Objectives: To test whether hospital-based physicians made different intensive care unit and life-sustaining treatment decisions for otherwise identical black and white patients with end-stage cancer and life-threatening hypoxia.

Design: We conducted a randomized trial of the relationship between patient race and physician treatment decisions using high-fidelity simulation. We counterbalanced the effects of race and case by randomly alternating their order using a table of random permutations. Physicians completed two simulation encounters with black and white patient simulator patients with prognostically identical end-stage gastric or pancreatic cancer and life-threatening hypoxia and hypotension, followed by a self-administered survey of beliefs regarding treatment preferences by race. We conducted within-subjects analysis of each physician's matched-pair simulation encounters, adjusting for order and case effects, and between-subjects analysis of physicians' first encounter, adjusting for case.

Setting: Peter M. Winter Institute for Simulation Education and Research at the University of Pittsburgh, Pennsylvania.

Subjects: Thirty-three hospital-based attending physicians, including 12 emergency physicians, eight hospitalists, and 13 intensivists from Allegheny County, Pennsylvania.

Intervention: Race of patient simulator.

Measurements and main results: Measurements included physician treatment decisions recorded during the simulation and documented in the chart and beliefs about treatment preference by race. When faced with a black vs. a white patient, physicians did not differ in their elicitation of intubation preferences (within-subject comparison, 28/32 [88%] vs. 28/32 [88%]; p = .589; between-subject comparison, 13/17 [87%] vs. 13/17 [76%]; p = .460), intensive care unit admission (within-subject comparison, 14/32 [44%] vs. 12/32 [38%]; p = .481; between-subject comparison, 8/15 (53%) vs. 7/17 (41%); p = .456), intubation (within-subject comparison, 5/32 [16%] vs. 4/32 [13%]; p = .567; between-subject comparison: 1/15 [7%] vs. 4/17 [24%]; p = .215), or initiation of comfort measures only (within-subject comparison: 16/32 [50%] vs. 19/32 [59%]; p = .681; between-subject comparison: 6/15 [40%] vs. 8/17 [47%]; p = .679). Physicians believed that a black patient with end-stage cancer was more likely than a similar white patient to prefer potentially life-prolonging chemotherapy over treatment focused on palliation (67% vs. 64%; z = -1.79; p = .07) and to want mechanical ventilation for 1 wk of life extension (43% vs. 34%; z = -2.93; p = .003), and less likely to want a do-not-resuscitate order if hospitalized (51% vs. 60%; z = 3.03; p = .003).

Conclusions: In this exploratory study, hospital-based physicians did not make different treatment decisions for otherwise identical terminally ill black and white elders despite believing that black patients are more likely to prefer intensive life-sustaining treatment, and they grossly overestimated the preference for intensive treatment for both races.