Identifying Flight Schedule Characteristics Increasing Pilots Absenteeism at an Airline Using a Data Mining and Simulation Approach

Sickness absenteeism among flight crews is a pervasive problem disruptive to operations and costly for the employer. According to literature, exposure to certain schedule attributes has been associated with adverse health issues. However, the relationship between schedule characteristics and sickness absenteeism remains unclear. Therefore, the aim of this study is to identify schedule characteristics increasing the odds of sickness absenteeism based on historical data. Here, data records for each flight crew member were obtained from a Dutch low-cost airline in the period between 1 January 2018 and 24 January 2020. Schedule characteristics with an adverse effect on both the circadian and/or social rhythm, as identified in literature, were extracted from the available data, and included in the model. Exploration on these potential harmful schedule attributes was done using two generalised additive models. After adjusting for the socio-demographic and work-related confounding variables, simulations revealed that employees exposed to night shifts, backward, and forward rotations over a thirty-day period were significantly more likely to report sick. Furthermore, employees who flew four sectors showed higher odds to call in sick compared to employees who flew two sectors. Based on the results, it is recommended to schedule either sufficient rest periods after exposure or limit the occurrence of the identified schedule attributes.