• Adolescent substance abuse treatment;
  • meta-analysis


Aims  To examine the comparability, reliability and predictive validity of two instruments used to assess alcohol use and dependence: the Global Appraisal of Individual Needs (GAIN) and the Form 90 Timeline Followback (TLFB) method.

Design, setting and participants  Adolescents (n = 101) admitted to a residential treatment program in the United States were interviewed at intake with the GAIN, and again within a week with a variation of TLFB, called Form 90. Alcohol and cannabis measures were compared and used to predict the number of past-month substance abuse and dependence symptoms.

Measurement  Self-report measures of days of alcohol and cannabis use in the 90 days prior to intake, peak number of drinks/joints used, peak blood alcohol content (BAC) and alcohol and cannabis abuse and dependence symptom counts.

Findings  Results revealed that the measures had: (a) excellent comparability (r = 0.7–0.8) across the two instruments; (b) deteriorating reliability after reported peak BAC levels exceeded 0.50 and peak joints exceeding 19; and (c) similar and strong relationships between use measures and the number of abuse/dependence symptoms across measures and instruments.

Conclusions  In a sample of 101 adolescents who were admitted to residential treatment for alcohol or drug dependence, the corresponding measures from the two instruments produced comparable results. If the cross-validation of these two measures generalizes to adolescents treated in out-patient settings and other adolescent treatment populations, the GAIN and Form 90 may provide useful core alcohol measures for meta-analyses.