Quantification of the healthy worker effect: a nationwide cohort study among electricians in Denmark

Abstract

BackgroundThe healthy worker effect (HWE) is a well-known phenomenon. In this study we used the extensive registration of all Danish citizens to describe the magnitude of HWE among all Danish electricians and evaluated strategies for minimizing HWE bias of the association between occupation and mortality.MethodsAll Danish male citizens aged 26-56 years in the period 1984-1992 were followed for three years in several registers. We evaluated HWE bias among electricians because they were unexposed to detrimental occupational exposures. We compared electricians to three reference groups (general population, construction industry and carpenters/brick layers) and utilized analytical methods for minimizing HWE bias (lag time analyses, age-stratified analyses, marginal structural model and restriction to employed, newly employed or long-term workers).ResultsThe mortality rate was higher among electricians, who the year following active employment received incapacity benefits or were on long-term sick leave. Electricians receiving incapacity benefits, on long-term sick leave, unemployed, or with increased comorbidity index had lower odds of re-employment. Electricians had lower mortality rate (rate ratio,0.60;95%CI,0.52-0.69) compared to the general population, while electricians leaving employment had increased mortality (1.90;1.50-2.40). Adjusting for several social events slightly attenuated the estimates, while the marginal structural model did not minimize bias. Electricians had the same mortality as the construction industry and carpenters/brick layers. Mortality was comparable to the general population after three or more years of lag time.ConclusionsIn this nationwide study, employment as electricians had marked effect on mortality. Appropriate reference selection and lag time analyses minimized the HWE bias.

Inequalities in Transplant Waiting List Activation across Italian Dialysis Centers

“…It is extremely difficult to discern whether socioeconomic disparities in transplant wait-listing arise from discrimination, patients preferences [32], or sensible differences in outcome expectations by healthcare providers (i.e. employed patients might be healthier [33]). Additionally, insurance-related barriers to transplantation in countries without a universal healthcare system might be important as well [7]: the incentive structure of welfare systems usually plays a major role in healthcare utilization decisions, especially for expensive treatments.…”

The healthy worker effect: Do health problems predict participation rates in, and the results of, a follow-up survey?

“…The HWE should therefore under no circumstances be disregarded as a potential bias based on the findings of the current study alone. To further understand how the HWE can bias research, future studies should investigate the impact of health on initial survey participation, for instance, through the use of registry data on sickness absence (Thygesen et al 2011).…”

Commentary for the Then and Now Forum

“…[15][16][17] The HWHE occurs during the initial selection process of subjects when they first gain employment [15][16][17] and results from the overall better health status and favorable survival of the actively employed compared with the general population. Given that the HWE has been attributed to the use of an inappropriate reference comparison group, 17,18 and could be minimized if the general population was not used for comparison, 19 internal 20,21 and external reference groups [22][23][24] have been, and continue to be, used as the comparison groups-in lieu of the general population. The selection process of workers from a cohort leaving employment differentially occurs over time [15][16][17] usually leading to downward bias with underestimation of the effects of possibly harmful workplace conditions.…”