Accident risk as a function of hour at work and time of day as determined from accident data and exposure models for the German working population.

OBJECTIVES Recent studies indicate that accident risk may be a function of hour at work and time of day. Further evidence was sought for these assumptions, along with the answer to the question of whether the risk of accident can be conceived as an interaction between hours at work and time of day. METHODS Data on more than 1.2 million accidents for the year 1994 were provided, all listed according to the time of day and hour at work. Since information about how long each day and at what time of day people work is not available in Germany, different exposure models had to be estimated. For estimating the risk of having an accident relative accident risks were calculated from the ratio of accident frequencies to the exposure data. RESULTS An exponentially increasing accident risk was observed beyond the 9th hour at work. The relative accident risks differed considerably according to the respective exposure model with regard to time of day. A highly significant interaction effect was found for hour at work by time of day, the percentage of accidents at different hours at work varying according to the particular time of day when work is started. For the 3 "traditional" shiftwork starting times, it was shown that, with later starting times, the relative accident risk increased dramatically beyond the 8th hour at work. CONCLUSIONS Since the results clearly indicate that there are time-related effects on occupational accident risk, more detailed analyses are called for. More elaborated exposure models should be used to assess the efficiency of work schedules with extended workhours, especially under shiftwork conditions. The results also indicate the necessity of recording and providing adequate data bases for such analyses.