Are Sinhala and Tamil PHQ-9 versions comparable across the Sri Lankan population? Psychometric equivalence across translations and between demographic subgroups

Abstract

Background: Although indications suggest that Sri Lanka is facing serious mental health challenges within its diverse populations, the extent of the problem is poorly quantified and even tools required to explore its nature are unavailable. The present study tests the psychometric equivalence of the PHQ-9, a widely used measure of depression, across two translated versions to Sinhala (Sinhala-English) and Tamil (Tamil-English) languages. It also tests equivalence across different demographic groups. Methods: The study uses data from a nationally representative sample of 6,444 individuals, from the 2018-19 Wave 1 of the Sri Lanka Health and Ageing Study, who completed the PHQ-9 along with other measures of health and wellbeing. For analysing the data, standardized process of measurement invariance testing was performed using multigroup confirmatory factor analysing technique. Results: Results indicated equivalence across the two translated versions, and across gender, ethnicity, level of education, and socioeconomic status. Configural, scalar and strict invariances were established successfully for each variable. Conclusions: The translated PHQ-9 tools were found to perform psychometrically similarly across diverse populations in Sri Lanka. These versions of the PHQ-9 can therefore be used for screening of depressive symptoms and research purposes in Sri Lanka irrespective of population characteristics in terms of language, gender, ethnicity, level of education or socioeconomic status without concerns over whether they are equivalent.