In response to findings from a New York Times investigation on fraudulent billing for wound care in fee-for-service Medicare, researchers from the University of California, San Diego, Stanford, and the VA Palo Alto Health Care System sought to better understand similar reimbursement incentives in ophthalmology for the use of tissue in eye care. Their retrospective cohort study of Medicare Part B claims found a dramatic increase in both utilization and costs of sutureless ocular amniotic membrane grafts (AMGs) between 2011 and 2020. The authors explained that sutureless AMGs may have little clinically significant benefits in conditions such as bacterial keratitis, neurotrophic corneal ulcers, chemical injuries, and recurrent pterygia, though they are essential for managing vision-threatening chemical injuries and autoimmune inflammation.
“The financial benefits of AMGs for providers are more clear-cut than the clinical benefits for patients,” they wrote. “We…demonstrate that the increased costs that CMS has paid for ocular AMGs from 2011 to 2020 are largely driven by clinicians using AMGs for dry eye.”
The analysis, published recently in Ophthalmology, examined a nationally representative 20% sample of Medicare fee-for-service claims from 2011 to 2020. It included 49,180 AMG claims that were submitted for 23,894 unique patients by 6,416 unique providers. The mean patient age at first AMG was 74.6 years, and 68% of recipients were female.
The researchers found that the number of sutureless AMG claims increased from 319 in 2011 to 8,031 in 2020. Initially, corneal ulcers (39.8%) and corneal epithelial defects (12.2%) were the leading diagnoses associated with AMG placement, but by 2020, dry eye accounted for 44.2% of claims (from 7.2% in 2011). Optometrists, who submitted no AMG claims in 2011, accounted for 43.3% of all AMG claims by 2020. Dry eye diagnoses represented the majority (59%) of AMGs billed by optometrists, compared with a wider mix of indications among ophthalmologists.
Total submitted claims for AMGs increased from $3.6 million in 2011 to $95.6 million in 2020, and allowed charges rose from $1.4 million to $53.4 million over the same period. The authors noted, “These increases are notable because they occurred during a period in which an increasing portion of the Medicare population is covered under Medicare Advantage, rather than fee-for-service Medicare; the total costs of AMGs to the Medicare program as a whole are likely substantially higher.” For dry eye specifically, charges rose from $250,000 in 2011 to $41.4 million in 2020 (allowed charges increased from $110,000 to $24.1 million). A small number of providers accounted for a disproportionate share of claims: 1% of clinicians submitted 28.2% of all AMG claims.
Statistical models showed that patients were more likely to receive an AMG when treated by an optometrist rather than an ophthalmologist (hazard ratio [HR] = 1.16, P<.001). Patients who were seen by optometrists were also more likely to receive an AMG at their first dry eye diagnosis (odds ratio [OR] = 1.29, P<.001), and they were more likely to receive multiple AMGs (OR=1.25, P<.001). Costs per dry eye patient were substantially higher among providers who used AMGs, and the researchers observed a steeper increase in dry eye costs in optometry practices compared with ophthalmology practices.
They also found substantial discounts for AMGs, compared with the list prices for tissue that are reimbursed by CMS. They wrote, “By more closely aligning tissue reimbursement with the true acquisition costs for sutureless AMGs, policymakers could ensure continued coverage for a medical device with established benefits and encourage ongoing research into its adoption for novel clinical indications, while also limiting the costs associated with inappropriate use of AMGs in situations where they are unlikely to benefit patients.”
The researchers were not able to determine whether individual claims for AMGs were appropriate because analyses of insurance claim data do not allow for such observations. Although they investigated the possible effects that patients’ pre-enrollment histories might have on their findings, they were still limited by the care that patients received through Medicare.