Identifying psychological morbidity among people with cancer using the Hospital Anxiety and Depression Scale: time to revisit first principles?

Background: The aim of this review was to describe the findings and methodological quality of studies, which sought to validate the Hospital Anxiety and Depression Scale (HADS) against the Structured Clinical Interview for DSM in cancer populations. We also sought to compare the cut points recommended by these validation studies with the way in which the HADS is currently used to determine prevalence of psychological morbidity in cancer populations. Methods: An electronic database search was conducted of Medline from 1983 to October 2010 for validation studies of the HADS in cancer populations. Reference lists of HADS reviews were hand searched. To examine which cut points are commonly used in cancer specific literature to identify the prevalence of psychological disorders, studies published in 2009 were identified via an electronic database search of Medline. Results: Ten studies, which validated the HADS against the Structured Clinical Interview for DSM in cancer patient populations, were found and examined in detail. None met all methodological criteria associated with the selection of a screening instrument. Recommendations for optimal HADS thresholds varied substantially across these studies. The most commonly used threshold for determining caseness in the 2009 literature on prevalence of psychological distress among patients with cancer was a subscale score of ≥ 8. This threshold was poorly supported by the results of the 10 cancer HADS validation studies examined. Conclusions: Caution is warranted in interpreting the results of prevalence studies using the HADS. There is a need to develop evidence about the optimal thresholds for defining caseness using the HAD