The Centers for Medicare & Medicaid Services should consider banning artificial intelligence from being used to deny Medicare Advantage coverage pending a “systematic review,” a group of 49 congressional leaders is urging.
The group, led by Sen. Elizabeth Warren (D-MA) and Reps. Jerry Nadler (D-NY) and Judy Chu (D-CA), sent a joint letter to the regulator agency Tuesday, detailing their concerns and suggestions regarding the use of AI for informing or determining coverage denials for Medicare Advantage patients.
Skilled nursing providers have been sounding the alarm for years on Medicare Advantage coverage access, especially when informed by AI and other algorithms. Sector leaders have frequently noted that these methods can deny or prematurely end coverage for patients who need it to afford necessary long-term care.
Providers and consumer advocates both spoke out in favor of the lawmakers’ letter this week.
“LeadingAge’s nonprofit and mission driven members … have firsthand experience of Medicare Advantage (MA) plans’ inappropriate use of prior authorization to deny, shorten and limit MA enrollees’ access to medically necessary Medicare benefits,” wrote Katie Smith Sloan, president and CEO of LeadingAge, in a report issued by Nadler’s office Tuesday. “Implementation by [the] Centers for Medicare and Medicaid Services (CMS), which we fully support, would ensure MA plans fulfill their obligation to provide enrollees equitable access to Medicare services.”
Sector leaders have increasingly argued that MA plans’ methods of prior authorizations and denials of coverage are a threat to both residents’ access to coverage and providers’ access to vital Medicare funding.
“Skilled nursing providers have struggled, and continue to struggle, with Medicare Advantage coverage decisions that are based solely on AI algorithms,” Nisha Hammel, vice president of reimbursement policy and population health at the American Health Care Association told McKnight’s Long-Term Care News Wednesday.
“Relying on automated systems alone and making decisions without speaking to the person directly or consulting medical professionals can unnecessarily end coverage sooner than is medically necessary. We agree with members of Congress on their concern regarding this issue, and we applaud their request to increase oversight of AI use.”
Pushing for enforcement
In January, CMS announced regulations that will require MA plans to provide a rationale for any denials of coverage, and concurrently promised providers that the agency will use its “full array of tools” to enforce compliance.
The congressional leaders wrote that they appreciated CMS’ ongoing efforts, but shared worries that current rules and enforcement will not be enough to stop plans from denying necessary coverage.
“We believe CMS must be more proactive in monitoring plans’ use of AI and algorithm-driven tools,” the policymakers wrote. “MA plans cannot be allowed to side-step oversight by claiming that these tools are mere ‘guidance.’ Given that we do not know what inputs are used for the algorithms and AI tools currently being used, it is difficult to know the accuracy of the information they generate and whether the inputs comply with the regulations.”
The congressional leaders asked CMS to exercise strong control of AI tools — even as far as to actively limit their use for coverage decisions or prohibit them entirely.
“Absent a prohibition on the use of such tools altogether, CMS should limit their use until a systematic evaluation can be conducted on how these tools are impacting care,” they wrote. “We urge CMS to take a more affirmative role in prohibiting plans’ use of software or tools with discriminatory biases … These tools apply a generalized need for care to an individual beneficiary’s situation, resulting in generalizations instead of person-centered approaches to care, which is antithetical to the mission of the Medicare program.”
The lawmakers also asked CMS for clarification on what specific reporting will be required from MA plans, as well as on how CMS plans to differentiate between algorithms that sufficiently consider patients’ individual circumstances and those that do not.
