Medicare Pilot Program Faces AI-Induced Delays and Concerns

Introduction

Newly released government records indicate that a Medicare pilot program employing artificial intelligence (AI) for prior authorization reviews has caused extended treatment delays and raised concerns about insufficient testing. These documents emerged due to a Freedom of Information Act (FOIA) lawsuit by the Electronic Frontier Foundation (EFF). They provide an in-depth look at the Wasteful and Inappropriate Service Reduction (WISeR) model, launched in January by Medicare with AI-assisted reviews for certain medical procedures.

Concerns and Findings

EFF is pressing the Centers for Medicare & Medicaid Services (CMS) for transparency on AI use in Medicare. According to Michael Ryan, a finance expert, the integration of AI into prior authorization without appropriate testing has created bottlenecks, delaying treatments that should take 72 hours. A request can sit unanswered for weeks, diminishing administrative efficiency and delaying patient care.
Newsweek reached out to CMS for comment.

Operational Challenges of WISeR

WISeR is a significant federal attempt to incorporate AI in Medicare administration, operational in six states, including Texas and Arizona, among others. It covers 13 medical services targeted to reduce fraud and misuse. Providers must secure authorization via AI-driven platforms operated by contractors.
Based on documents from the EFF lawsuit, WISeR has experienced operational issues. One instance involved a request pending for 83 days, in addition to inappropriate denials and reports of patient harm. The systems were not properly tested pre-launch, leading to delays beyond the promised 72-hour timeline.

System Testing and Transparency

EFF highlights concerns regarding the inadequacy of AI system testing before deployment. Questions remain about the thoroughness of safeguards, as AI-driven authorizations risk unwarranted and discriminatory delays or denials. Even though CMS states a qualified clinician reviews all denials, research implies AI recommendations can sway human decisions.

CMS has pledged to update WISeR based on feedback from doctors and patients. The agency emphasizes its commitment to timeliness, accuracy, and transparency, promising corrective actions to meet these objectives.

Previous Reports and Future Implications

Records uphold complaints from doctors and patients since the pilot’s launch. A June investigation noted patient confusion and increased travel demands due to new prior authorization requirements. Providers criticized the rollout, citing administrative challenges.
Kevin Thompson, CEO of 9i Capital Group, expressed concern about treatment delays due to longer approval times. Despite issues, CMS plans to expand WISeR as a cost-saving tool.

EFF queries CMS on technology evaluation pre-launch, focusing on accuracy, bias, and other factors. The group insists on transparency, arguing AI-generated suggestions influence human reviews.

Future of AI in Medicare

Michael Ryan believes the lawsuit won’t lead to Medicare abandoning AI. WISeR runs through 2031, with CMS envisioning technology-assisted authorization in Medicare’s future. However, proving consistent safety and efficiency before expansion is essential.

For inquiries on this report, contact Newsweek editors Jason Lemon and Anthony Murray.