Predictors of short- and long-term sickness absence in female post office workers in Poland

Abstract

Background: The aim of this study was to highlight major predictors of the frequency of sickness absence in a group of workers directly involved in customer service. Material and Methods: The study was carried out on a random sample of 229 women employed as assistants and clerks in post offices. The survey was based on the Subjective Work, Health Status and Life Style Characteristics Questionnaire, and sickness absence data for the years 2004.2006. Results: The negative binominal regression model of sickness absence risk revealed the following significant predictors of short-term absence spells (1.29 days):1) marital status, sickness absence risk for single women was (rate ratio (RR)) = 1.56 (95% confidence interval (CI):1.01.2.39) vs. married women; 2) post offices employing 7 workers had a rate ratio of sickness absence of 1.6 (95% CI: 1.04.2.42); 13.25 workers. RR = 2.03 (95% CI: 1.41.2.93); > 25 workers. RR = 1.82 (95% CI: 1.15.2.88) compared with an average number of 8.12 workers; 3) shift work, RR = 1.57 (95% CI: 1.14.2.14); 4) breaks from work. the risk of absence in the case of any breaks amounted to RR = 1.5 (95% CI: 1.07.2.07) in comparison with the statutory breaks; 5) self-rated health reported as moderate relative to good health, RR = 1.71 (95% CI: 1.26.2.32); and 6) occurrence of respiratory diseases resulted in the risk of RR = 1.51 (95% CI: 1.08.2.08). The Poisson regression model of long-term sickness absence spells (≥ 30 days) revealed the following significant predictors: 1) number of clients per shift: 51.100 clients, RR = 3.62 (95% CI: 1.07.22.6) compared with a lower number of clients; 2) self-rated health, assessed as moderate, RR = 1.97 (95% CI: 1.06.3.78) and 3) household chores performed for at least 4 h a day, RR = 0.4 (95% CI: 0.18.0.79). Conclusions: Association between sickness absence and workload as well as work organization indicates directions of corrective actions, which could reduce the scale of the problem.