When Words Shape Care: Educating First-Year Medical Students on Identifying and Replacing Bias in Electronic Health Documentation

Although medical education increasingly addresses bias in patient care, less attention has been paid to biased language in clinical documentation within the Electronic Health Record (EHR), a critical concern as patients have increasing access to their records. Because biased language can influence clinicians' perceptions and negatively affect patient relationships, focused medical student training is needed to promote equitable, patient-centered care.

We implemented a one-hour workshop on biased language in medical records using case-based discussions. First-year students (MS1s) completed pre- and postsurveys assessing their ability to identify biased language. Results were analyzed using paired t-tests, descriptive statistics, and thematic analysis.

Sixty-nine MS1s attended the workshop; 68 completed the presurvey (98.5%), and 63 the postsurvey (91.3%). Students demonstrated significant improvement in their ability to identify biased language in patient cases (P < .001) and survey scores (P < .001). Significant improvements were also observed in students' ability to identify the word "claims" as biased language (P < .001), identify language that defined patients by their illness (P < .001), and recognize negative patient descriptions (P < .001). Students valued the workshop's use of examples of biased language and neutral language alternatives, and many recommended incorporating this training into the medical school curriculum.

The workshop improved MS1s' recognition of biased language in EHR documentation. Incorporating documentation bias training early in medical school may strengthen students' awareness of biased language and support more patient-centered documentation.