0155 SLEEP WHILE ON-CALL OVERNIGHT DOES NOT RESTORE PERFORMANCE AMONG FIRST-YEAR RESIDENT PHYSICIANS

Abstract

It has recently been reported that providing resident physicians with 3- or 5-hour protected windows for sleep during an on-call night increases the amount of sleep obtained and improves next-day performance relative to an on-call night with no protected window, in contrast to earlier findings. These studies have not reported, however, the relationship between the amount of sleep during this protected window and the degree of improvement in performance, and whether performance is sufficiently restored to warrant such an intervention. Here, we explored the dose-response relationship between amount of sleep during an on-call night and next-day performance. We analyzed data from 34 postgraduate-year 1 resident physicians (23 males; 28.0 ± 1.8 years old) studied for 3 weeks on a Q3 schedule (24-30-hour on-call extended duration work shift [EDWS] every other shift). Resident physicians completed daily sleep/work logs (validated by ambulatory polysomnography) and 10-min psychomotor vigilance tests (PVTs) every ~6 hours during each EDWS. We used a generalized estimating equation to examine the correlation between spontaneous sleep amounts (0, >0–1, >1–2, >2–3, >3–4, and >4 h bins) overnight during the EDWS (23:00-07:00) and next-day (05:00-19:00) PVT attentional failures (reaction time >500 ms). Time of day of PVT and study week were included as covariates. A total of 296 PVTs were analyzed. Compared to no sleep overnight, >4 hours of actual sleep (4.15–6.33 h) was required before a significant reduction in next-day PVT attentional failures was observed (p=0.0237). Sleeping <4 hours did not significantly improve performance (p>0.05). Even following >4 hours of sleep, PVT performance remained impaired, with 2 or more attentional failures on 68% of tests and 5 or more attentional failures on 39% of tests. Our data show that resident physicians need >4 hours of sleep on-call overnight to improve next-day PVT performance. Performance remained suboptimal despite this relative improvement, highlighting the substantial sleep deficiency inherent in EDWS. Given that 3- and 5-hour ‘protected’ windows result in far less than 4 hours of total sleep, our findings challenge reports that such protected time during an on-call shift is sufficient to maintain optimal next-day performance. NHLBI-4U01HL111478/NIOSH-R01OH07567/AHRQ-R01HS12032/NCRR-M01RR02635/UL1-RR025758/NHLBI-T32HL079010/NSBRI-NASA-NCC-9–58.