Construct validity of the PHQ-9 in university students in Colombia: A Rasch analysis approach

Introduction: The Patient Health Questionnaire 9 (PHQ-9) is one of the most widely used screening instruments for major depressive episodes. However, there are no published studies on Rasch-type analysis of the PHQ-9 among Spanish-speaking university students. Objective: To evaluate the psychometric properties of the PHQ-9 in university students using Rasch-type models and to assess possible biases of the items according to gender. Methods: This cross-sectional observational study evaluated the psychometric performance of the PHQ-9 in health sciences students at the University of Cartagena (Colombia). A random sampling stratified by academic program, semester, and sex was used, obtaining a sample of 550 participants (9 excluded for incomplete responses). Participants signed an informed consent, and the study was approved by an ethics committee. Rasch analysis was used to assess model fit, differential item functioning, dimensionality, local independence, and reliability. Adequate internal consistency (α=0.83, ω=0.89) and factorial validity were found. Results: A cross-sectional study was conducted with 550 health science students from Colombian university. The data were analyzed using a Rasch model, in which the following psychometric characteristics were verified: 1) differential item functioning, 2) dimensionality and local independence, and 3) overall fit. Only item 2 showed a tendency toward differential functioning. Conclusions: One-dimensionality and local independence of the items, moderate reliability, and good general fit were found, although there was a gap between the degree of depression measured by the PHQ-9 and the participants' responses. The Spanish version of the PHQ-9 for Colombian university students showed adequate item-level psychometric properties for screening for major depressive episodes.