Costlier AMD Prescriptions Linked to Manufacturer Payments to Ophthalmologists

Published on: 

A cross-sectional analysis of Medicare Part B data spending shows ophthalmologists who received manufacturer funding were less likely to prescribe bevacizumab.

Ophthalmologists who receive payments from drug manufacturers are more likely to select costlier therapies for age-related macular degeneration (AMD), according to new cross-sectional data.

A new study from a pair of health policy investigators showed an association between manufacturer payments to ophthalmologists and their reduced likelihood of prescribing bevacizumab (Avastin; Genentech) versus ranibizumab (Lucentis; Genentech) or aflibercept (Eylea; Regeneron)—the former option being significantly less costly on off-label use per 2020 averages.

Sean Dickson, JD, MPH, with the political advocacy and trade association AHIP, and Kately E. James, BA, conducted a retrospective cross-sectional study to observe links between ophthalmologist-received manufacturer payments and selection of treatments for AMD, as well as prescriber characteristics associated with the selection of lower-cost therapies for the common chronic retina condition.

The selected observation of the 3 anti-VEGF treatments was due to their equal recommendation of use by the American Academy of Ophthalmology (AAO) for the treatment of AMD, despite their significant disparities in cost. Per Dickson and James, average 2020 prices for off-label use were as follows for each agent:

  • Bevacizumab, $70.86 per Medicare claim
  • Ranibizumab, $333.55
  • Aflibercept, $921.65

“These cost differences have implications for both the Medicare program and beneficiaries, as these patients are liable for 20% of the treatment cost,” Dickson and James wrote. “Aflibercept was the second most expensive drug by total spending for the Medicare Part B program in 2020, with more than $3 billion in annual costs, while ranibizumab was the sixth most expensive drug by total spending, costing more than $1 billion; spending on aflibercept and ranibizumab together was more than 10% of the total Medicare Part B drug spending for 2020.”

The pair conducted their analysis using longitudinal Medicare Part B data from 2013 – 2019. Data regarding manufacturer payments to ophthalmologists were obtained via Open Payments database. Primary outcome was percentage of bevacizumab prescribed by ophthalmologists among all AMD therapies.

Investigators additionally identified ophthalmologist characteristics including duration of practice, Medicare Administrative Contractor region, and patient characteristics. They calculated savings through an estimated projected change in bevacizumab use had ophthalmologists not accepted manufacturer payments.

Their final analysis included 21,584 ophthalmologists. Those who accepted manufacturer payments were 38.9% less likely to prescribe bevacizumab for the treatment of AMD (28.0%; 95% CI, 24.6 – 42.5) versus ophthalmologists who did not receive payments (45.8%; 95% CI, 44.5 – 47.1).

Ophthalmologists treating dual-eligible beneficiaries were 50% likely to prescribe bevacizumab (50.0%; 95% CI, 40.6 – 68.3) in the highest quartile, versus 36.1% among those in the lowest quartile (95% CI, 33.5 – 38.8; P <.001).

Those who treated patients with greatest mean beneficiary risk scores reported lower bevacizumab use (38.0%; 95% CI, 23.7 – 44.1) in the highest quartile versus the lower quartile (48.2%; 95% CI, 45.5 – 50.8; P <.001).

The investigators’ estimated Medicare spending on AMD treatments had ophthalmologists who received manufacturer payments prescribed similarly to those who did not would have been more than $642 million total from 2013 – 2019, or 2.0% savings.

“Prior research reported an association between ophthalmologist use of lower-cost therapies and the race and ethnicity of patients, but we did not observe such an association,” Dickson and James noted. “Suggesting that an association exists between patient dual-eligibility status and race and ethnicity may have confounded prior analyses. The present study found that ophthalmologists with a greater percentage of dual-eligible beneficiaries in their patient population were significantly more likely to prescribe bevacizumab.”

Noting that ophthalmologists with longer practice duration were also significantly more likely to prescribe bevacizumab to patients with AMD, investigators hypothesized that the continued off-label use of bevacizumab—approved by the US Food and Drug Administration (FDA) prior to aflibercept and ranibizumab—may be often predating the introduction of the costlier anti-VEGF agents to the market.

The analysis is limited by differences in billing for off-label bevacizumab, the patient characterized observed in the analysis not being specific to those being treated for AMD, and the inability to account for practice-level outcomes that may influence an ophthalmologist’s prescribing strategy. However, the pair concluded their study was the first to estimate an increase in Medicare spending on AMD therapies associated with pharmaceutical manufacturer payments to ophthalmologists.

“Given that ophthalmologists appeared to gain greater profits when choosing higher-cost therapies, policymakers should counter this incentive by developing manufacturer payment models that encourage ophthalmologists to choose lower-cost therapies,” they wrote.


Dickson SR, James KE. Medicare Part B Spending on Macular Degeneration Treatments Associated With Manufacturer Payments to Ophthalmologists. JAMA Health Forum. 2023;4(9):e232951. doi:10.1001/jamahealthforum.2023.2951