Factors associated with intern fatigue

Abstract

Background: Prior data suggest that fatigue adversely affects patient safety and resident well-being. ACGME duty hour limitations were intended, in part, to reduce resident fatigue, but the factors that affect intern fatigue are unknown.

Objective: To identify factors associated with intern fatigue following implementation of duty hour limitations.

Design: Cross-sectional confidential survey of validated questions related to fatigue, sleep, and stress, as well as author-developed teamwork questions.

Subjects: Interns in cognitive specialties at the University of California, San Francisco.

Measurements: Univariate statistics characterized the distribution of responses. Pearson correlations elucidated bivariate relationships between fatigue and other variables. Multivariate linear regression models identified factors independently associated with fatigue, sleep, and stress.

Results: Of 111 eligible interns, 66 responded (59%). In a regression analysis including gender, hours worked in the previous week, sleep quality, perceived stress, and teamwork, only poorer quality of sleep and greater perceived stress were significantly associated with fatigue (p < 0.001 and p = 0.02, respectively). To identify factors that may affect sleep, specifically duty hours and stress, a secondary model was constructed. Only greater perceived stress was significantly associated with diminished sleep quality (p = 0.04), and only poorer teamwork was significantly associated with perceived stress (p < 0.001). Working >80 h was not significantly associated with perceived stress, quality of sleep, or fatigue.

Conclusions: Simply decreasing the number of duty hours may be insufficient to reduce intern fatigue. Residency programs may need to incorporate programmatic changes to reduce stress, improve sleep quality, and foster teamwork in order to decrease intern fatigue and its deleterious consequences.

Figure 1

Summary of relationships found among perceived stress, sleep, and fatigue. Figure illustrates results of multiple regression analyses. Note: Figure 1 shows the hypothesized relationships. It is possible, however, that the causal pathway may be reversed or that the relationships noted may be bi-directional.