This generally well-conducted review concluded that several psychosocial interventions were effective in treating adolescents with substance abuse. The authors' conclusions are likely to be reliable. However, the potential for selection bias, and the lack of details as to which individual studies were used to grade the interventions, should be kept in mind.

To assess the effectiveness of interventions for substance abuse in adolescents.

The following sources were searched for studies published in English: MEDLINE (1988 to 2003), PsycINFO (1988 to 2003), Social Science Abstracts (1988 to 2003), Criminal Justice Abstracts (1966 to 2003), the Cochrane Database of Systematic Reviews, the Cochrane Controlled Trials Register and the C2 registries of the Campbell Library. A National Library of Medicine computerised bibliographic search was also conducted. Alcohol and drug treatment websites were also searched, as were references lists of identified studies, recent relevant books and government documents. The search terms were reported.

Study designs of evaluations included in the review

Controlled clinical trials were eligible for inclusion.

Specific interventions included in the review

Studies that compared a substance abuse intervention with a control treatment (either waiting-list control or another treatment) were eligible for inclusion. Studies of pharmacological treatment were only included if this was part of a psychosocial intervention. The included studies evaluated family therapy, multisystemic therapy, cognitive-behavioural therapy, coping skills training and life skills training. The interventions lasted between 4 and 44 weeks (details were reported).

Participants included in the review

Studies that targeted adolescents were eligible for inclusion. Studies targeting both adults and adolescents were only included if the results were reported separately for adolescents. The participants in the included studies were typically aged 14 to 21 years (modal age 15 years) and were predominantly male. They were also predominantly white, with African Americans being well represented and Latinos also included. About half of the studies were of juvenile offenders, probationers or youths referred by courts, and most participants were in lower socioeconomic groups.

Outcomes assessed in the review

Studies that used a measure of drug or alcohol use were eligible for inclusion. Studies that only assessed compliance, safety, other problem behaviours or prevention-only outcomes were not included. The included studies assessed alcohol and/or drug use, and months of abstinence, mostly using self-reports. Some studies obtained verification of the reports from parents, urinalysis and arrest records.

How were decisions on the relevance of primary studies made?

The authors did not state how studies were selected for the review, or how many reviewers performed the selection.

Validity was assessed and scored using a modified version of the Methodological Quality Rating Scale described by Miller et al. (reference given). This assesses study design, replicability, baseline characteristics, quality control, length of follow-up, dosage of intervention, collateral verification, objective verification, drop-outs/attrition, statistical power, independent assessment of the outcomes, statistical analyses and the number of study sites. Potential scores range from 1 to 16. Studies scoring 1 to 8 points were considered to be less methodologically strong, while those scoring 9 to 16 points were considered to be relatively strong methodologically. Two reviewers independently assessed validity. Inter-rater agreement was measured.

Two reviewers independently extracted and coded the data into a coding form. Inter-rater reliability was assessed. One reviewer checked all coding forms for accuracy and any discrepancies were resolved by discussion. Where sufficient data were reported, the effect size (ES; Cohen's d) was calculated as the difference in post-treatment scores between treatments. Where only the results of statistical tests were reported (t, F or r), the ES was estimated from these. ESs were classified as small (0.20), medium (0.50) or large (0.80). Interventions were classified according to quality and ES: ES>0.20 with at least 1 year' follow-up or replication and using strong study design; ES>0.20 with relatively strong study designs and less than 1 year' follow-up and no replication; negligible or unwanted effect with less strong study design; negligible or unwanted effect with relatively strong study design; and indeterminate effect, mixed or incomplete findings.

How were the studies combined?

Study details and key findings were tabulated, and a brief narrative synthesis was provided.

How were differences between studies investigated?

Study characteristics were tabulated and differences between the studies were discussed in the text.

Fifteen studies (n=1,928) reported in 18 articles were included. Thirteen of these were randomised controlled trials.

The quality of the studies was high (mean score 12.0, range: 8 to 15). Inter-rater agreement was 0.96 for the data extraction and coding and 95% for the validity assessment.

Many treatments reduced substance abuse and increased abstinence. The ES ranged from 0.51 for coping skills training to -1.25 for behavioural therapy. Benefits found immediately after treatment were often not sustained at follow-up, with ESs ranging from 0.39 for cognitive-behavioural therapy to -0.87 and -0.86 for cognitive-behavioural therapy and multidimensional family therapy, respectively.

The interventions were graded as follows.

ES>0.20 with at last 1 year' follow-up or replication and using strong study design: multidimensional family therapy and cognitive-behavioural group treatment.

ES>0.20 with relatively strong study designs and less than 1 year' follow-up and no replication: behavioural therapy, multisystemic therapy, combined cognitive-behavioural therapy and functional family therapy, family systems therapy, functional family therapy, combined Botvin life-skills with additive programmes, and psychoeducational therapy.

Negligible or unwanted effect with less strong study design: supportive group counselling, interactional group treatment, aftercare services and residential treatment services.

Negligible or unwanted effect with relatively strong study design: support for individual counselling, family education, adolescent group treatment and individual cognitive-behavioural treatment.

Several psychosocial interventions were effective in treating adolescents with substance abuse. The strongest available evidence was for multidimensional family therapy and cognitive-behavioural group therapy.

The review addressed a clear question that was defined in terms of the participants, intervention, outcomes and study design. Many relevant sources were searched and attempts were made to locate unpublished studies, thus minimising the potential for publication bias. The restriction to English language studies might have resulted in the omission of some relevant data, which the authors acknowledged. Methods were used to minimise reviewer errors and bias in the assessment of validity and extraction of data, but it was unclear whether similar steps were taken at the study selection stage. Validity was assessed using specified criteria and the results were reported.

Combining studies using a hierarchy of evidence based on study quality and ES appeared appropriate. Explicit details of which studies were used to grade the evidence for each intervention would have enabled the grading to be verified. Keeping this and the potential for selection bias in mind, the review was generally well-conducted and the authors' conclusions are likely to be reliable.

Practice: The authors stated that it is important that practitioners use interventions for which there is the greatest evidence.

Research: The authors stated that further controlled trials are required to assess the longer term effects of treatment for substance abuse in adolescents, particularly those with co-existing disorders who take medication, heavy-substance abusers and substance-dependent youths. The authors also stated the need for greater social work research in this field.

This is a critical abstract of a systematic review that meets the criteria for inclusion on DARE. Each critical abstract contains a brief summary of the review methods, results and conclusions followed by a detailed critical assessment on the reliability of the review and the conclusions drawn.