The Validity and Reliability of the PHQ-9 and PHQ-2 on Screening for Major Depression in Spanish Speaking Immigrants in Chile: A Cross-Sectional Study

Abstract

Background: The study aimed to explore the psychometric properties of two versions of the Patient Health Questionnaires (PHQ-9 and PHQ-2) on screening for Major Depressive Disorder (MDD) among Spanish-speaking Latin American adult immigrants in Santiago, and to explore factors associated with a higher risk of occurrence of MDD among them. Methods: A representative sample of 897 Spanish-speaking immigrants completed the PHQ-9. The Composite International Diagnostic Interview (CIDI) was employed to evaluate MDD. Internal consistency and structural validity were evaluated using Cronbach’s α coefficient and confirmatory factor analysis (CFA). Convergent validity with the 7-item General Anxiety Disorder Scale (GAD-7) was assessed using Spearman’s correlations. Sensitivity, specificity, positive predictive values, and area under the receiver operating characteristic (ROC) curve were calculated for different cut-off points. Logistic regression analysis was used to identify factors associated with the risk of MDD. Results: Cronbach’s α coefficient of the PHQ-9 was 0.90; item-total correlation coefficients ranged from 0.61 to 0.76 and correlation with the GAD-7 was moderate (r = 0.625; p < 0.001). CFA on three alternative models suggests a plausible fit in the overall sample and among two of the subsamples: Peruvians and Venezuelans. Taking the results of CIDI as the gold standard for MDD, the area under the ROC curve was 0.91 (95% confidence interval (CI): 0.83~1.0). When the cut-off score was equal to 5, values of sensitivity, specificity, and Youden’s index were 0.85, 0.90, and 0.75, respectively. Multivariate logistic regression analyses showed that the influence of having three or more children (OR = 3.91, 95% CI: 1.20~12.81; p < 0.05), residency in Chile of up to three years (OR = 1.79, 95% CI: 1.07~3.00; p < 0.05), active debt (OR = 2.74, 95% CI: 1.60~4.70; p < 0.001), a one (OR = 2.01, 95% CI: 1.03~3.94; p < 0.05) and two or more events of adversity during childhood (OR = 5.25, 95% CI: 1.93~14.3; p < 0.01) on the occurrence of MDD was statistically significant. Reliability (α = 0.62), convergent (r = 0.534; p < 0.01) and criterion (AUC = 0.85, 95% CI: 0.67~1.00) validity coefficients of the PHQ-2 were weaker than for the PHQ-9. Conclusions: The PHQ-2 and the PHQ-9 are reliable and valid instruments for use as screeners for MDD among Spanish-speaking populations of Latin America.