The impact of prolonged continuous wakefulness on resident clinical performance in the intensive care unit: a patient simulator study.

Abstract

To evaluate the impact of prolonged continuous wakefulness on resident performance under controlled experimental conditions. Experimental within-subjects comparison. High-fidelity patient simulator. Twelve residents in an Internal Medicine Program at various stages of training (range, 1-35 mos). Performance was studied during 26 hrs of continuous wakefulness at four time points (8:00-10:00 am, 2:00-4:00 pm, 2:00-4:00 am, and 8:00-10:00 am the next day) using high-fidelity patient simulation. At each session, residents managed eight simulated dysrhythmias according to advanced cardiac life support protocols (advanced cardiac life support scenarios) and then managed a simulated critically ill patient (e.g., patient with meningitis) to test more complicated clinical decision-making (complex scenario). The frequency of previously defined major medical errors (i.e., action or inaction that likely would have resulted in significant harm in a real patient) was assessed by a scorer blinded to the time of the session. For each complex scenario, a global score between 0 and 100 was also given for overall performance. The impact of wakefulness on performance was assessed by using longitudinal mixed-effects models. For the complex scenarios, the mean number of errors increased from 0.92 +/- 0.90 in the first session to 1.58 +/- 0.79 in the fourth session (p = .09), and mean global score decreased from 56.8 +/- 14.6 to 49.6 +/- 12.6 (p = .02). For the advanced cardiac life support scenarios, the mean number of major errors committed in the advanced cardiac life support scenarios decreased during the study period (p = .01). However, essentially all of the improvement occurred between the first and second time points, suggesting that a substantial learning effect accounted for the findings. During prolonged continuous wakefulness of medical residents, clinical performance in the management of a simulated critically ill patient deteriorates. The practice of scheduling residents for extended work shifts (>24 hrs) should be reconsidered.