In his novel The Magic Mountain, Thomas Mann describes wealthy European consumptives who retreat to a sanitorium high in the Swiss Alps, convinced that thin air aids in treating pulmonary tuberculosis. In their self-contained community, they affirm their faith in contemporary medical practices while delaying the demands of life in the \u201cflatlands.\u201d<\/p>\n
Medical researchers, it seems, are retreating to a sanatorium of their own. Based on our analysis, leading medical journals increasingly lean left and are abandoning scientific principles in service of progressive ideology.<\/p>\n
Finally, a reason to check your email.<\/p>\n
Sign up for our free newsletter today.<\/p>\n
We began by conducting a keyword search of every article published in the JAMA Network\u2014a group of 13 medical journals affiliated with the American Medical Association\u2014between April 1 and May 31. The phrase \u201cdiversity, equity, and inclusion\u201d appeared 56 times, more often than atherosclerosis (45) and osteoporosis (16). Another progressive-coded term, \u201cinequity,\u201d showed up 99 times\u2014more than asthma (75) or opioid use disorder (65).<\/p>\n
Phrase Searched
\n\u00a0Number of Results<\/p>\n
chronic kidney disease
\n128<\/p>\n
inequity
\n99<\/p>\n
asthma
\n75<\/p>\n
opioid use disorder
\n65<\/p>\n
structural discrimination
\n61<\/p>\n
diversity, equity, and inclusion
\n56<\/p>\n
congestive heart failure
\n53<\/p>\n
atherosclerosis
\n45<\/p>\n
structural racism
\n36<\/p>\n
historically marginalized group
\n33<\/p>\n
osteoporosis
\n16<\/p>\n
intersectionality
\n13<\/p>\n
Number of search results from the JAMA Network (April 1, 2025\u2013May 31, 2025)<\/p>\n
To assess the political valence of the articles, we performed a qualitative review of the 56 DEI papers. We asked two free online artificial intelligence chatbots\u2014xAI\u2019s Grok 3 and Microsoft\u2019s Copilot\u2014to rate the political bias of each of those articles on a scale from -5 (strongly left-leaning) to 5 (strongly right-leaning). We instructed the AI systems to assess each article according to various criteria, including language, framing, tone, and policy implications. We then averaged the chatbots\u2019 scores.<\/p>\n
The results revealed a decided left-wing slant. The average score across all 56 articles was -2.02. Seven scored -3.5 and six scored -3. Not one article leaned right.<\/p>\n
What explains these findings? One potential explanation is that the network\u2019s editors believe their readership is predominantly left-wing. That assumption was made plain in a recent editor\u2019s note<\/a>, published in response to the National Institute of Health\u2019s cap on indirect costs. \u201cMany JAMA readers hope that the NIH and HHS will lose these legal cases,\u201d the note read, \u201cand that the new 15% indirect cost rate will be struck down.\u201d Fiscally conservative readers may have had a different view<\/a>.<\/p>\n These findings reflect a broader leftward shift in medical research\u2014one that has contributed to the scientific establishment\u2019s erosion of rigor and objectivity. This shift threatens the foundations of the profession, the integrity of research, and the quality of medical education.<\/p>\n DEI-inspired scientific studies are often plagued by poor design: sampling bias, lack of proper control groups, leading survey questions, and circular reasoning. Researchers frequently treat incidental associations as proof of causation. These methodological flaws are baked into published papers, which then get cited, flaws and all, by others. Over time, some DEI-based studies acquire unwarranted credibility and authority.<\/p>\n Take Brad Greenwood and colleagues\u2019 research from 2020<\/a>, which claimed that \u201cnewborn-physician racial concordance\u201d\u2014meaning the child and the doctor are the same race\u2014\u201cis associated with a significant improvement in mortality for Black infants.\u201d But the original authors failed to control for babies with very low birth weight. When Manhattan Institute researchers replicated the study<\/a>, they found that racial concordance had no such effect. Nonetheless, the original study has received more than 800 citations, including 21 by articles within the JAMA Network.<\/p>\n Another study with similar left-wing premises has also made the rounds. Kelly Hoffman and colleagues<\/a> claimed in a 2016 paper that \u201cBlack Americans are systematically undertreated for pain relative to white Americans.\u201d Their conclusion was based on a sample of 194 medical students and 28 residents, with no attending physicians included except for \u201c10 experienced physicians,\u201d whose recommendations served as the benchmark against which sample participants\u2019 knowledge was evaluated.<\/p>\n In one of the paper\u2019s two case vignettes, these physicians identified narcotics as the appropriate pain management for kidney stones. Had they consulted the 7th edition of Tintinalli\u2019s Emergency Medicine Manual from 2012<\/a>, however, those physicians would have known that NSAIDs are the \u201cprimary analgesic of choice\u201d for that condition, not opiates. While this error calls into question the study\u2019s conclusions, the paper nevertheless has been cited over 3,000 times, according to Google Scholar.<\/p>\n Residents of the academy\u2019s Magic Mountain are sustained by skewed evidence and increasingly disconnected from reality. They would do well to step away from the illusion of certainty and return to the foundations of scientific inquiry\u2014rigorous debate, objective analysis, and a plurality of perspectives.<\/p>\n Howard H. Fenn<\/a> is a clinical psychiatrist certified by the American Board of Psychiatry and Neurology with additional qualifications in geriatric psychiatry. Kurt Miceli<\/a> serves as medical director for Do No Harm, a membership organization representing health-care professionals, patients, and policymakers focused on keeping identity politics out of medicine.<\/p>\n Photo: Maskot \/ DigitalVision via Getty Images<\/p>\n