Back to 2016 Annual Meeting
Resident Fatigue: Home call’s effect on sleep and fatigue on urology residents over apower weekend
Stephen T. Ryan, MD1, Moritz H. Hansen, MD1, Christine W. Duarte, MD2, Adam E. Ludvigson, MD1.
1Maine Medical Center, Portland, ME, USA, 2Maine Medical Center for Outcomes Research & Evaluation, Portland, ME, USA.
BACKGROUND – The Accreditation Council for Graduate Medical Education (ACGME) has mandated limited work-hours and required fatigue mitigation for all resident physicians(RP). Detailed guidelines exist for RP call in-hospital but few rules pertain to home call (HC), with no guidelines for fatigue detection or mitigation. A popular method of the HC system on the weekends is “Power Weekends”(PWs), where residents are the primary contact for 72-hours over the weekend. We tested an approach to record sleep during PWs and describe the relationship between sleep and a fatigue mitigation protocol.
METHODS – PW duty rotates among 4 RPs. Our fatigue mitigation protocol requires RPs to report being “Not Fatigued” or “Fatigued” after a PW. If “Fatigued” the RP is dismissed at noon and the event is recorded. A wrist mounted accelerometer (FitBit) measured total time asleep (TTA) in minutes each 24-hours for all RPs over a 7-month period. De-identified data was stratified into 3 categories: not on call “Baseline”(B), and On-Call either “Not Fatigued”(N) or “Fatigued”(F). TTA was compared pairwise among the 3 call categories using 2-sided t-tests for each individual. The combined data was analyzed with a mixed model, fit with a fixed scenario effect using the 3 call categories. Random resident and resident/scenario effects were used to account for the correlation within each resident/scenario combination. The dependent variable was TTA on Friday, Saturday, Sunday, or the TTA for the weekend.
RESULTS – All RP reported that the FitBit monitors were not an undue burden. Data was available for 86/120 total weekends over 4 RPs (55 B, 24 N, 6 F). Table1 contains average TTA±standard deviations. Table2 contains the significance values for Baseline verses Not-Fatigued weekends for each individual, and significance values using the mixed model. Average TTA±confidence intervals are in the figure. Overall, Baseline TTA was 1354 (±79) minutes. A call weekend was 1195 (±90) minutes for N and 1037 (±135) minutes for F. Using Friday, Saturday, and Sunday as the dependent variable, Sunday night was the only statistically significant night for Not Fatigued verses Fatigued.
CONCLUSIONS – TTA on a baseline weekend was higher than an On-Call weekend, and significant for 3 of 4 residents. Using a mixed model and accounting for individual resident variation, the TTA on a baseline weekend was higher than on an On-Call weekend. The average TTA for the fatigued weekends was lower than not-fatigued weekends, but was not statistically significant. Evaluating for the impact of individual nights, Sunday night is the strongest predictor of making a normal On-Call weekend convert into a Fatigued weekend (p=0.0004).
Back to 2016 Annual Meeting