Does Multitasking Improve Performance? Evidence from the Emergency Department

This paper examines the effect of multitasking on overall worker performance, as measured by processing time, throughput rate, and output quality using microlevel operational data from the field. Specifically, we study the multitasking behavior of physicians in a busy hospital emergency department (ED). By drawing on recent findings in the experimental psychology literature and the nascent work in cognitive neuroscience, we develop several hypotheses for the effect of multitasking on worker performance. We first examine how multitasking affects a physician's processing time. We find that the total time taken to discharge a given number of patients has a U-shaped response to the level of physician multitasking; that is, multitasking initially helps to reduce the time taken, but only up to a certain threshold level, after which it increases in the level of multitasking. In addition, multitasking significantly impacts quality of care. Although lower levels of multitasking are associated with improved quality of care, at higher levels, additional multitasking leads to a smaller number of detected diagnoses and an increased likelihood of a 24-hour revisit rate to the ED. These findings have important implications for the design and organization of work in general and for the delivery of critical care in particular.