Medicare Part B Spending on Macular Degeneration Treatments Associated With Manufacturer Payments to Ophthalmologists

Importance Age-related macular degeneration (ARMD) therapies aflibercept and ranibizumab are among the highest-cost Medicare Part B drugs, even though off-label use of lower-cost bevacizumab is clinically noninferior. Payments from manufacturers of these ARMD therapies to ophthalmologists are hypothesized to be factors in ophthalmologists' therapeutic choice, controlling for ophthalmologist and patient characteristics. Objective To assess the association between manufacturer payments to ophthalmologists and choice of ARMD treatment as well as to identify ophthalmologist-level characteristics associated with prescribing lower-cost ARMD therapies. Design, Setting, and Participants This retrospective cross-sectional study of longitudinal (2013-2019) Medicare Part B data was conducted from December 2021 to December 2022. Ophthalmologists prescribing aflibercept (manufactured by Regeneron Pharmaceuticals Inc), rabinizumab, or bevacizumab (both manufactured by Genentech Inc) for ARMD treatment of Medicare Part B beneficiaries were included. Data on manufacturer payments to ophthalmologists were obtained from the Open Payments database. Main Outcomes and Measures The primary outcome was the percentage of bevacizumab prescribed by ophthalmologists among all ARMD therapies. Regression analysis assessed variation in bevacizumab prescribing by acceptance of manufacturer payments as well as by ophthalmologist and patient characteristics. Ophthalmologist characteristics were duration of practice and Medicare Administrative Contractor region, and patient characteristics were aggregated at the ophthalmologist level and included mean beneficiary age, percentage of dual-eligible beneficiaries, mean beneficiary risk score, and percentage of White beneficiaries. Savings were estimated by projecting the change in bevacizumab use had ophthalmologists not accepted manufacturer payments, controlling for all ophthalmologist and patient characteristics and comparing with observed use and costs. Results A total of 21 584 ophthalmologists (18 489 males [85.7%]) were included. Ophthalmologists who accepted manufacturer payments were significantly less likely to prescribe bevacizumab (28.0% [95% CI, 24.6%-42.5%] of patients) compared with those who did not accept manufacturer payments (45.8% [95% CI, 44.5%-47.1%]). Ophthalmologists who saw dual-eligible beneficiaries had greater bevacizumab prescribing (50.0% [95% CI, 40.6%-68.3%] in the highest quartile vs 36.1% [95% CI, 33.5%-38.8%] in the lowest quartile; β coefficient, 0.139; P < .001), while those who saw patients with higher mean beneficiary risk scores had lower bevacizumab use (38.0% [95% CI, 23.7%-44.1%] in the highest quartile vs 48.2% [95% CI, 45.5%-50.8%] in the lowest quartile; β coefficient, −0.102, P < .001). Had ophthalmologists who accepted manufacturer payments prescribed ARMD drugs as those who did not accept payments, Medicare spending on these treatments would have been $642 779 703.08 lower from 2013 to 2019, a 2.0% savings. Conclusions and Relevance Results of this cross-sectional study suggest that drug manufacturer payments to ophthalmologists were associated with selection of higher-cost therapies for ARMD, which is a factor in increased Medicare and patient spending. Development of manufacturer payment models that encourage ophthalmologists to choose lower-cost therapies are needed.