The Hospital Anxiety and Depression Scale (HADS): structurally unsound and unfixable

Abstract

The Hospital Anxiety and Depression Scale (HADS) is a widely used self-report measure for screening and assessing medical patients for anxiety and depressive symptoms. Recently, doubts were raised about highly variable factor structure, discrepant cutpoints, and inability to distinguish between anxiety and depression. To salvage large amounts of published studies and unpublished data, proposals are being made to reconceptualize the HADS as a unidimensional measure of general distress. We demonstrate that problems are intrinsic and unresolvable, due to decisions made by the original developers. The HADS was constructed with concerns about avoiding careless responding and acquiescence. Developers found a self-defeating solution in presenting respondents with overwhelming cognitive demands, posed by items that shifted from anxiety versus depression, as well as the direction and content of both items and response keys from each item to the next. These problems are hiding in plain sight. Discrepancies in structure, cutpoints, and discriminant validity reflect a high level of respondent confusion and misresponse. Much can be learned from problems of the HADS for designing valid and reliable measures consistent with more clearly defined purposes.