New Medicare AI model linked to inappropriate care denials

For beneficiaries enrolled in Original Medicare (OM), a new Artificial Intelligence (AI) model launched by the Centers for Medicare & Medicaid Services (CMS) is reportedly leading to inappropriate delays and denials of care. The program, known as the Wasteful and Inappropriate Service Reduction (WISeR) Model, began in January of this year in six states and focuses on prior authorization for certain medical procedures [Medicare Rights Center].

The WISeR Model is currently active in New Jersey, Ohio, Oklahoma, Texas, Arizona, and Washington. It covers specific procedures, such as some pain injections, spinal surgeries, and nerve treatments [Medicare Rights Center]. While Original Medicare has historically had fewer prior authorization requirements compared to Medicare Advantage plans, this new AI system marks a shift for a limited set of services [Medicare Rights Center].

Growing Concerns Over Delays

Reports indicate that the implementation of WISeR has increased wait times for beneficiaries needing procedures, forcing them to navigate through denial and appeal processes [Medicare Rights Center]. This has raised concerns among consumer advocates and even prompted previous reporting on delayed care for seniors in earlier stages of the program [Medicare Rights Center].

Potential for Expansion

Despite the current issues, there are reports that CMS may intend to expand the WISeR Model to include more types of treatments, potentially extending to oncology or cardiac care [Medicare Rights Center]. Such an expansion would impact a broader range of beneficiaries and services, making the current challenges even more pressing.

Flawed Design and Financial Incentives

A central concern with the WISeR Model's design is a perceived bias towards denying care. The AI companies responsible for making the initial coverage decisions reportedly receive greater financial incentives when they block access to medical services [Medicare Rights Center]. Although a qualified human clinician is meant to review AI decisions, studies suggest that these clinicians can be influenced by initial denials made by AI systems [Medicare Rights Center].

Many beneficiaries who receive an initial denial may not fully pursue the appeals process, meaning inappropriate denials often go unchallenged. This system raises questions about whether the model is truly serving its stated purpose of reducing wasteful services or if it's inadvertently creating barriers to necessary medical care for those who rely on Original Medicare [Medicare Rights Center].