Screening tools for depression among community dwelling older adults in Chile: concordance and relative performance of the PHQ-2 compared with the PHQ-9

Depression is the most common mental health disorder in old age. In Chile, less than half of older adults with severe depressive symptoms have received diagnoses and treatment. Brief screening tools could increase the ability of health providers to detect possible cases of depression. The aim of this study was to assess the concordance of the PHQ-2 compared with the PHQ-9, which is a validated depression screening in older adults enrolled in primary care in Chile. Data from 253 community dwelling older adults were analyzed. Cohen’s kappa and Matthews correlation coefficient at different cutoffs, and area under the ROC curve, relative sensitivity, and relative specificity of the PHQ-2 with respect to the PHQ-9 were estimated. Logistic regression models were used to explore the association of sociodemographic, socioeconomic, and health variables with a positive screen for depression. Cutoffs of PHQ-2 ≥ 2 and PHQ-9 ≥ 6 showed the highest concordance (kappa and MCC = 0.61). Considering a cutoff of PHQ-9 ≥ 6 as reference, the PHQ-2 (≥ 2) had a relative sensitivity of 83.53% and relative specificity of 80.36%. A positive screen for depression with the PHQ-9 was more likely among people with a lower level of education and those with a negative self-assessment of health. Women were more likely to screen positive with the PHQ-2. Concordance between both scales was moderate. The performance and clinical utility of the PHQ-2 should be assessed against a gold standard in Chile. To promote equity in mental health access among older adults, the role of health status, level of education, and gender in the performance of these screening tools should be analyzed.