Do workers underreport morbidity? The accuracy of self-reports of chronic conditions

We use matched Danish health survey and register data to investigate discrepancies between register-based diagnoses and self-reported morbidity. We hypothesize that false negatives (medical diagnoses existing in the register but not reported in the survey) arise partly because individuals fear career repercussions of being discovered suffering a chronic or severe illness that potentially lowers productivity. We find evidence of substantial underreporting, which is indeed systematically higher for individuals in the labor market.

► We compare Danish survey self-reports of chronic conditions collected in 2000 with matched hospital register data from the years 1997 to 1999.
► Survey self-reports miss as much as 50 percent of all conditions treated in hospitals in the preceding three years.
► Underreporting rates are found to vary systematically with disease severity.
► Workers underreport conditions more than non-workers.
► Other socio-demographic background, including education, is not associated with underreporting rates.