Psychometric validation of the Diagnostic Assessment Research Tool: Alcohol use disorder module

Abstract

Aims: Structured clinical interviewing is considered the gold standard in psychiatric diagnosis. The Diagnostic Assessment Research Tool (DART) is a novel modularized, non-copywritten, semi-structured interview; however, no studies have examined the psychometric properties of its alcohol use disorder (AUD) module. The primary aims of this study were to: (i) validate the factor structure of the DART AUD module and (ii) examine measurement invariance across several key demographic and subgroup factors. Methods: Participants were community members in Hamilton, Canada and Boston, USA who self-identified as making a significant AUD recovery attempt (N = 499). Internal reliability was examined via the Kuder-Richardson 20 statistic, and correlations between symptom count and drinking quantity/frequency were examined. Then, symptom-level data were included in a confirmatory factor analysis to examine model fit of a single hypothesized factor structure. Finally, measurement invariance analyses were conducted for sex, age, ethnicity (White vs. racialized), and study site. Results: This study found evidence for adequate internal reliability (rKR20 = 0.75), and symptom scores correlated with drinking quantity and frequency (r = 0.16-0.43). Confirmatory factor analysis results suggested excellent fit for the unidimensional one-factor AUD model (χ2 = 0.09, confirmatory factor index = 0.99, Tucker Lewis index = 0.99, standardized root mean square residual = 0.06, root mean square error of approximation = 0.02). Measurement invariance analyses revealed that the factor structure was equivalent between sex, age, ethnicity, and study site. Conclusions: Findings provide strong evidence for the psychometric validity of the DART AUD module and support its use in research and clinical practice. The DART represents a credible alternative to other diagnostic interviewing tools for AUD.