Artificial intelligence (AI)-powered decision-making tools that are used in health care utilization management (UM) are increasingly used by providers and plans to automate the medical review and prior authorization processes, direct post-acute care, and make determinations concerning admission and discharge planning. The Center for Medicare Advocacy (the Center) has released a report, authored by health policy intern Lyla Saxena, focusing on AI-powered decision-making tools used to make Medicare coverage decisions and their impact on beneficiaries.
In the Center’s experience, AI-powered decision-making tools may prompt providers and plans to make decisions about the authorization or continuation of care that are more restrictive than Medicare coverage guidelines, including denials that violate the coverage standard for skilled care that was clarified by the Jimmo v. Sebelius class action settlement. While Medicare requires an individualized assessment of each beneficiary’s qualification for coverage in certain care settings, AI-tools offer recommended decisions that are based on general rules and previous patient experiences. Further, plans, providers, and beneficiaries often do not fully understand the scope of these tools’ development and use. This is partly due to the proprietary nature of UM-focused AI-powered decision-making tools, which prevents the public from seeing and understanding the tools and challenging their results. Among other things, the paper calls on the Medicare program to conduct better oversight of AI-powered decision-making tools, including requiring greater transparency surrounding their development and use.
Read or download the full Report at https://medicareadvocacy.org/wp-content/uploads/2022/01/AI-Tools-In-Medicare.pdf