and less restorative than sleep at night (Knauth et al., 1980; Lavie et al., 1989; Walsh et al., 1981).

Workers are more likely to report greater fatigue at the end of 12-hour work shifts than at the end of 8-hour workshifts (Mills et al., 1983; Rosa 1995; Ugrovics and Wright, 1990). There are exceptions, however: mineworkers reported no differences in fatigue after 8- and 12-hour shifts despite high physical workloads (Duchon et al., 1994), and computer operators reported reduced tiredness throughout the shift after switching from 8-hour to 12-hour shifts (Williamson et al., 1994). Although the timing and duration of meal breaks and “coffee” breaks were not described in these studies, in the case of unionized mineworkers, it is likely they were allowed to stop working for brief periods during their work shift.

Sustained operations (shifts of 12 or more hours with limited opportunity for rest and no opportunity for sleep) (Krueger, 1989) often occur among health care providers who staff busy emergency rooms and intensive care units (ICUs), work overtime shifts on nursing units, or work as members of surgical teams that perform lengthy or consecutive procedures (Krueger, 1989). The majority of anesthesiologists and anesthesia residents report having made errors in the administration of anesthesia when fatigued (Gravenstein et al., 1990). The California Nurses Association (CNA) website (CNA, 2001a) reports several serious errors committed by nurses mandated to work 16-hour shifts, in addition to cases in which nurses did not make errors but were at high risk for doing so. For example, a nurse who worked on average one mandatory double shift (16 hours) every 2 weeks for a 2-month period reported that “by 4 a.m. I was so exhausted that I would stop between going from one baby to the next and completely forget why I was going to the other bedside. Another time, again about 4 a.m., I would sometimes stop in the middle of the floor and forget what I was doing.”

Studies have shown that accident rates increase during overtime hours (Kogi, 1991; Schuster, 1985), with rates rising after 9 hours, doubling after 12 consecutive hours (Hanecke et al., 1998), and tripling by 16 consecutive hours of work (Akerstedt, 1994). Data from aircraft accident investigations of the National Transportation Safety Board (NTSB) also show higher rates of error after 12 hours (NTSB, 1994a). Finally, night shifts longer than 12 hours and day shifts longer than 16 hours have consistently been found to be associated with reduced productivity and more accidents (Rosa, 1995).

Laboratory studies have shown that moderate levels of prolonged wakefulness can produce performance impairments equivalent to or greater than levels of intoxication deemed unacceptable for driving, working, and/or operating dangerous equipment (Dawson and Reid, 1997; Lamond and Dawson, 1998). In one study, performance on neurobehavioral tests remained relatively stable during the first 17 hours of testing, a period the

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