Diagnostic validity of drinking behaviour for identifying alcohol use disorder: Findings from a representative sample of community adults and an inpatient clinical sample.

Abstract

Alcohol consumption is an inherent feature of alcohol use disorder (AUD), and drinking patterns may be diagnostically informative. This study had three aims: (1) to examine the classification accuracy of several individually analysed drinking behavior measures in a large sample of US community adults; (2) to extend the findings to an adult clinical sample; and (3) to examine potential sex differences. In cross-sectional epidemiological and clinical datasets, receiver operating characteristic (ROC) curves were used to evaluate diagnostic classification using area under the curve (AUC), accuracy, sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV). Two samples were examined: a large random sample of US community adults who reported past-year drinking (n = 25 773, AUD = 20%) and a clinical sample from a Canadian inpatient addiction treatment centre (n = 1341, AUD = 82%). Classifiers included measures of quantity/frequency (e.g. drinks/drinking day, largest drinks/drinking day, number of drinking days and heavy drinking frequency). The clinical criterion (reference standard) was AUD diagnostic status per structured clinical interview (community sample) or a symptom checklist (clinical sample). All drinking indicators were statistically significant classifiers of AUD (AUCs = 0.60-0.92, Ps<0.0001). Heavy drinking frequency indicators performed optimally in both the community (AUCs = 0.78-0.87; accuracy = 0.72-0.80) and clinical (AUCs = 0.85-0.92; accuracy = 0.77-0.89) samples. Collectively, the most discriminating drinking behaviours were number of heavy drinking episodes and frequency of exceeding drinking low-risk guidelines. No substantive sex differences were observed across drinking metrics. Quantitative drinking indices appear to perform well at classifying alcohol use disorder (AUD) in both a large community adult and inpatient sample, robustly identifying AUD at rates much better than chance and above accepted clinical classification benchmarks, with limited differences by sex. These findings broadly support the potential clinical utility of quantitative drinking indicators in routine patient assessment.