The MAHA children’s health report mis-cited our research. That’s sloppy

P.T. Barnum, American showman and circus master, famously said, “There’s no such thing as bad publicity.”

Or did he?

There is no definitive evidence that Barnum was the actual source.

But why did we ask? Well, that just came naturally to us, two old medical academics.  Thanks to our training and our work as writers, researchers, and teachers, we know that it is good to think twice about claims and their veracity. This matters less when the stakes are low, like the P.T. Barnum quip, but it matters a lot when the stakes are high — for example, important health policy claims that might affect millions of people, such as those in the recent MAHA Report. Citations bolster research reliability because they put claims in the context of prior work and facilitate verification by pointing to sources. And AI agrees. When we asked CoPilot “Do citations matter in writing about science?” it said, yes, “they’re the backbone of academic integrity—without them, science wouldn’t progress in a structured, credible way!” Of course, asking AI is hardly the ideal way to check citations (spoiler alert).

In our little academic world, being cited is gratifying. It means someone actually read something we wrote, found it useful, and chose to build on it. So, we were both intrigued learning that two of our papers were cited in the MAHA report. Being cited in a major government publication is not just recognition, it’s also publicity. But in this case, not good publicity. Here’s why.

Last week, both of us received calls from journalists about citations in the MAHA report. Kravitz was asked if he knew about a publication cited in the MAHA report (footnote No. 520) meant to support the claim that “more studies on direct-to-consumer television advertisements are directed to adults” than to children. The problem was that the cited paper did not exist. But the clever journalist was able to trace the phony citation to a real paper Kravitz wrote years before. The citation (since corrected) was supposedly published in the Lancet, a prestigious medical journal whose editorial gates Kravitz long aspired to breech. But it wasn’t. Worse, from the standpoint of ego (always an issue in academia), the paper was attributed to a different author. Most importantly, it did not directly address the claim it was used to corroborate.

Woloshin got a call from a journalist inquiring about his New England Journal of Medicine article, cited in the MAHA report (No. 519). The journalist asked if it was ironic that the report cited a journal Health and Human Services Secretary Robert F. Kennedy Jr.  recently trashed as unreliable.

Of course it was ironic. But it was also wrong. The article was published in a different (alas, less prestigious) journal in a different year and, most importantly, on the wrong subject. The citation, used to substantiate a claim about TV drug ads, was actually about print journalists’ coverage of disease awareness campaigns.

These errors were just two among many cited by the nonprofit outlet NOTUS. For example, the MAHA report cited a study about Texas children covered by Medicaid but used it to support a claim about all 50 states. Another claim, that stopping “antidepressants, stimulants, antipsychotics, and other psychiatric drugs … often leads to disabling and pronged physical dependence … and withdrawal,” was not supported by its citation, a review only about antidepressants and not focused on children. Four other citations were to nonexistent papers. Some believe these errors occurred because MAHA report authors may have relied on an AI agent to find citations.

White House press secretary Karoline Leavitt minimized these (and other) errors in the Report as “some [minor] formatting issues.”  But that is wrong. This is sloppy work which undermines faith in the whole report. A government report meant to influence the lives of millions of Americans should be thoroughly vetted and based on the most reliable sources. The government has reissued the report, but problems remain: The citation to Kravitz has been partially corrected, while Woloshin’s is unchanged.

Scientific and technological successes of the past century come in part from a culture that respects prior work and endeavors to get the facts right. The process isn’t perfect. As we and others have noted, scientific integrity remains under threat from predatory journals, porous peer review, poor results reproducibility, and other scientific crimes and misdemeanors. But it is undeniable that journals have made great progress in the last decades — for example, by creating standards for conflict of interest disclosure, pre-registration of randomized trial protocols (to prevent changing methods and outcomes after the fact), and publishing papers critical of the journals when their efforts fall short.

In contrast, given the Trump administration’s recent actions — such as agency closures, grant cancellations, and layoffs of technical experts — the errors in the MAHA report reinforce serious concerns about its dedication to rigorous inquiry and factual accuracy.

How could the MAHA authors have done better? We asked AI:

“I have to write a report to outline the leading drivers of childhood disease. Please suggest some basic principles for if and how to use and cite the research literature.”

AI was happy to oblige.  Here are the top 3 “key principles for using and citing research literature in your report:”

1. “Prioritize studies from reputable journals to ensure accuracy and credibility in your report.”

2. “Reference original studies directly rather than relying on summaries or secondary interpretations.”

3. “Select literature that directly applies to childhood disease to maintain focus and avoid misrepresenation.”

The post ended with general advice and an offer of help:

“This concise approach keeps the most important guidelines front and center while ensuring your report maintains integrity.  Would you like help incorporating these principles into a draft section of your report?”

Great answer. And funny. But also, given the stakes, very sad.

Steven Woloshin, M.D., M.S., is professor of medicine at the Geisel School of Medicine’s Dartmouth Institute. He is an unpaid member of the editorial boards of JAMA Internal Medicine and the International Congress on Peer Review and Scientific Publication — the views expressed here are his own. Richard L. Kravitz, M.D., M.S.P.H., is a distinguished professor at the UC-Davis Department of Internal Medicine.