The Centers for Medicare & Medicaid Services recently announced a pilot program that will use artificial intelligence and machine learning. This program will be used, among other things, to oversee Medicare prior authorization procedures.
Prior authorization, as used today, warrants increased oversight. Polling shows that voters find these insurance processes difficult to manage and believe that delays and disruptions to care are a major problem. Prior authorization is an unnecessary tool used by health plans that does just that: it creates additional administrative burdens by requiring preemptive approval for a medication already approved by the doctor, causing delays and disruptions in crucial treatment plans.
As many patients know, finding effective treatment can be an exhausting, burdensome or even painful process. Once a patient sees a drug that works for them, facing additional bureaucratic hurdles that inhibit their access is demoralizing and potentially dangerous.
Often, patients cannot afford a delay — or outright denial — in access to the treatment they need. Innovative personalized therapies, which usually treat rare, chronic and life-threatening illnesses, are particularly affected by these policies. These patients already feel vulnerable, and prior authorization programs further subject them to the will of health plans looking to cut costs. This is not patient-centered healthcare, and it is un-American.
Prior authorization was likely well-intentioned. However, it now serves as a way for health plans, not doctors and providers with extensive knowledge, to have the final say in what treatments patients can access. This allows these coverage providers to prioritize their bottom lines over patient welfare.
CMS Administrator Mehmet Oz sees how this program can work against patients. Recently, he spoke to health plan providers, reemphasizing the need to reform and streamline prior authorization in Medicare Advantage programs. This follows the administration’s work to ensure health insurers’ voluntary commitments to reforming prior authorization over the summer.
AI can and should be used to improve patient access, including expediting the review of prior authorization claims. However, the use of AI for healthcare decisions also warrants caution. As with all AI use, guardrails should be put in place to ensure it is used wisely.
Physicians have expressed concern that AI may increase prior authorization denials. Thankfully, CMS is working to reduce the risk of baseless or excessive denials. When discussing the recent CMS announcement, spokesperson Alexx Pons highlighted that no request for coverage under Medicare will be denied without review by a “qualified human clinician” and that AI companies involved are “prohibited from compensation arrangements tied to denial rates.”
It is encouraging to see efforts to streamline prior authorization policies to ensure patients can access their medications. However, to ensure no patient is unjustly delayed or denied a treatment they need, the policy should be used sparingly, carefully, or done away with altogether.
Policymakers and lawmakers who want to protect patients should seriously examine this harmful policy. Policies that restrict patient access to save dollars come with the far higher cost of worsening health, time patients cannot get back, and even loss of life.
Other CMS policies — such as coverage with evidence development and step therapy — also restrict patient access to treatment. Access for All urges the Medicare administrator to assess the need for these policies and consider revising or eliminating their applications. Short of abandoning prior authorization, the agency must ensure that any AI use is implemented with extreme caution to protect patients from needless denials.
Access for All hopes it works to ensure no unnecessary bureaucratic hurdles hinder access to care.