Reliability and validity evidence of a new interpretation bias task in patients diagnosed with drug use disorder: a preliminary study of the Word Association Task for Drug Use Disorder (WAT-DUD).

Background: Interpretation bias tasks such as word association tests have shown a moderate relation with substance use, but most studies have been conducted in nonclinical samples and these tasks are difficult to rate. Objectives: To provide: (1) reliability evidence of the Word Association Task for Drug Use Disorder (WAT-DUD), a novel and easy-to-rate instrument for measuring interpretation bias and (2) validity evidence based on the relationship between the WAT-DUD and variables associated with patterns of drug use and treatment outcomes. Methods: 186 patients (67 outpatients and 119 inpatients, 90% males) participated in the study. The task consisted of a simultaneous conditional discrimination where an image (either explicit or ambiguous) was the sample and two words (drug-related or not) served as comparison stimuli. The Substance Dependence Severity Scale, the Cocaine Craving Questionnaire-Now, and the Multidimensional Craving Scale were also used. Results: The ambiguous images items showed adequate reliability in terms of internal consistency (α = .80) and test-retest reliability (79.7% on average). The interpretation of images as drug-related was positively correlated with craving for cocaine (r = .20; p = .029), alcohol (r = .30; p = . 01), and alcohol withdrawal (r = .31; p = .01) along with severity of alcohol dependence (r = .23; p = .04). No relationship was found with the severity of cocaine dependence, or its symptoms of abstinence. Conclusion: WAT-DUD shows psychometric properties that support its use in research contexts, although more research is needed for its use in the clinical setting.