Clinical and behavioral predictors of short-term disability absence from work

Background and Objective: Extended absences from work due to illness are costly to workers, their employers and society at large. Many disability absences may be preventable in the near term. This study analyzes prospectively how observable medical conditions, biometric indicators and health behaviors predict short-term disability (STD) absences from work.

 Methods: An integrated database of employees’ medical and disability claims, lab results and survey responses is used to estimate the likelihood of an STD insurance claim within twelve months of an observation.  Logistic regression analysis is used to estimate the log odds of disability claim activity for employees with different medical conditions and health status indicators, and as a function of claim activity in the twelve months prior to an observation.

 Results: Employees with healthier diets and sleeping habits were at lower risk of a non-pregnancy STD claim in the following twelve months. Unhealthy HbA1c, blood glucose, HDL cholesterol, and triglyceride values were predictive of later STD claims, as were high stress and psychological distress levels. Several categories of medical conditions – mental disorders and digestive diseases in particular – were predictive of a later STD claim. However, treatment for symptoms and ill-defined signs were substantively less important, and employees with routine care were less likely to experience STD absences. This suggests that disability experiences are not associated with care-seeking in general.

Conclusions: Early identification of health risks among employees who have not yet experienced a disability leave from work may provide opportunities to prevent future episodes of work disability.