LO69: Evaluating the impact of night shifts on emergency medicine resident competence in simulated resuscitations

Abstract

Introduction: Sleep deprivation negatively affects cognitive and behavioural performance. Emergency Medicine (EM) residents commonly work night shifts and are then expected to perform with competence. This study examines the impact of night shifts on EM resident performance in simulated resuscitation scenarios. Methods: A retrospective cohort study was completed at a single Canadian academic centre where residents participate in twice-annual simulation-based resuscitation objective structured clinical examinations (OSCEs). OSCE scores for all EM residents between 2010-2016 were collected, as well as post-graduate year (PGY1-5), gender, and shift schedules. OSCEs were scored using the Queen’s Simulation Assessment Tool (QSAT) evaluating four domains: primary assessment, diagnostic actions, therapeutic actions and communication, and an overall global assessment score (GAS). A night shift was defined as a late evening (beyond 23:00) or overnight shift within the three days before an OSCE. A mixed effects linear regression model was used to model the association between night shifts and OSCE scores while adjusting for gender and PGY. Results: A total of 136 OSCE scores were collected from 56 residents. PGY-5 residents had 37.1% (31.3 to 34.0%; p<0.01) higher OSCE scores than those in PGY-1 with an average increase of 8.8% (7.5 to 10.1%; p<0.01) per year. Working one or more night shifts in the three days before an OSCE reduced the total and communication scores by an average of 3.8% (p=0.04) and 4.5% (p=0.04) respectively. We observed a significant gender difference in the effects of acute shift work (p=0.03). Working a night shift one night prior to an OSCE was not associated with total score among male residents (p=0.33) but was associated with a 6.1% (-11.9 to -0.2; p=0.04) decrease in total score among female residents. This difference was consistent across PGY and was primarily due to an 8.5% (-15.5 to -1.6%; p=0.02) decrease in communication scores and a 6.7% (-13.1 to -0.3%; p=0.04) reduction in GAS. Conclusion: Proximity to night shifts significantly impaired the performance of EM trainees in simulated resuscitation scenarios, particularly in the domain of communication. For female residents, the magnitude of difference in total scores after working such shifts one night prior to a resuscitation OSCE was approximately equal to the difference seen between residents one year apart in training.