Evidence Base on Outpatient Behavioral Treatments for Adolescent Substance Use: Updates and Recommendations 2007–2013

Abstract

This article updates the evidence base on outpatient behavioral treatments for adolescent substance use (ASU) since publication of the previous review completed for this journal by Waldron and Turner (2008). It first summarizes the Waldron and Turner findings as well as those from more recent literature reviews and meta-analytic studies of ASU treatment. It then presents study design and methods criteria used to select 19 comparative studies subjected to Journal of Clinical Child & Adolescent Psychology level of support evaluation. These 19 studies are grouped by study category (efficacy or effectiveness) and described for sample characteristics, methodological quality, and substance use outcomes. Cumulative level of support designations are then made for each identified treatment approach: ecological family-based treatment, group cognitive-behavioral therapy, and individual cognitive-behavioral therapy are deemed Well Established; behavioral family-based treatment and motivational interviewing are deemed Probably Efficacious; drug counseling is deemed Possibly Efficacious; and four integrated treatment models combining more than one approach are deemed Well Established or Probably Efficacious. The remainder of the article (a) articulates fidelity, mediator, and moderator effects reported for evidence-based approaches since 2008 and (b) recommends four enhancements to the prevailing business model of ASU outpatient services to accelerate penetration of evidence-based approaches into the underserved consumer base: pursue partnerships with influential governmental systems, utilize web-based technology to extend reach and control costs, adapt effective methods for linking services across sectors of care, and promote uptake and sustainability by emphasizing return on investment.

ACKNOWLEDGMENTS

We thank the CASAColumbia librarian, David Man, for his invaluable assistance in conducting the article search and reviewing abstracts for initial study inclusion.

Notes

^a Unless otherwise noted, study setting was outpatient treatment.

^b Codes for study category: II = Efficacy Study, III = Effectiveness Study.

^c Abbreviations used to indicate methodological attributes deemed missing from a given study, based on modified Nathan and Gorman (2002) quality of evidence criteria.

^d Assumed due to author report that 3% of sample did not complete follow-up survey and no indication that a missing data imputation or estimation method was used.

^e TAU-Plus: A TAU study condition that has been recognized by a federal agency as providing effective community-based treatment for ASU.

¹To derive a sample size threshold for the current review we followed Venter, Maxwell, and Bolig (2002), who investigated statistical power to detect differences across longitudinal measurements between randomized groups and concluded (1) statistical power in a 3-wave longitudinal design typically is equivalent to a 2-wave design in which the initial assessment is treated as a covariate in an ANCOVA model and (2) substantial increases in statistical power are not obtained over a 2-wave design unless at least five measurement points are included. Basing our sample size estimations on a 2-wave design, stipulating a moderate difference between groups regarding pre- to post-treatment change equivalent to Cohen's (1988) d = 0.50, and assuming a conservative correlation of 0.50 between longitudinal measurements, the GLMPOWER procedure in the SAS/STAT 9.3 software package determined that the sample size necessary to detect a difference between groups with power = 0.80 using an ANCOVA model with the post-treatment measurement as the dependent variable, a dichotomous treatment group indicator as the single independent variable, and the pre-treatment measurement as the single covariate, is 49 participants per condition (note that this sample size threshold diminishes given higher correlations between longitudinal measurements).

^a TAU-Plus: A TAU study condition that has been recognized by a federal agency as providing effective community-based treatment for ASU.

^b Recovery defined as living in the community (vs. being incarcerated, or residing in inpatient treatment or other controlled environment) and reporting no past month substance use, abuse, or dependence problems at the 12 month interview.

^a Cumulative support from studies reviewed in either Waldron and Turner (2008) or the current review.

^b Level of evidence designations for specific FBT-E models are listed following Waldron and Turner (2008).

^c Due to space considerations the table does not list all available comparative results for specific FBT-E models.