Diagnostic accuracy, internal consistency, and convergent validity of the Greek version of the patient health questionnaire 9 in diagnosing depression in rheumatologic disorders.

Abstract

The Patient Health Questionnaire 9 (PHQ-9) was developed to screen for depressive disorders in community, primary care, and medical settings. We aimed to estimate its diagnostic accuracy, internal consistency, reliability, and convergent validity in diagnosing major depressive disorder (MDD) in Greek patients with rheumatologic disorders. In a 2-phase sampling design study, we recruited 475 patients with established rheumatologic disorders. One of 2 of the high scorers (PHQ-9 score ≥9, n = 85) and 1 of 3 of the low scorers (PHQ-9 score 0-8, n = 128) were interviewed using the Mini International Neuropsychiatric Interview to confirm MDD. A receiver operator characteristic curve analysis was performed to confirm the optimum threshold value. The scale's dimensional structure was tested with factor analysis, and internal consistency reliability was assessed with Cronbach's alpha. Psychological distress (Symptom Check List-90-Revised [SCL-90-R]), disability (Health Assessment Questionnaire disability index), and health-related quality of life (HRQOL; World Health Organization Quality of Life Instrument [WHOQOL-BREF]) were also assessed to test convergent validity with bivariate correlations. At an optimum threshold of 10, the PHQ-9 showed a sensitivity of 81.2% and a specificity of 86.8%. The area under the curve was 0.91. The PHQ-9 presented unidimensional structure with good scale reliability (α = 0.82). The PHQ-9 score presented the greatest correlations with SCL-90-R depression (r = 0.736) and WHOQOL-BREF mental HRQOL scales (r = -0.571), and all other correlations with disability and HRQOL were in the expected direction. At a cutoff of 10, the PHQ-9 is an accurate, reliable, and valid measure for screening for MDD among Greek rheumatologic patients.