A test of the DSM-5 severity scale for alcohol use disorder.

Background: For the DSM-5-defined alcohol use disorder (AUD) diagnosis, a tri-categorized scale that designates mild, moderate, and severe AUD was selected over a fully dimensional scale to represent AUD severity. The purpose of this study was to test whether the DSM-5-defined AUD severity measure was as proficient a predictor of alcohol use following a brief intervention, compared to a fully dimensional scale.

Methods: Heavy drinking primary care patients (N=246) received a physician-delivered brief intervention (BI), and then reported daily alcohol consumption for six months using an Interactive Voice Response (IVR) system. The dimensional AUD measure we constructed was a summation of all AUD criteria met at baseline (mean=6.5; SD=2.5). A multi-model inference technique was used to determine whether the DSM-5 tri-categorized severity measure or a dimensional approach would provide a more precise prediction of change in weekly alcohol consumption following a BI.

Results: The Akaike information criterion (AIC) for the dimensional AUD model (AIC=7623.88) was four points lower than the tri-categorized model (AIC=7627.88) and weight of evidence calculations indicated there was 88% likelihood the dimensional model was the better approximating model. The dimensional model significantly predicted change in alcohol consumption (p=.04) whereas the DSM-5 tri-categorized model did not.

Conclusion: A dimensional AUD measure was superior, detecting treatment effects that were not apparent with tri-categorized severity model as defined by the DSM-5. We recommend using a dimensional measure for determining AUD severity.