The Marijuana Screening Inventory (MSI-X): reliability, factor structure, and scoring criteria with a clinical sample.

Objective: This study continues the psychometric evaluation of a 31-item Marijuana Screening Inventory (MSI-X) with adults referred to a substance abuse clinic, by determining MSI-X reliability, factor structure, scoring cutoff accuracy, sensitivity, and specificity with a DSM-IV-TR criterion measure.

Method: A "marijuana inventory" containing demographic, MSI-X, and DSM IV-TR diagnostic items was administered to 107 adults undergoing substance-use evaluation.

Results: The MSI-X reliability was .90 with factor analysis deriving nine factors explaining 72.2% of the variance. Varimax rotation supports retention of all 31 items on nine factor-based scales. Receiver operating characteristic analysis determined MSI-X accuracy and cutoff points in relation to four DSM IV-TR diagnostic classifications. The MSI-X obtained the highest probability (.91) for accuracy in identifying both cannabis dependence and abuse, with six the optimal cutoff for maximum sensitivity (.83) and specificity (.89). Thus, MSI-X scores of > or = 6 are considered high risk. A cutoff score of 3 was associated with (probability, .90; sensitivity, .85; specificity, .81) identifying cannabis abuse only risk, providing a 3 to 5 score in the moderate risk range.

Conclusions: Clinical sample data supports the psychometric usefulness of the MSI-X as a screening tool. Marijuana lifetime use was 90% and past-year use 48%. The MSI-X identified 43% of lifetime users and 29% of past year users with moderate to high risk marijuana patterns deserving comprehensive evaluation. More males (15.9%) than females (7.5%) obtained MSI-X high-risk scores. The MSI-X empirically derived cutoff scores are within one point of the theoretical clinical cutoffs previously reported.