Comparative validity and responsiveness of PHQ-ADS and other composite anxiety-depression measures

ABSTRACT Background Composite measures that assess the overall burden of anxiety and depressive symptoms have been infrequently evaluated in the same study. The objective of this study was to compare the validity and responsiveness of the Patient Health QuestionnaireAnxiety-Depression Scale (PHQ-ADS) and other composite anxiety-depression measures. Methods The sample comprised 256 primary care patients enrolled in a telecaretrial of chronic musculoskeletal pain and comorbid depression and/or anxiety. Measures included the PHQ-ADS; the 8-item and 4-item depression and anxiety scales from the PROMIS profiles; the PHQ-anxiety-depression scale (PHQ-4); the SF-36 Mental Health scale; and the SF-12 Mental Component Summary scale. Correlationsamong these measures and health-related quality of life measures were examined. Responsiveness was evaluated by standardized response means, area under the curve (AUC) analyses, and treatment effect sizes in the trial. Results Convergent and construct validity was supported by strong correlations of the composite depression-anxiety measures with one another and moderate correlations with health-related quality of life measures, respectively. All composite measures differentiated patients who were better at 3 months, whereas the PHQ-ADS and PHQ-4 also distinguished the subgroup that had worsened. AUCs for composite measures were generally similar, whereas treatment effect sizes were largest for the PHQ-ADS. Limitations The study sample was predominantly male veterans enrolled from primary care who had chronic musculoskeletal pain and moderate levels of depression and anxiety. Conclusions Composite depression and anxiety scales are valid and responsive measures that may be useful as outcomes in research and clinical practice.