Mathematical procedure to adjust for the healthy worker effect: the case of firefighting, diabetes, and heart disease.

Abstract

This article presents a mathematical procedure to adjust for one component of the healthy worker effect (HWE), namely, the healthy hired effect, on diabetes in the case of firefighting and heart disease. Three examples from real studies are given to illustrate, step-by-step, the application of the mathematical procedure. The mathematical procedure can be applied to adjust for other components of the HWE (e.g., the low-risk hired effect on obese individuals and smokers). In such cases, additional information will be needed to use the mathematical procedure. Results of applying the mathematical procedure in the case of firefighting and heart disease revealed the rather unexpected results that adjusting for diabetes selection on hiring leads to only a 3% to 9% increase in the magnitude of ratio statistics such as the standardized mortality ratio. It might be argued that the HWE from one component such as the healthy hired effect on diabetes might be small, but together with other components, the HWE might be large. Further investigation will be needed to support this argument.