Background: Brief depression screening questionnaires may increase detection of depression in primary care settings but there have been few validation studies carried out in typical populations in low-income countries.
Methods: Cultural validation of the Patient Health Questionnaire (PHQ-9/PHQ-2), the 20-item Self-Reporting Questionnaire (SRQ-20) and the Kessler scales (K6/K10) was carried out in 306 adults consecutively attending primary care facilities in small towns in Ethiopia. To assess criterion validity, the gold standard assessment for presence of Major Depressive Disorder (MDD) was made by Ethiopian psychiatric nurses using the Mini International Neuropsychiatric Interview.
Results: The prevalence of gold standard MDD was 5.9%, with irritability more common than depressed mood or anhedonia. The area under the receiver operating characteristic curve indicated good performance of the PHQ-9, SRQ-20, K6 and K10 (0.83-0.85) but only fair for the PHQ-2 (0.78). No cut-off score had acceptable sensitivity combined with adequate positive predictive value. All screening questionnaires were associated with disability and the PHQ-9 and SRQ-20 were associated with higher health service contacts, indicating convergent validity. Construct validity of all scales was indicated by unidimensionality on exploratory factor analysis.
Limitations: Test-retest reliability was not assessed.
Conclusions: Brief depression screening questionnaires were found to be valid in primary care in this low-income country. However, these questionnaires do not have immediate applicability in routine clinical settings. Further studies should evaluate utility of indicated screening embedded within health system changes that support MDD detection. Investigation of irritability as a core depression symptom is warranted.
Keywords: Depression; Developing country; Primary health care; Screening; Sub-Saharan Africa; Validation.
Copyright © 2015 Elsevier B.V. All rights reserved.