Translation and adaptation of the Alcohol-adapted Perceived Discrimination and Devaluation scale for the Tanzanian culture

Introduction: There is a lack of research on stigma about alcohol use disorders (AUDs) in low- and middle-income countries (LMICs). This study evaluated the psychometric properties of a Swahili version of the alcohol-adapted Perceived Devaluation-Discrimination scale (PDD) among an injury population seeking care in Tanzania. Method: Our sample consisted of 341 injury patients at the Kilimanjaro Christian Medical Center (KCMC) Emergency Department in Tanzania. The alcohol-adapted PDD was translated and back-translated to develop the Swahili version for assessing perceived alcohol stigma (PAS). Its reliability and factor structure were analyzed regarding evidence of content and internal structure validation. Results: 1-factor (perceived devaluation-discrimination) and 2-factor (perceived devaluation, perceived discrimination) confirmatory factor analytic models were performed. The 2-factor model adjusted for negatively worded items fitted the data best with the exclusion of 2 items that yielded low factor loadings (Comparative Fit Index = 0.99, Tucker-Lewis Index = 0.98, Root Mean Square Error of Approximation = 0.03). However, the adjusted 1-factor model had problematic factor loadings and did not show an adequate factor structure. Without adjusting for negatively worded items, both 1-factor and 2-factor models had good reliability and factor loadings, but did not fit the data well. Conclusions: An adjusted 2-factor solution of PAS, with 10 items, had superior parsimony. The alcohol-adapted PAS appears to be a psychometrically sound measure with high reliability and a good-fitting model structure in the Tanzanian context.