Validity of the patient health questionnaire-2 (PHQ-2) for the detection of depression in primary care in Colombia

Background: The PHQ-2 is a screening test for major depressive disorder (MDD) derived from PHQ-9, which has shown to be useful in the detection of cases of clinical significance in previous studies. The psychometric properties of PHQ-2 in the Colombian population are unknown. Methods: PHQ-2 were assessed in 243 patients in general medical consultations using the Mini-International Neuropsychiatric Interview (MINI) as the gold standard. Internal consistency, convergent validity and criterion validity were calculated by analyzing the Receptor Operating Characteristics (ROC) and the area under the curve (AUC). Results: Cronbach's alpha and McDonald's omega coefficients were 0.71. Spearman's rho coefficients for correlations with PHQ-9 and HADS-D scores were 0.63 and 0.59 (p<0.01). AUC was 0.89. The optimal cut point was >= 2 with the following indicators: sensitivity 0.87 specificity 0.74; Youden index 0.60; PPV 0.47; NPV 0.95; (LR +) 3.24; (LR-) 0.18 (95% CI 0.09-0.37). Kappa coefficient between PHQ-2 and MINI for depression was .458 and 0.46 for HADS-D. Limitations: Since this study was done with people attending hospital, which could has implications for the prevalence of depression, affecting the validity indicators of the instrument. Conclusions: PHQ-2 show an acceptable test performance in the context of the study. However, the test could yield a considerable amount of false positives that would require specialized evaluation to establish a reliable diagnosis.