Psychometric properties of the Spanish version of the Cannabis Use Problems Identification Test among Chilean university students: A validation study

Abstract

BackgroundIn Chile, concerns mount about escalating cannabis use. Thus, it is important to have tools for early identification of at-risk users. The Cannabis Use Problems Identification Test (CUPIT) is a useful screening tool, and the aim of this study was to examine the psychometric properties of its Spanish version among Chilean university students. MethodsThe CUPIT was translated into Spanish, pre-tested in a focus group (n=8), and then tested through an online survey (n=3798, 28% response rate). Of the 1061 respondents, 578 reported 12-month cannabis use. Internal reliability, internal structure, and concurrent validity (using the Cannabis Abuse Screening Test [CAST]) were obtained. Test–retest reliability was calculated (n=150) at 3–4 weeks (30% of attrition rate). Discriminative validity was evaluated comparing CUPIT subscales and four DSM-IV diagnostic groups. Receiving operator characteristic (ROC) curve analysis assessed sensitivity and specificity. ResultsTest–retest Pearson correlation between total CUPIT scores of 0.90 (p<0.001), and highly significant Kendall Tau-b coefficients for individual items (p<0.001) indicated excellent reliability. Concordance between the CUPIT and CAST (Pearson correlation 0.73, p<0.001) indicated good concurrent validity. ANOVA revealed significant differences in CUPIT scores between the four DSM-IV diagnostic groups (p<0.001), indicative of good discriminative validity. ROC analysis (gold standard: DSM-IV abuse/dependence) yielded an AUC value of 0.72, indicating acceptable discriminative capability. ConclusionsThe Spanish CUPIT is reliable, valid, and accepted by the university population studied, and, thus, a potentially useful tool for identifying both problematic and at-risk users.