Exposure to working-hour characteristics and short sickness absence in hospital workers: A case-crossover study using objective data.

INTRODUCTION: Shift work characteristics, such as the number of night shifts or quick returns, are linked to disturbed sleep and greater work-life conflict, but little is known about their association with short sickness absences. Studies utilizing objective working-hour characteristic data can provide knowledge on specific shift work characteristics that may associate with sick leave. We investigated the association between working-hour characteristics in shift work and the incidence of short (1-3 days) sickness absences. PARTICIPANTS AND METHODS: The individual-level data were collected from the employers' electronic working time records of the Finnish Public Sector study. The final analytical sample was restricted to the first incidence of short sickness absence (1-3 days) from January 1, 2008 to the end of 2015. We studied employees with shift work contracts (n = 12 156) and day work contracts (n = 6225). Of these, 89% were female workers, with a mean age of 37.1 years, and 12% worked part-time. We calculated type of shift (morning, day, evening, and night), quick returns (< 11 h between two work shifts), length of work shift and weekly working time. We utilized case-crossover study design to compare the working-hour characteristics of the 28 days preceding sickness absence (exposure window) and those of the 28 days before the exposure window (control window), using conditional logistic regression models for odds ratios with 95% Confidence Intervals (95%CI).
RESULTS: Exposure to ≥ 2, and especially ≥ 4 consecutive night shifts associated with an increased likelihood of short sickness absence (OR 1.24, 95%CI 1.12-1.38 and OR 1.54, 95%CI 1.10-2.15, respectively) among shift workers, whereas a high number (> 25%) of evening shifts and having ≥ 2 consecutive evening shifts associated with lower odds of sickness absence. Over 40-hour weeks (OR 1.29, 95%CI 1.20-1.39), over 48-hour weeks (OR 1.24, 95%CI 1.01-1.54) and quick returns (OR 1.02, 95%CI 1.01-1.02) increased the likelihood of sickness absence. Having only a few (0-1 or 2-3) quick returns over 28 days decreased the likelihood of sickness absence (ORs 0.46 and 0.67), whereas having ≥ 5 quick returns over 28 days increased it (OR 1.88, 95%CI 1.76-1.99). These associations were similar among both shift and day workers.
CONCLUSIONS: Long working hours, several consecutive night shifts and quick returns increase the risk of short sickness absence. These working-hour characteristics should be avoided in shift scheduling to minimize sickness absence and increased costs for the employer.