We examined emergency physician disposition decisions and computed tomography (CT) ordering as markers of decision fatigue over an eight-hour shift.

Administrative database analysis of patients presenting to an academic, tertiary care, emergency department (ED) over two years. Patients were grouped by the hour of the shift that they were initially assessed by an emergency physician. For each hour, we evaluated the proportions of patients who had CT head, chest, or abdomen, consultations, and consultations not resulting in admission. For patients discharged without consultation, we evaluated return visits within 72 h and ED length-of-stay (LOS). Statistical significance was assessed using random effects regression accounting for clustering by physician.

We analyzed 87,752 patients and there were no important differences in consultations, consultations not resulting in admission, or return visits in relation to the hour of shift the patient was seen. Rates of CT head and abdomen and ED LOS decreased as the shift progressed. From the first to the last hour, CT head ordering decreased from 15.8% to 12.2% (OR 0.73, 95% CI 0.66–0.80, p < 0.0001), CT abdomen ordering decreased from 9.6% to 7.6% (OR 0.72, 95% CI 0.64–0.80, p < 0.0001), and ED LOS decreased from 5.5 h to 4.9 h (relative difference 0.83, 95% CI 0.81–0.85, p < 0.0001).

Emergency physician decisions about patient disposition did not change throughout the shift. The rates of CT head and abdomen and ED LOS decreased as the shift progressed. We did not find evidence of decision fatigue among emergency physicians over an eight-hour shift.