Four studies involving 594 participants were included in this review. The studies assessed the effectiveness of six psychosocial interventions: CBT, TSF, BI, HHP, MI, and BMI. There was significant clinical and methodological heterogeneity among the included studies, which precluded meta-analysis. Comparing different psychosocial interventions, there was only one study for each comparison. Most of the comparisons were not statistically significant, except for decreased alcohol use at three months and nine months in the BI study . Surprisingly, these results favored the control intervention. This could be interpreted in light of the main limitations of this study, namely, the standard intervention provided to the control group was ‘too strong’ to enable reasonable comparison with the intervention group, and the intervention group had a high proportion of people with alcohol addiction who received the 15-minute-long brief alcohol intervention. This is in contradiction to the manual for BIs, which states that people with alcohol addiction should not receive BI, but should be referred to a specialized, more intensive treatment . Also other systematic reviews examining the general population indicated that BI was effective for harmful/ hazardous use, but not for dependence [12, 15]. Finally, participants receiving BMI were significantly more likely to reduce their alcohol use by seven or more days in the past 30 days at six months’ follow-up, compared to the control group .
Our review was systematic, but not without weaknesses. We did not limit our searches to studies published in English; however, studies in non-English languages may have been missed because they are commonly less frequently indexed in the selected databases. Unpublished studies may also have been missed. Unpublished studies are likely to have negative results, which is why they are not published. The major limitation of the review process was that most trials did not provide enough published data or did not provide data in a form that could be extracted for meta-analysis. Although the lead authors of all four studies were emailed, only two responded and provided further data. Furthermore, we could not include a number of potentially relevant studies, because they involved drug users without problem alcohol use in their samples.
Similar to our work, two previous narrative reviews were unable to identify evidence to answer our question or to conduct a meta-analysis [17, 18]. Subsequently, they based their conclusions on evidence coming from mixed-type studies (for example, case studies and RCTs) or studies that included illicit drug users without a concurrent problem alcohol use. We excluded these types of studies. Furthermore, the review by Arias et al.  discussed 14 reports/studies related to treatment of co-occurring alcohol and cocaine/opioid dependence, two of which were included in our review.
This review is unintentionally tapping into an important question: what constitutes standard maintenance/outpatient treatment? It appears that all standard treatments contain some type of psychosocial support, which varies considerably, and this makes it difficult to evaluate the added value of additional services. This was true for studies included in our review and, in addition, the process of assessment or quick feedback following the assessment, or both, resulted in improved alcohol outcomes among the participants.