OBJECTIVE: Licensing and reimbursement decisions for age-related macular degeneration (AMD) therapies may be driven by different cost and outcome relationships across payers and thus any model should be flexible to accommodate these variances transparently.
METHODS: To accurately assess cost effectiveness of treatment as desired by European countries, a comprehensive model considering all direct costs is important. A Markov framework was used to model a cohort’s lifetime movement through visual acuity (VA) states of >20/40, 20/40 to >20/80, 20/80 to >20/200, 20/200 to >20/400, and >20/400. Presenting lesion subtypes (i.e., predominately classic, minimally classic, and occult) were considered. Unlike previous models, patients could experience more than one gain or loss of VA over their lifetimes. This comprehensive model was initially populated from a US perspective. Drug, procedure, monitoring, outpatient, vision rehabilitation, and adverse events (AEs) costs were obtained from standard US published sources. Expert interviews were conducted to determine AE treatment patterns and vision rehabilitation resource use. Relative risks and costs associated with effects associated with declining VA were extracted from a Medicare analysis. Efficacy data was taken from published literature for two years of treatment for all lesion subtypes. Efficacy beyond two years was extrapolated from clinical trials. Utilities were taken from similar published sources as previous models.
RESULTS: Results are expressed as vision years, quality-adjusted life years (QALYs), drug, treatment, AE, and other costs as well as incremental cost per vision year and QALY gained. Outcomes were discounted 3% per annum. This model is more robust than previous models as patient movement is not limited and represents natural disease progression over a lifetime, costs attributed to declining VA are included, and all lesion subtypes have been analyzed.
CONCLUSIONS: Thus, cost-effectiveness of treatment of AMD with respect to differing licensing and reimbursement is more accurately analyzed.
