Physicians’ Implicit Racial Bias and Verbal Dominance and Perceived Discrimination in Goals of Care Conversations (RP115)

Outcomes 1. Describe the relationship between physician implicit racial bias, communication behavior and patient and family outcomes including perceived discrimination and feeling heard and understood.2. Select appropriate communication outcomes to target for communication interventions aimed at increasing health equity in end-of-life care. Key Message Understanding the relationship between physician implicit racial bias, communication behavior and patient and family perceived discrimination in end-of-life communication will help identify communication outcomes, such as verbal dominance, that interventions should target. Importance Implicit racial bias affects patient-provider communication which may contribute to disparities in burdensome end-of-life care. Objective(s) We examined the relationship between physicians’ implicit racial bias and paraverbal communication behaviors and between those behaviors and perceived discrimination in a simulated end-of-life communication. Scientific Methods Utilized Physicians completed a 20-minute videotaped simulated goals of care discussion with a Black actor portraying a patient's daughter. The patient was a 70-year-old woman with advanced metastatic lung cancer and acute respiratory failure. Physicians completed the Implicit Association Test (IAT) for racial bias after the simulation. Actors completed the Discrimination in Medical Settings Scale and a single-item measure of feeling Heard and Understood. Two investigators coded paraverbal communication (including verbal dominance as a ratio of physician to actor speech time) from videotaped encounters. Results Of 43 physicians, 22 (51%) were female, 22 (51%) White, 17 (40%) Asian, 4 (9%) Hispanic and 1 (2%) Black. 8 were oncologists (19%), 15 intensivists (35%), 16 internists (37%), and 4 geriatricians (9%). Two did not complete the IAT. Average IAT-D score was 0.31 (SD 0.51). Physicians paused on average 4.6 times (SD 4.5) and interrupted 1.3 times (SD 1.8) per encounter. Average speech rate was 149.9 syllables per minute (SD 17.2). Average verbal dominance was 2.19 (range 1.04, 5.78). Average Heard and Understood score was 3.7 (SD 1.2), and Discrimination was 1.6 (SD 0.8) (scale of 1-5 for both). There were no significant correlations between IAT-D score and any paraverbal communication behaviors. Verbal dominance correlated with increased perceived discrimination (r=0.64, p< 0.01) and lower ratings of feeling heard and understood (r=-0.36, p< 0.05). Conclusion(s) Implicit racial bias was not correlated with paraverbal communication behaviors. Verbal dominance was highly correlated with perceived discrimination. Impact Understanding which behaviors correlate with perceptions of discrimination will help identify communication outcomes, such as verbal dominance, that should be targeted by interventions. Keywords Communication and prognostication/Cultural diversity