoma patients is important to prevent vision loss, but medication persistence is poor. Efficacy and tolerability influence physicians decisions and patient persistence, and differences between medications may impact persistence. Objective: To examine differences in physicians decisions to continue, switch, or discontinue therapy across three prostaglandin analogs (PGAs) latanoprost, bimatoprost, and travoprost using claims data supplemented by evaluation of physicians charted therapeutic decisions. Methods: A year of pharmacy claims data for 6271 patients with a first (index) fill between 5/1/2001 and 11/30/2004 for PGA monotherapy were classified as persistent, switched, restarted, or discontinued with initial PGA use. An analysis of index therapy continuation during the first 2 years reflected chart reviews for 223 patients with PGA monotherapy as the index prescription. Results: Ten percent of patients had uninterrupted use of the initial PGA alone or in combination for a year. More than half (56) stopped and then restarted, 16 switched, and 19 discontinued the initial PGA. Patients using latanoprost were more likely to be persistent (11) compared to bimatoprost (9) or travoprost (5; p<0.0001 overall comparison). Overall, 68 of patients on latanoprost persisted or restarted after a gap compared to 61 for bimatoprost and 58 for travoprost (p<0.0001). Patient charts demonstrated a parallel pattern in physicians decisions to continue latanoprost (56), bimatoprost (45), and travoprost (40). Study limitations included the inability to establish causal links between variables, to account for sample use, or to document reasons for patient-driven changes in therapy. The study should be replicated in a more recent database including a larger population. Conclusions: Uninterrupted use of ocular hypotensive therapy for a full year is relatively rare. Differences in physicians decisions to continue, switch, or discontinue PGAs were observed in claims data, and parallel trends were observed in patient medical records.