Evidence-Based Psychosocial Treatments for Adolescent Substance Abuse

Abstract

This review synthesized findings from 17 studies since 1998 regarding evaluation of outpatient treatments for adolescent substance abuse. These studies represented systematic design advances in adolescent clinical trial science. The research examined 46 different intervention conditions with a total sample of 2,307 adolescents. The sample included 7 individual cognitive behavior therapy (CBT) replications (n = 367), 13 group CBT replications (n = 771), 17 family therapy replications (n = 850) and 9 minimal treatment control conditions (n = 319). The total sample was composed of approximately 75% males, and the ethnic/racial distribution was approximately 45% White, 25% Hispanic, 25% African American, and 5% other groups. Meta-analysis was used to evaluate within-group effect sizes as well as differences between active treatment conditions and the minimal treatment control conditions. Methodological rigor of studies was classified using Nathan and Gorman (2002) criteria, and treatments were classified using criteria for well-established and probably efficacious interventions based on Chambless et al. (1996). Three treatment approaches, multidimensional family therapy, functional family therapy, and group CBT emerged as well-established models for substance abuse treatment. However, a number of other models are probably efficacious, and none of the treatment approaches appeared to be clearly superior to any others in terms of treatment effectiveness for adolescent substance abuse.

This research was supported in part by grants from the National Institute on Drug Abuse (R01 DA11955; R01DA13350; R01DA13354) and National Institute on Alcohol Abuse and Alcoholism (R01 AA12183). We also thank Hyman Hops and Lynette F. Cofer for editorial feedback on drafts of this article.

Notes

Note: tx = treatment; SUD = substance use disorder; CD/ODD = conduct disorder/oppositional defiant disorder; CBT-I = cognitive behavioral therapy-individual; BFT = behavioral family therapy; MI + CBT = motivational interviewing plus CBT; CBT-G = CBT-group; MET/CBT12 = motivation enhancement therapy/CBT-12 sessions; MET/CBT5 = MET/CBT-5 sessions; ACRA = adolescent community reinforcement approach; MDFT = multidimensional family therapy; MST = multisystemic therapy; SET = structural ecosystems therapy; FAM = family therapy; SAU = service as usual; BSFT = behavioral strategic family therapy; SOFT = strengths oriented family therapy; 7C = seven challenges; TFT = transition family therapy; AGT = adolescent group therapy; FFT = functional family therapy; IBFT = integrated behavioral and family therapy.

Note: We chose not to include some of the data from the Robbins et al. (in press). The authors raised concerns about the validity of these findings due to the high differential (across condition) attrition rate for the African American participants. ES = effect sizes; CBT-I = cognitive behavioral therapy-individual; MET/CBT12 = motivation enhancement therapy/CBT-12 sessions; MET/CBT5 = MET/CBT-5 sessions; CBT-G = CBT-group; na = not applicable; FSN = family support network; ACRA = adolescent community reinforcement approach; AGT = adolescent group therapy; MEI = multifamily educational intervention; SET = structural ecosystems therapy; FAM = family therapy; TLFB = timeline followback; BSFT = behavioral strategic family therapy; 7C = seven challenges; SOFT = strengths oriented family therapy; TFT = transition family therapy; FFT = functional family therapy; IBFT = integrated behavioral and family therapy; MM12 = minnesota 12 step; TFLB = timeline followback procedure.

¹The current analysis did not control for initial within-replication differences in drug use, as we did not have a direct estimate for each study condition. However, if we had statistically controlled for these initial differences (assuming a within replication correlation of .55, a value which was based on our analysis of more than 500 adolescents receiving similar substance abuse treatments), the between treatment modality would have been increased by approximately 30%. We derived the .55 correlation from a pooled estimate from our own study of nearly 500 adolescents in substance abuse treatment programs. As a consequence, the family, group CBT, and individual CBT modalities would each be significantly different from the minimal treatment control condition.

∗References marked with an asterisk indicate studies included in the meta-analysis.