Better evidence for a better world

Campbell evidence and gap maps

Coming soon – Campbell EGMs are a new evidence synthesis product. Plain language summaries of our EGMs will be published on this website. The interactive EGMs and full EGM reports will be available in our journal on the Wiley Online Library platform: click here.

Learn more about Campbell EGMs

Other EGMs

Campbell has produced maps on other topics, sometimes in partnership with other organisations.

See our other EGMs
12-step programs for reducing illicit drug use

Additional Info

  • Authors: Martin Bøg, Trine Filges, Lars Brännström, Anne-Marie Klint Jørgensen, Maja Karrman Fredriksson
  • Published date: 2017-02-15
  • Coordinating group(s): Social Welfare
  • Type of document: Title, Protocol, Review, Plain language summary, Data
  • See the full review:

About this systematic review

This Campbell systematic review examines the effectiveness of 12-step programs in reducing the use of illicit drugs. The review summarises findings from 10 studies, nine of which were conducted in the US.

What are the main results?

There is no difference in the effectiveness of 12-step interventions compared to alternative psychosocial interventions in reducing drug use during treatment, post treatment, and at 6- and 12-month follow-ups. 12-step programmes combined with additional treatment did have a significant effect at six-month follow-up, but this finding is based on few studies and is not found at 12-month follow-up.

There is some evidence that 12-step programmes retain fewer of their participants than other programmes, but the evidence has shortcomings. No effect was found on other secondary outcomes.


The effects of substance dependence have serious implications for the individual, the family and friends of the substance dependent individual, and society at large. Practitioners and public health policy makers have an interest in finding effective treatments that are also cost-effective. This review examined the effectiveness of 12-step programs aimed at illicit drug dependent participants compared to no intervention, treatment as usual, and other interventions.


The main objective of this review was to systematically evaluate and synthesise effects of 12-step interventions for participants with illicit drug dependence against no intervention, treatment as usual, and alternative interventions. The primary outcome of interest was drug use. Secondary outcomes of interest comprised criminal behaviour, prostitution, psychiatric symptoms, social functioning, employment status, homelessness and treatment retention.

Search methods

An extensive search strategy was used to identify studies meeting inclusion criteria. We searched electronic bibliographic databases in January 2010, October 2011, July 2013, August 2015, and September 2016. Searches for this review were performed on multiple international and Nordic databases. In total 11 databases were searched including PsycInfo, SocIndex, and Medline. A substantial range of grey literature sources were searched including governmental repositories, targeted web sites and trial registers. We checked the reference lists of primary studies, hand-searched relevant key journals, and searched the Internet using Google and Google Scholar. We also contacted researchers who had published in the area of 12-step interventions. Neither language nor date restrictions were applied to the searches. The conclusions of this review are based on the most recent searches performed September 2016.

Selection criteria

Studies had to meet the following criteria in order to qualify for inclusion in the review:

  • Intervention - only studies that considered 12-step interventions were eligible for inclusion.
  • Study Design - only studies using a RCT/QRCT design or a QES with a well-defined control group were eligible for inclusion.
  • Comparison - studies that compared 12-step to either no intervention or to other interventions were eligible for inclusion.
  • Participants - only studies where the drug of choice of participants was an illicit drug (established either by self-report or via clinician) were eligible for inclusion. Where only a subset of study participants were illicit drug users, a study was only eligible if it reported outcomes separately for the subgroup of illicit drug users.

Data collection and analysis

Descriptive and numerical characteristics of included studies were coded by one review author. A second review author independently checked coding, and any disagreements were resolved by consensus. We used an extended version of the Cochrane Risk of Bias tool to assess risk of bias of included studies. One review author evaluated the risk of bias of all included studies. A second review author independently checked the assessment and disagreements were resolved by consensus. Random-effects meta-analysis was used to synthesise effect sizes. We compared 12-step to other interventions, and 12-step with add-on to other interventions with the same add-on. For each comparison we conducted separate meta-analyses by time: during treatment, at treatment end, and at 6- and 12-month follow-up. Sensitivity of the results to risk of bias was assessed. Publication bias was assessed by the use of funnel plots.

Main results

The total number of potentially relevant records was 21,974 (database search: 17,416, grey literature search: 2,639, hand search and others: 1,919), of these 428 records were screened in full text. Thirteen reports met the inclusion criteria, with six reports contributing data on three independent studies. In total 10 studies were included in the review.

Seven of the included studies used a RCT design, two studies used a QRCT design, and one study used a QES design. One study, assessed as high risk of bias, was excluded from data synthesis. Thus, nine studies with a total of 1,071 participants contributed data to the analyses. These nine studies all considered outpatient settings where interventions were manual-based and delivered by trained therapists. In seven studies, treatment was partially or fully delivered in group therapy sessions. The reported statistical analyses were not corrected for this design element.

Seven studies contributed data to the comparison of 12-step intervention to alternative psychosocial interventions during treatment, at treatment end, and at 6-and 12-month follow-up. The seven studies did not all contribute data to all time points. Analyses did not reveal any statistically significant differences, for the primary outcome of drug use, between 12-step and the alternative set of interventions.

Three studies contributed data to the comparison of 12-step intervention with an add-on to alternative psychosocial interventions with an add-on. Drug use was assessed during treatment, post treatment, and at 6- and 12-months follow-up. All studies did not contribute data to all time points. We found no statistically significant effect size estimates during and post treatment. We found statistically significant effect size estimates at 6-month follow-up favouring 12-step with an add-on compared to alternative interventions with add-on (Hedges’ g =0.48, 95% CI: 0.06 to 0.90, and g=0.45, 95% CI: 0.03 to 0.88). No statistically significant effect size estimates were found at 12-months follow-up.

There was no strong indication of heterogeneity between studies (I2 did not exceed 75%). Results were robust to sensitivity analysis, and there was no observed evidence of publication bias.

Authors’ conclusions

The results of this review suggest that 12-step interventions to support illicit drug users are as effective as alternative psychosocial interventions in reducing drug use. This conclusion should be seen against the weight of evidence. A total of seven studies contributed data to analyses comparing 12-step interventions and alternative psychosocial interventions. The power to detect differences was low, and estimated effect sizes were small. In addition most studies delivered treatment as group therapy, but did not correct the analysis for the dependence between participants assigned to the same group.

Only one study reported results of the effects of self-help group attendance on drug use. This study was excluded from synthesis following the risk of bias assessment. Given the preponderance with which self-help 12-step interventions are delivered in practice, further evidence regarding the effectiveness of this type of intervention is needed.

See the full review

Contact us