Employee health index based on health checkups and its association with future absent workdays

Employee health is an important factor that influences the productivity and economic success of companies. The goal of health management in companies is to improve employee health by implementing different kinds of interventions, including checkups (voluntary preventive health examinations) that cover a broad set of tests and questions, to give a comprehensive snapshot of the employee's health and uncover priorities for prevention and health promotion. By analyzing health checkup information on a company-wide level, subgroups with increased need for health management interventions may be identified before productivity loss occurs. This study thus analyzed data from a company's health checkup to form an employee health index (EHI) using principal component analysis (PCA) to adequately summarize the data. Additionally, this study investigated whether EHI values were associated with the number of absent workdays on the department level. The study population consisted of employees of a German automotive company. In total, data from 71,054 voluntary health checkups conducted between 2006 and 2016 were analyzed. PCA resulted in five components in three different health aspects (musculoskeletal disorders, mental health, and physiology). Based on these components, departments were categorized as having low, medium, and high needs for health management interventions. Poisson regression was used to analyze differences in absent workdays between departments with low compared to medium or high need for health management interventions. Components describing musculoskeletal disorders and physiology were associated with a change in the rate of absent days per employee. However, results were inconsistent over different time intervals. Therefore, although the presented EHI was to a certain extent associated with future absent workdays, using checkup data to investigate more specific health problems and outcomes related to individual job-related activities or tasks instead of general measures seems to be a more appealing approach.