Utility of the Alcohol Use Disorder Inventory Test in determining caseness for alcohol use disorder among persons receiving antiretroviral therapy.

ABSTRACT

We sought to determine the utility of the Alcohol Use Disorder Identification Test (AUDIT) in screening for alcohol use disorder (AUD) among 688 users of antiretroviral therapy (ART). We used the receiver operating characteristic (ROC) curve analysis to assess whether the AUDIT reliably detected AUD against the Structured Clinical Interview for the DSM (SCID). Of the samples, 22.09% (CI 19.05% to 25.38%) of participants met the diagnostic criteria for AUD. Among men (n = 135) and women (n = 553), the prevalence estimates were 31.9% and 19.7%, respectively. For men, a cut-off score of 6 predicted AUD with 88.36% (95%CI 74.92-96.11) sensitivity and 88.04% (95%CI 79.61-93.88) specificity. For women, a cut-off score of 4 on the AUDIT yielded an optimal sensitivity of 92.66% (95% CI 86.05%-96.78%) and a specificity of 93.24% (95% CI 90.49-95.40%). For men, the AUDIT yielded a positive predictive value (PPV) of 75.55% (95% CI 66.26%-85.87%) and a negative predictive value (NPV) of 91.49% (95% CI 87.63-97.37%); for women the PPV and NPV were 77.09% (70.34%-82.68%) and 98.11% (96.37%-99.02%), respectively. The instrument's good sensitivity and specificity indicate it is likely to be useful for screening and referral of ART users who are probably cases of AUD.