Making Sense of Variations in Prevalence Estimates of Depression in Cancer: A Co-Calibration of Commonly Used Depression Scales Using Rasch Analysis

Background: The use of different depression self-report scales warrants co-calibration studies to establish relationships between scores from 2 or more scales. The goal of this study was to examine variations in measurement across 5 commonly used scales to measure depression among patients with cancer: Hospital Anxiety and Depression Scale-Depression subscale (HADS-D), Centre for Epidemiologic Studies Depression Scale (CES-D), Patient Health Questionnaire-9 (PHQ-9), Beck Depression Inventory-II (BDI-II), and Depression Anxiety and Stress Scale-Depression subscale (DASS-D). Methods: The depression scales were completed by 162 patients with cancer. Participants were also assessed by the major depressive episode module of the Structured Clinical Interview for Diagnostic and Statistical Manual of Mental Disorders, 4th Edition. Rasch analysis and receiver operating characteristic curves were performed. Results: Rasch analysis of the 5 scales indicated that these all measured depression. The HADS and BDI-II had the widest measurement range, whereas the DASS-D had the narrowest range. Co-calibration revealed that the cutoff scores across the scales were not equivalent. The mild cutoff score on the PHQ-9 was easier to meet than the mild cutoff score on the CES-D, BDI-II, and DASS-D. The HADS-D possible cutoff score was equivalent to cutoff scores for major to severe depression on the other scales. Optimal cutoff scores for clinical assessment of depression were in the mild to moderate depression range for most scales. Conclusions: The labels of depression associated with the different scales are not equivalent. Most markedly, the HADS-D possible case cutoff score represents a much higher level of depression than equivalent scores on other scales. Therefore, use of different scales will lead to different estimates of prevalence of depression when used in the same sample.