John M. Booker, Connie Roseman
Specific behavioral consequences of seasonal affective disorder have not been closely examined. Length of daylight is evaluated in relation to medication errors in a medical center located in the far north. Factors such as numbers of patient admissions, discharges, and deaths were controlled with data collected in Anchorage, Alaska, over 5 consecutive years, 1985–1989. These data revealed that 58% of all medication errors occurred during the first quarter of the year. Medication errors were 1.95 times more likely in December than September. The best statistical prediction was for errors associated with levels of darkness 2 months earlier. There may be not only an impairment of work performance among hospital nursing staff that reaches a peak in late winter but, more importantly, medication errors appear to follow a pattern that is closely associated with the annual cycle of daylight and darkness.