A task analysis of emergency physician activities in academic and community settings.

Study Objective: We characterize and compare the work activities, including peak patient loads, associated with the workplace in the academic and community emergency department (ED) settings. This allows assessment of the effect of future ED system operational changes and identifies potential sources contributing to medical error.

Methods: This was an observational, time-motion study. Trained observers shadowed physicians, recording activities. Data included total interactions, distances walked, time sitting, patients concurrently treated, interruptions, break in tasks, physical contact with patients, hand washing, diagnostic tests ordered, and therapies rendered. Activities were classified as direct patient care, indirect patient care, or personal time with a priori definitions.

Results: There were 203 2-hour observation periods of 85 physicians at 2 academic EDs with 100,000 visits per year at each (N=160) and 2 community EDs with annual visits of 19,000 and 21,000 (N=43). Reported data present the median and minimum-maximum values per 2-hour period. Emergency physicians spent the majority of time on indirect care activities (academic 64 minutes, 29 to 91 minutes; community 55 min, 25 to 95 minutes), followed by direct care activities (academic 36 minutes, 6 to 79 minutes; community 41 minutes, 5 to 60 minutes). Personal time differed by location type (academic 6 minutes, 0 to 66 minutes; community 13 minutes, 0 to 69 minutes). All physicians simultaneously cared for multiple patients, with a median number of patients greater than 5 (academic 7 patients, 2 to 16 patients; community 6 patients, 2 to 12 patients).

Conclusion: Emergency physicians spend the majority of their time involved in indirect patient care activities. They are frequently interrupted and interact with a large number of individuals. They care for a wide range of patients simultaneously, with surges in multiple patient care responsibilities. Physicians working in academic settings are interrupted at twice the rate of their community counterparts.