Will AI fix prior authorization—or make it worse?

Editor's take

The Centers for Medicare & Medicaid Services (CMS) is testing an AI system to expedite prior authorization decisions, aiming to streamline the process for healthcare providers and patients. This initiative directly addresses the administrative burden and delays that currently plague prior authorization, a system often criticized for its inefficiency and impact on patient care access. By potentially reducing turnaround times for approvals, CMS seeks to improve operational efficiency within Medicare and Medicaid.

The success of this pilot program carries significant implications for the broader adoption of AI in healthcare administration. If effective, it could set a precedent for other payers and government programs, potentially reshaping how medical necessity is verified and claims are processed. The stakes are high, as a positive outcome could unlock substantial cost savings and improve patient experience, while a failure might reinforce skepticism about AI's readiness for such critical decision-making roles and prolong existing frustrations.

Future developments to monitor include the specific AI models and data sets employed by CMS, the measurable impact on authorization approval rates and processing times compared to manual reviews, and the extent of provider adoption and feedback. Transparency regarding the AI's decision-making logic will be crucial for building trust and ensuring accountability. Continued observation of any unintended consequences, such as algorithmic bias or increased denial rates for complex cases, will be essential in evaluating the program's long-term viability.