Cost-utility Analysis of Screening for Diabetic Retinopathy in Japan: A Probabilistic Markov Modeling Study

<i>Purpose</i>: To evaluate the cost-effectiveness for a screening interval longer than 1 year detecting diabetic retinopathy (DR) through the estimation of incremental costs per quality-adjusted life year (QALY) based on the best available clinical data in Japan. <i>Methods</i>: A Markov model with a probabilistic cohort analysis was framed to calculate incremental costs per QALY gained by implementing a screening program detecting DR in Japan. A 1-year cycle length and population size of 50,000 with a 50-year time horizon (age 40–90 years) was used. Best available clinical data from publications and national surveillance data was used, and a model was designed including current diagnosis and management of DR with corresponding visual outcomes. One-way and probabilistic sensitivity analyses were performed considering uncertainties in the parameters. <i>Results</i>: In the base-case analysis, the strategy with a screening program resulted in an incremental cost of 5,147 Japanese yen (¥; US$64.6) and incremental effectiveness of 0.0054 QALYs per person screened. The incremental cost-effectiveness ratio was ¥944,981 (US$11,857) per QALY. The simulation suggested that screening would result in a significant reduction in blindness in people aged 40 years or over (−16%). Sensitivity analyses suggested that in order to achieve both reductions in blindness and cost-effectiveness in Japan, the screening program should screen those aged 53–84 years, at intervals of 3 years or less. <i>Conclusions</i>: An eye screening program in Japan would be cost-effective in detecting DR and preventing blindness from DR, even allowing for the uncertainties in estimates of costs, utility, and current management of DR.