Selecting a depression measure for research: A critical examination of five common self-report scales

Abstract

Depression is frequently assessed in science, as it is known to contribute to, exacerbate, and result from other health declines, but selecting among numerous depression instruments can be challenging for researchers. Here, we critically examine and compare self-report depression measures, describing theoretical foundations, psychometric features, and performance across diverse populations of adults-salient considerations for selection in research. Five commonly used depression measures were evaluated: the Beck Depression Inventory (BDI), the Center for Epidemiological Studies Depression (CES-D), the Hospital Anxiety and Depression Scale (HADS), the Patient Health Questionnaire (PHQ-9), and The Patient Reported Outcome Measures Information System Depression (PROMIS-D). Each measure was critically evaluated according to the Consensus-based Standards for the Selection of Health Measurement Instruments (COSMIN) properties. Variations in theoretical basis and COSMIN quality were evident across the measures. Only the PROMIS-D received COSMIN ratings of 'very good' or 'acceptable' across all properties. Content and structural validity were deemed 'inadequate' or 'doubtful' for the BDI, PHQ-9, and HADS. Adequacy of criterion and construct validity varied across measures. HADS received 'doubtful' or 'inadequate' ratings across all properties. All but the PROMIS-D received 'doubtful' ratings for internal consistency. The CES-D and PROMIS-D received the highest cross-cultural validity ratings. Continued use of depression measures that lack theoretical grounding, robust psychometrics, or equivalence across diverse populations should be reconsidered. The PROMIS-D scale demonstrates sound theoretical foundations, superior COSMIN ratings, and robust scoring. Research to test cross-cultural validity of measures across race and ethnicity is needed to advance health for all.

Keywords: Depression; Health equity; Measurement; Mental health; Psychometrics; Symptoms.