Clinicians are more likely to express doubt in the medical records of Black patients than White patients, according to a new study published in PLOS One. Researchers analyzed over 13 million electronic health records from a large health system and found that notes about non-Hispanic Black patients were significantly more likely to contain language questioning the patient’s sincerity or competence.
Terms like “claims,” “insists,” or “poor historian” were flagged by artificial intelligence tools as indicators of doubt. While fewer than 1% of all notes contained such language, Black patients faced disproportionately higher odds of being described in ways that undermined their credibility. Compared to White patients, notes about Black patients were 29% more likely to contain credibility-undermining language and 50% more likely to question their competence.
The study’s authors say this pattern may contribute to ongoing racial disparities in healthcare. “For years, many patients – particularly Black patients – have felt their concerns were dismissed by health professionals,” they wrote. They argue that biased documentation can stigmatize patients and affect the quality of care they receive.
The researchers call for medical training that addresses unconscious bias and for AI tools used in clinical documentation to be programmed to avoid biased language. They believe their findings represent “the tip of an iceberg” and hope to raise awareness of credibility bias in healthcare.
See: “Clinicians more likely to express doubt in medical records of Black patients” (August 13, 2025)
