Whether a health insurer denies or approves a medical procedure, there’s a good chance AI was involved in the review. Increasingly, insurers are using artificial intelligence for prior authorization, concurrent review and retrospective and utilization review — terms MedCity News recently defined in an overview of utilization review.
“AI is in all of them,” said Indiana University Bloomington law professor Jennifer D. Oliva, J.D., in an interview. She is referring to health insurers using AI in all forms of review and for multiple other purposes, as the National Association of Insurance Commissioners (NAIC) found last year. The Val Nolan Faculty Fellow at IU’s Maurer School of Law, Oliva is an expert in how health insurers use AI to boost income by limiting spending on patients’ health care. She wrote two in-depth review articles about this topic, one of which was published in the Indiana Law Journal.
In contrast, health insurers say their use of AI in all forms of treatment review is designed to find proof that care is medically necessary, Reuters reported in March. As businesses across industries adopt AI to cut costs, health insurers are increasingly using it before and after approving medical services, treatments and medications, according to a February report from Stanford University.
Stories to cover as AI use spreads
The growing use of AI creates plentiful opportunities for journalists to examine how the use of the technology affects patients and physicians, as well as how states are attempting to regulate it. In its survey, the NAIC found that 84% of health insurers reported using AI or machine learning across multiple product lines. Among respondents, 68% said they use AI for prior authorization.
While facing shrinking profit margins and higher medical costs this year, some of the nation’s largest health insurers told investors during first-quarter earnings calls they were accelerating the use of AI, as Casey Ross reported for STAT in February. UnitedHealth Group executives, for example, said they planned to use AI to cut $1 billion in costs this year.
STAT’s Pulitzer-finalist “Denied by AI” series examined how insurers used AI-driven algorithms to limit patient care while increasing profits. The series was a 2024 Pulitzer Prize finalist and a second-place winner in AHCJ’s 2023 Awards for Excellence in Health Care Journalism.
In an interview, Oliva said health insurers’ use of AI stretches back 10 to 15 years, adding that the practice of having humans monitor AI use varies among insurers. Many states and health plans require human oversight of AI systems, with some states mandating that reviews involve a medical professional. Health plans disclose these practices in various ways, in some cases only in small print at the end of documents, she said.
Story angle: Monitor a federal program on AI — and a related lawsuit
If you are seeking stories about insurers’ use of AI, consider covering the Wasteful and Inappropriate Service Reduction (WISeR) Model, a program the federal Centers for Medicare and Medicaid Services (CMS) started this year. The five-year program operates in six states — Arizona, New Jersey, Ohio, Oklahoma, Texas and Washington — to test whether AI and machine learning can streamline prior authorization and medical reviews for services vulnerable to fraud, waste and abuse, Becker’s Hospital Review reported.
Under the model, providers in participating states must obtain AI-backed prior authorization before delivering certain services to traditional Medicare patients, Healthcare Dive reported.
Providers reported denials for some care they said fell within coverage guidelines, Rebecca Adams reported for The Washington Post in March. Because different companies administer WISeR in each state, patient and provider experiences have varied. Some patients reported waiting longer than usual for such services as pain relief and minimally invasive care for spinal compression fractures, Adams wrote.
Critics raise transparency concerns
On March 25, the Electronic Frontier Foundation, a digital rights advocacy organization, filed a Freedom of Information Act lawsuit against CMS seeking details about how WISeR algorithms work.
“There is little information about how the AI algorithms used in WISeR work, including what training data they rely on,” an EFF news release said. “It remains unclear whether WISeR has any safeguards against systemic flaws such as algorithmic bias, privacy violations and wrongful denials of care.”
EFF also argues the model may incentivize denials because vendors share in cost savings tied to reduced utilization. The organization sued CMS seeking records about how the AI systems are trained, monitored and audited. CMS declined to comment on pending litigation.
Story angle: Follow state legislation
Journalists also could follow state legislation related to insurers’ use of AI, Oliva said. With little federal oversight of AI’s use, California, Colorado, Illinois and Utah have considered legislation to limit health insurers’ use of AI, she added.
In February, the Alabama Senate passed legislation requiring a physician or other qualified provider to be involved in AI-related prior authorization denials, Becker’s Health IT reported. If the House passes the bill and the governor signs it into law, it could become effective Oct. 1. In Florida, a bill requiring human review of all care denials passed the House but stalled in the Senate, the Palm Beach Post reported.
Story angle: Follow the impact of a CMS proposed rule
CMS on April 10 issued a proposed rule requiring payers to support electronic prior authorization for some medications and make decisions on requests within shorter timeframes, Healthcare Finance reported.
Under the rule, certain insurers would be required to make prior authorization decisions for covered outpatient drugs within 24 hours of receiving a request. This rule could take effect as soon as Oct. 1, 2027. Given the volume of requests insurers handle and the shorter proposed turnaround times journalists could examine whether insurers rely more heavily on AI to speed approvals and denials.
Expert sources to interview
In addition to Oliva, you could interview the following experts on health law and insurer’s use of AI:
- Erin Fuse Brown, J.D., MPH, of Brown University.
- Elizabeth McKuskey, J.D., of the Boston University School of Public Health.
- Michelle Mello, J.D., Ph.D., of Stanford University.
Resources
- Regulating Healthcare Coverage Algorithms – Indiana Law Journal
- The AI Arms Race in Health Insurance Utilization Review – Health Affairs
- How AI is leading to more prior authorization denials – American Medical Association
- Medicare’s AI experiment leads to delayed care for some seniors – Washington Post
- AI-powered prior authorization is coming to Medicare. Experts are concerned – Advisory Board
- In a financial pinch, major health insurers are turning to AI for help – STAT
- Is AI denying your insurance claim? It’s happening more than you think – Palm Beach Post
- Payers need to speed up prior authorization for drugs under proposed rule – Healthcare Finance
- CMS trials prior authorization for select Medicare services – AHCJ blog post
- Denied by AI – Link to a series by STAT reporters Casey Ross and Bob Herman
- STAT’s “Denied by AI” series a model of solid investigative journalism – AHCJ blog post
- 9 story ideas to cover now about AI-driven algorithm use in health care – AHCJ tip sheet
