Saving AHRQ and the USPSTF

AcademyHealth CEO Aaron Carroll, MD recently submitted testimony to the House Appropriations Subcommittee on Labor, Health and Human Services, Education, and Related Agencies about the dire condition of the Agency for Healthcare Research and Quality (AHRQ), where I spent 4 years as a medical officer early in my career. Dr. Carroll points out the immense return on investment that AHRQ has provided over the years - for example, saving $7.7 billion in U.S. health care costs by reducing hospital-acquired infections from 2014 to 2017 on a budget of around $300 million per year - and its unique, irreplaceable function among federal health agencies:

NIH [National Institutes of Health] studies diseases. AHRQ studies how health care is delivered. These are different missions. NIH can tell us that a treatment works in a clinical trial. AHRQ tells us whether that treatment reaches patients in a rural hospital, whether it is implemented safely, what it costs, and whether a critical access hospital in a rural county can actually use it. No other federal agency performs this function. Eliminating AHRQ does not transfer these capabilities elsewhere. It simply ends them. 

Notably, Congress rejected HHS Secretary Robert F. Kennedy Jr.'s 2025 proposal to eliminate AHRQ. But Dr. Carroll observes that the Trump administration has effectively carried out this plan anyway, by laying off most of the agency's staff and the entire grants management division, crippling its ability to function as a funder of health services research:

AHRQ has not awarded a single new grant since April 2025. An estimated $80 million in FY25 appropriated research funding was allowed to expire unused—a pattern consistent with the Government Accountability Office’s ongoing impoundment investigation. In FY26, the agency has not funded any of the noncompetitive continuing grants it is statutorily obligated to pay. The FY27 congressional justification now explicitly states a policy of “no new grants,” ending AHRQ’s four-decade role as the nation’s primary funder of health services research—a decision Congress never authorized.  

Similarly, former New York City and Philadelphia Health Commissioner Thomas Farley, MD wrote today on his Substack that the U.S. Preventive Services Task Force is being "quietly strangl[ed]" by being deprived of AHRQ support staff, not being convened since March 2025, and not appointing replacements for 5 members whose terms expired on December 31. He cites the recent ACC/AHA dyslipidemia guidelines as an example of what fills the preventive care vacuum when the USPSTF (which wrote its own cholesterol guideline in 2022) is effectively silenced:

Are cholesterol tests for kids and coronary artery scans for adults now scientifically justified? Here’s the problem: I do not know. It takes more expertise and time than I have to sift through all the many complicated studies to figure that out. ... But I do know that (by my count) 12 of the 33 members of the writing committee and 17 of the 29 members of the review committee for the ACC/AHA guidelines have financial ties to biotech companies that are likely to make money from this testing and treatment. (None of the USPSTF members have these conflicts.) And I know this rule: if you’re wondering whether you need a new pair of shoes, don’t ask a shoe salesman.

The muddle about cholesterol testing, statin treatment and coronary artery scans is just one example of what we are losing from the USPSTF’s paralysis. ... Thanks to Kennedy, dozens of other important questions on the USPTF consideration list are also languishing. Each month that the Task Force is in deep freeze our ignorance accumulates. ... Surely we can afford to have a group of experts who are not motivated by profit guiding us on which medical services actually keep us healthy. With the USPSTF dead in the water, the war on science begins to feel like a war on us.

Nearly a year ago, I wrote a Medscape commentary that appealed to readers to "Save the USPSTF." The USPSTF still needs saving. So does AHRQ. So does the entire taxpayer-funded scientific apparatus at HHS devoted to keeping people healthy that RFK Jr. has wrecked.

AI health tools for the general public fall short

A 2025 American Family Physician editorial by Dr. Joel Selanikio discussed how artificial intelligence (AI) tools had accelerated an existing trend of “patients bypassing physicians to diagnose and treat themselves,” which began with over-the-counter drugs and online search engines. This direct-to-consumer health care approach received a boost in January with OpenAI’s launch of ChatGPT Health, which invites users to upload their medical records and health data from apps for personalized recommendations.

AI chatbots can provide helpful responses to health questions in several low-stakes contexts, as outlined in this handout from Dewey Labs: translating medical jargon, brainstorming possible causes of symptoms, summarizing research or test results, and preparing questions for an upcoming doctor’s visit. However, a recent study in Nature Medicine highlighted ChatGPT Health’s significant limitations in triaging patients with acute problems to appropriate levels of care.

Dr. Ashwin Ramaswamy and colleagues compared the chatbot’s responses to “60 clinician-authored vignettes across 21 clinical domains under 16 factorial conditions (960 total responses)” to triage levels assigned independently by three physicians: non-urgent, semi-urgent, urgent, and emergency. ChatGPT Health performed well in triaging semi-urgent and urgent clinical situations, but it over-triaged 65% of non-urgent situations and under-triaged 52% of true emergencies. For example, it recommended evaluation in 24 to 48 hours for patients with diabetic ketoacidosis and impending respiratory failure rather than sending them directly to the emergency department. Just as concerning, patients with suicidal ideation were less likely to receive crisis interventions when they had identified a method of self-harm than when they had no identified method:

The crisis guardrail finding may be the most consequential failure mode exhibited in the entire study. … A guardrail that fires for ‘haven’t thought through how I would do it’ but not for ‘thought about taking a lot of pills’ is not calibrated to clinical risk and users have no basis to anticipate when it will or will not fire. The capability to recognize mental health crises and connect users with crisis resources is a basic prerequisite for any consumer health platform. Our data show this prerequisite has not been reliably met.

In another study, three AI chatbots were provided with 10 detailed medical scenarios and tested on their ability to diagnose the condition and recommend appropriate management. In the United Kingdom, 1,298 adults were provided the scenarios and randomized to use one of the chatbots or a usual source of their choice (typically an online search engine). When researchers input the full scenarios, the chatbots diagnosed 95% of the conditions and correctly managed them 56% of the time. However, when intervention participants shared elements of the scenarios in live conversations, the chatbots performed much worse, correctly diagnosing 34% of the time and recommending appropriate management in 44%; this result is no better than control participants using a search engine. Researchers observed that participants often failed to provide enough information to make the diagnosis, and slight changes in symptom emphasis or wording of questions frequently led to dramatic differences in advice.

Bottom line: For patient-facing chatbots such as ChatGPT Health to diagnose and triage problems appropriately and safely, it isn’t enough to passively process the incomplete clinical data they are provided. They will need to get much better at asking the right questions to elicit information that patients may not be aware is relevant.

**

This post first appeared on the AFP Community Blog.