Objective: In sub-Saharan Africa, large proportions of patients who are on antiretroviral therapy (ART) engage in excessive alcohol use, which may lead to adverse health consequences and may go undetected. Consequently, health care workers need brief screening tools to be able to routinely identify and manage excessive alcohol use among their patients. Various brief versions of the valid and reliable 10-item Alcohol Use Disorders Identification Test (AUDIT) (i.e., the AUDIT-C, AUDIT-3, AUDIT-QF, AUDIT-PC, AUDIT-4, and m-FAST) may potentially replace the full AUDIT in busy HIV care settings. This study aims to assess the utility of these six brief versions of the AUDIT relative to the full AUDIT for identifying excessive alcohol use among patients in HIV care settings in South Africa.
Method: Participants were 188 (95 women) patients from three ART clinics within district hospitals in the City of Tshwane Metropolitan Municipality who reported past-12-month alcohol use. Performance of each brief AUDIT measure for identifying excessive alcohol use was evaluated against that of the full AUDIT (with a cutoff score of ≥6 for women and ≥8 for men) as the gold standard. We used receiver-operating characteristic (ROC) analysis.
Results: Most brief AUDIT measures had an area under the receiver operating curve (AUROC) above .90 when compared with the full AUDIT (five of six for women and three of six for men). The AUDIT-PC, AUDIT-4, and m-FAST had the highest AUROCs, whereas the three brief measures comprising only consumption items had low specificities at the most optimal cutoff levels.
Conclusions: Various brief versions of the AUDIT may be appropriate substitutes for the full AUDIT for screening for excessive alcohol use in HIV clinics in sub-Saharan Africa.