Key points are not available for this paper at this time.
The sleep and performance of 12 male shift workers, operating a discontinuous, weekly alternating, three-shift system, were monitored over the course of one complete shift cycle. Compared with nocturnal sleep, day sleep was shorter in duration and was degraded in quality, and its sleep stages were temporally disrupted. Simple unprepared reaction time and four-choice reaction time were impaired at night, and simple reaction time deteriorated as a function of the number of days into the shift and the time on task. Sleep and performance changes can be primarily attributed to circadian factors; however, the deterioration in performance from night to night and with time on task is probably due to an accumulative sleep deficit. As far as sleep and performance are concerned, the best shift system is probably one having a short rotation cycle, with afternoon shifts or rest days preceding and following the night shift.
Tilley et al. (1982) studied this question.