al., 1990). Thus, fatigue causes decreased productivity; tired workers accomplish less, especially if their tasks demand accuracy (Krueger, 1994; Rosa and Colligan, 1988).

Since almost all physiological and behavioral functions are affected by circadian rhythms, the time of day when work occurs is important. Overall capacity for physical work is reduced at night (Cabri et al., 1988; Cohen and Muehl, 1977; Rosa, 2001; Wojtczak-Jaroszowa and Banaszkiewicz, 1974). Reaction times, visual search, perceptual–motor tracking, and short term memory are worse at night than during the daytime (Folkard, 1996; Monk, 1990). On-the-job performance also deteriorates; for example, railroad signal and meter reading errors increase at night, minor errors occur more often in hospitals, and switchboard operators take longer to respond to phone calls (Monk et al., 1996).

The human circadian rhythm strongly favors sleeping during the nighttime hours. Although one study notes that nurses working a permanent night shift or rotating shifts obtained more sleep on average than nurses working day or evening shifts, almost one-fifth of the nurses reported having struggled to stay awake while taking care of a patient at least once during the previous month (Lee, 1992). Another study found that falling asleep during the night shift occurred at least once a week among 35.3 percent of nurses who rotated shifts, 32.4 percent of nurses who worked nights, and 20.7 percent of day/evening shift nurses who worked occasional nights (Gold et al., 1992). It was also found that nurses working night or rotating shifts made more on-the-job procedural errors and medication errors because of sleepiness than nurses working other shifts. Sleepiness appeared to be confined to the night shift, as none of the shift rotaters or day/ evening nurses who worked occasional nights reported significant difficulties remaining alert on other shifts.

These subjective reports of sleeping on duty were recently verified by both activity (wrist actigraphy) and sleep (polysomnographic) recordings of 15 French nurses who worked at night (Delafosse et al., 2000). Only 4 of the 15 nurses were able to remain awake all night at work; the others averaged 86.5 (standard deviation [SD] ± 77.6) minutes of sleep while on duty.

Moreover, difficulties maintaining alertness at night are not confined to nurses; episodes of both subjective (or self-reported) and objective sleep were recorded while U.S. Air Force traffic controllers were on duty at night (Luna et al., 1997), and episodes of drowsiness at the wheel were observed in the majority of 80 commercial truck drivers (Wylie et al., 1996).

A person who is not sleep deprived performs tasks more efficiently after prolonged wakefulness and can cope better with nonstandard work hours (nights or rotating shifts) than someone with a sleep deficit (Dinges et al., 1996). Individuals working nights and rotating shifts rarely obtain optimal amounts of quality sleep. Their sleep is shorter, lighter, more fragmented,

The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement