A psychometric investigation on the diagnostic utility of the posttrauma risky behaviors questionnaire.

The Posttrauma Risky Behaviors Questionnaire (PRBQ) is a screening measure for posttrauma reckless and self-destructive behaviors (RSDBs). We examined (1) PRBQ's predictive relations with clinical (vs. not) endorsements of distinct RSDBs, and (2) PRBQ's optimal cutoff score yielding the most appropriate balance of sensitivity and specificity statistics. The sample included 354 adult trauma-exposed community participants (Mage=35.76 years; 57.90% female). Logistic regression analyses indicated that the PRBQ significantly differentiated individuals endorsing (vs. not) clinical levels of alcohol/drug misuse, disordered eating, problematic gambling, and compulsive buying. Receiver operating characteristic (ROC) curve analyses indicated that the 14-item PRBQ total score had moderate accuracy in differentiating individuals endorsing clinical vs. non-clinical levels of drug misuse, disordered eating, problematic gambling, compulsive buying, and engagement in RSDBs (PTSD's E2 Criterion); and low accuracy for alcohol misuse. ROC curve analyses indicated 3.5-6.5 as the optimal range of PRBQ cutoff scores. Thus, the PRBQ has good ability to discriminate participants endorsing (vs. not) clinical levels of alcohol/drug misuse, disordered eating, problematic gambling, and compulsive buying (lowest accuracy in identifying participants with clinical levels of alcohol misuse), and a minimum cutoff score between 3.5 and 6.5 may suggest using additional diagnostic assessments and clinical interventions.