Clinical Evaluation of the Alcohol Use Disorders Identification Test (AUDIT) in Moshi, Tanzania

Background: Alcohol use disorder is a major cause of morbidity and mortality in low- and middle-income countries. Alcohol screening using a validated tool is a useful way to capture high-risk patients and engage them in early harm reduction interventions. Our objectives were to 1) evaluate the psychometric evidence the Alcohol Use Disorders Identification Test (AUDIT) and its subscales in the general population of Moshi, Tanzania, and 2) evaluate the usefulness of the tool at predicting alcohol-related harms.

Methods: Two hundred and fifty-nine adults living in Moshi, Tanzania were included in the study. We used the AUDIT and its subscales to determine the classification of harmful and hazardous drinking. To analyze the internal structure of AUDIT and the model adequacy we used Confirmatory Factor Analysis (CFA). The reliability of AUDIT was analyzed for Cronbach's alpha, Omega 6 and Composite Reliability. The optimal cut off point for the AUDIT was determined by the receiver operating characteristic (ROC) curve, using the Youden approach to maximize sensitivity and specificity.

Results: The median score of the AUDIT was 1 (inter-quartile range: 0-7). The internal structure of the AUDIT showed factor loadings ranging from 0.420 to 0.873. Cronbach's alpha, Omega and Composite Reliability produced values above 0.70. The Average Variance Extracted was 0.530. For the AUDIT, a score of 8 was identified as the ideal cut-off value in our population.

Conclusions: This study validates AUDIT in the general population of Moshi and is one of the only studies in Africa to include measures of the internal structure of the AUDIT and its subscales.