Biased Language in Simulated Handoffs and Clinician Recall and Attitudes

Poor-quality handoffs can lead to medical errors when transitioning patient care. Biased language within handoffs may contribute to errors and lead to disparities in health care delivery.

To compare clinical information recall accuracy and attitudes toward patients among trainees in paired cases of biased vs neutral language in simulated handoffs.

Surveys administered from April 29 to June 15 and from July 20 to October 10, 2023, included 3 simulated verbal handoffs, randomized between biased and neutral, and measured clinical information recall, attitudes toward patients, and key takeaways after each handoff. Participants included residents in internal medicine, pediatrics, and internal medicine-pediatrics and senior medical students at 2 academic medical centers in different geographic regions of the US. Data were analyzed from November 2023 to June 2024.

Each participant received 3 handoffs that were based on real handoffs about Black patients at 1 academic center. These handoffs were each randomized to either a biased or neutral version. Biased handoffs had 1 of 3 types of bias: stereotype, blame, or doubt. The order of handoff presentation was also randomized. Internal medicine and pediatrics residents received slightly different surveys, tailored for their specialty. Internal medicine-pediatrics residents received the pediatric survey. Medical students were randomly assigned the survey type.

Each handoff was followed by a clinical information recall question, an adapted version of the Provider Attitudes Toward Sickle Cell Patients Scale (PASS), and 3 free-response takeaways.

Of 748 trainees contacted, 169 participants (142 residents and 27 medical students) completed the survey (23% overall response rate), distributed across institutions, residency programs, and years of training (95 female [56%]; mean [SD] age, 28.6 [2.3] years). Participants who received handoffs with blame-based bias had less accurate information recall than those who received neutral handoffs (77% vs 93%; P = .005). Those who reported bias as a key takeaway of the handoff had lower clinical information recall accuracy than those who did not (85% vs 93%; P = .01). Participants had less positive attitudes toward patients per PASS scores after receiving biased compared with neutral handoffs (mean scores, 22.9 [3.3] vs 25.2 [2.7]; P < .001). More positive attitudes toward patients were associated with higher clinical information recall accuracy (odds ratio, 1.12; 95% CI, 1.02-1.22).

In this survey study of residents and medical students, biased handoffs impeded accurate transfer of key clinical information and decreased empathy, potentially endangering patients and worsening health disparities. Handoff standardization is critical to addressing racial bias and improving patient safety.