Reliability and validity of the Brief DSM-5 Alcohol Use Disorder Diagnostic Assessment: A systematic replication in a crowdsourced sample

Effectively and efficiently identifying alcohol use disorder (AUD) is an essential goal for researchers and clinicians alike. To date, there are a limited number of self-reported tools specifically designed for evaluating DSM-5 criteria for AUD. The Brief DSM-5 AUD Diagnostic Assessment is a recently created participant self-reported measure with demonstrated reliability and validity in college student populations. The purpose of this study was to replicate and extend these findings by evaluating the psychometric properties of the Brief DSM-5 AUD Diagnostic Assessment in a geographically diverse adult sample with varying alcohol use history. Participants (N = 1986) were sampled from the crowdsourcing website Amazon Mechanical Turk (mTurk). The Brief DSM-5 AUD Diagnostic Assessment was completed to evaluate scale reliability and validity. A subset of individuals with answers indicative of AUD (N = 448) also completed a battery of standardized alcohol use questionnaires for additional validity assessments. High internal consistency reliability was observed (alpha = 0.92) along with strong item-total correlations. Convergent validity was supported by significant positive relationships between diagnostic category and measures of alcohol use consumption and severity. Modest relationships that were generally not statistically significant were observed with soda use measures supporting discriminant validity. AUD severity remained a significant predictor of alcohol use measures after accounting for AUDIT scores, thereby establishing incremental validity. These findings collectively support the Brief DSM-5 AUD Diagnostic Assessment as a reliable, valid, and rapid self-report tool for evaluating DSM-5 AUD in diverse research and clinical settings.