Does who is counted matter? Comparing disability measures and self-rated health

People with disabilities experience significant health disparities compared to those without disabilities. However, analysts disagree on which measures should be used to identify the disabled population. Differences in disability measurement hamper efforts to accurately estimate health risks experienced by people with disabilities. This study investigates how self-rated health differs based on responses to two common population measures of disability.

Using weighted data from the 2011 and 2012 National Health Interview Survey (N = 25,096), this study employed logistic regressions to estimate the predicted probabilities of reporting poor or fair self-rated health based on responses to the disability measures from the American Community Survey (ACS-6) and the Washington Group Short Set (WG-SS).

While the ACS-6 identifies a significantly higher prevalence of disability (16.9%) than the WG-SS (7.9%), the adjusted predicted probabilities of poor/fair health did not differ significantly between measures (0.329 for ACS-6; 0.358 for WG-SS). However, individuals counted only by the ACS-6, those who would be excluded if federal surveys adopted the more restrictive WG-SS, were roughly 20 percentage-points more likely to report poor/fair health than non-disabled respondents.

The WG-SS excludes a considerable proportion of the disabled population who experience significantly worse health than non-disabled people. Relying solely on the more restrictive WG-SS may omit millions of individuals, potentially hampering federal efforts to achieve health equity.