Are data on the uptake of disability benefits a useful addition to census data in describing population health care needs?

To determine if administrative data relating to the uptake of disability benefits held by government departments in Northern Ireland are fair and unbiased indicators of relative need for health care and therefore a useful addition to census data.

This is an ecological study of the 566 electoral wards in Northern Ireland. The variation in uptake of a health-related benefit (Disability Living Allowance) was regressed against three other indicators of health (limiting long-term illness, mortality and the numbers of patients admitted to hospital). All the indicators had been indirectly standardised by age and sex. The unstandardised residuals from this model were regressed against the social and geographical factors, namely area deprivation score, religious denominational composition and urban/rural dwelling to see if they influenced benefit uptake above and beyond any association with ill health.

The health factors alone explained 77.2% of the variation in benefit uptake, with limiting long-term illness being the major determinant; however, even after controlling for differences in health status, benefit uptake was shown to be higher in more urban areas and in areas that had a greater proportion of Catholics. Area deprivation was not associated with benefit uptake once health differences had been controlled for.

Administrative data on disability benefits can undoubtedly provide useful additional information for describing the levels of relative disadvantage or ill health of areas. However, because they also reflect variations in uptake, which appears to be confounded by social and geographical factors, we would urge caution when they are used to identify priority areas or to allocate resources.