The Trump administration launched the WISeR pilot in January to use artificial intelligence and machine learning to review prior-authorization requests for selected Medicare services in six states. The program covers roughly a dozen services, including some pain procedures, cervical fusions, wound treatments, and treatments for incontinence and impotence, and is scheduled to run through 2031. Federal documents obtained by the Electronic Frontier Foundation describe a rushed rollout: Innovaccer asked for a delay and temporarily configured its system to approve every request, while Zyter reported data problems involving Medicare Parts A and B. Providers also reported unexplained denials, long waits, poor communication, and delayed care. In one March 30 report, Virtix denied 3,233 of 6,096 reviewed requests, or 53 percent; at least one request remained pending for 83 days, despite a 72-hour target. Virtix was placed on a corrective action plan and later said its average turnaround time had fallen to 1.18 days for prior authorization. The article also examines WISeR’s payment model. CMS documents say contractors receive 25 percent of the regional benchmark cost of denied care, while the agency’s actuary warned that this creates an incentive to deny claims. Quality adjustments reduce that payment only modestly for poor performance, and relatively few denials are appealed, although many appealed decisions are overturned. Officials defended monitoring and possible contract termination, while lawmakers sought more documents and questioned the program’s legality and design. The Government Accountability Office said in May that proper procedures were not followed when the program was established. Despite the criticism, the administration reportedly plans to expand the system to areas including oncology, air ambulance transport, advanced imaging, cardiac procedures, implanted cardiac devices, and genetic and molecular tests.
