An adherence based cost-consequence model comparing bimatoprost 0.01% to bimatoprost 0.03%

Objectives: Estimate the long-term direct medical costs and clinical consequences of improved adherence with bimatoprost 0.01% compared to bimatoprost 0.03% in the treatment of glaucoma. Methods: A cost-consequence model was constructed from the perspective of a US healthcare payer. The model structure included three adherence levels (high, moderate, low) and four mean deviation (MD) defined health states (mild, moderate, severe glaucoma, blindness) for each adherence level. Clinical efficacy in terms of IOP reduction was obtained from the randomized controlled trial comparing bimatoprost 0.01% with bimatoprost 0.03%. Medication adherence was based on observed 12 month rates from an analysis of a nationally representative pharmacy claims database. Patients with high, moderate and low adherence were assumed to receive 100%, 50% and 0% of the IOP reduction observed in the clinical trial, respectively. Each 1 mmHg reduction in IOP was assumed to result in a 10% reduction in the risk of glaucoma progression. Worse glaucoma severity health states were associated with higher medical resource costs. Outcome measures were total costs, proportion of patients who progress and who become blind, and years of blindness. Deterministic sensitivity analyses were performed on uncertain model parameters. Results: The percentage of patients progressing, becoming blind, and the time spent blind slightly favored bimatoprost 0.01%. Improved adherence with bimatoprost 0.01% led to higher costs in the first 2 years; however, starting in year 3 bimatoprost 0.01% became less costly compared to bimatoprost 0.03% with a total reduction in costs reaching US$3433 over a lifetime time horizon. Deterministic sensitivity analyses demonstrated that results were robust, with the majority of analyses favoring bimatoprost 0.01%. Application of 1 year adherence and efficacy over the long term are limitations. Conclusions: Modeling the effect of greater medication adherence with bimatoprost 0.01% compared with bimatoprost 0.03% suggests that differences may result in improved economic and patient outcomes.