Cost-Effectiveness of Treatments for Diabetic Macular Edema: Simulated Bevacizumab-First Step Therapy Versus Real-World Practice

Abstract

Purpose: To compare the incremental cost-effectiveness of a clinical trial-simulated step-therapy versus real-world treatment for diabetic macular edema (DME). Methods: A theoretical Markov model (follow-up of 2 years and lifetime of 17 years) from the 2025 US societal perspective was used to compare the costs and cost-effectiveness between bevacizumab-first (Protocol AC) and real-world regimens from the Vestrum Health database. The modeling used mean characteristics from a reference case and analyzed low- and high-cost scenarios, total societal costs from formal and informal healthcare and non-healthcare sectors, and differences in utility (visual acuity outcomes) between arms. Results: Protocol AC bevacizumab-first in the reference case was 14% more expensive at 2 years, with a total adjusted societal cost of $69 850 versus $61 304 for real-world treatment. Although visual acuity gains were higher with Protocol AC, the incremental cost-utility ratio (ICUR) was $105 335/quality-adjusted life years (QALY) at 2 years and $151 032/QALY over 17 years, higher than most societal willingness-to-pay thresholds. In the low-cost scenario, Protocol AC was neither cost-saving nor cost-effective at 2 years (ICUR $82 283/QALY) but was cost-effective over 17 years (ICUR $591/QALY). In the high-cost scenario, Protocol AC was not cost-effective at 2 years (ICUR $219 420/QALY) or 17 years (ICUR $207 589/QALY). Probability sensitivity analysis showed that Protocol AC was more expensive in 87% of modeled scenarios and not cost-effective in 76%. Conclusions: Compared with real-world treatment, protocol AC bevacizumab-first treatment for DME was generally not cost-saving. Although better vision outcomes were achieved with bevacizumab-first, the protocol was generally not cost-effective due to greater treatment burdens.