Background Comorbid depression and substance use disorders are common and have poorer outcomes than either disorder alone, While effective psychological treatments for depression or substance use disorders are available, relatively few randomised controlled trials (RCTs) have examined the efficacy of these treatments in people with these comorbid disorders. Objectives To assess the efficacy of psychological interventions delivered alone or in combination with pharmacotherapy for people diagnosed with comorbid depression and substance use disorders. Search methods We searched the following databases up to February 2019: Cochrane Central Register of Controlled Trials, PubMed, Ernbase, CINAHL, Google Scholar and clinical trials registers. All systematic reviews identified, were handsearched for relevant articles. Selection criteria The review includes data from RCTs of psychological treatments for people diagnosed with comorbid depression and substance use disorders, using structured clinical interviews. Studies were included if some of the sample were experiencing another mental health disorder (e.g. anxiety); however, studies which required a third disorder as part of their inclusion criteria were not included. Studies were included if psychological interventions (with or without pharmacotherapy) were compared with no treatment, delayed treatment, treatment as usual or other psychological treatments. Data collection and analysis We used standard methodological procedures expected by Cochrane. Main results Seven RCTs of psychological treatments with a total of 608 participants met inclusion criteria. All studies were published in the USA and predominately consisted of Caucasian samples. All studies compared different types of psychological treatments. Two studies compared Integrated Cognitive Behavioural Therapy (ICBT) with Twelve Step Facilitation (TSF), another two studies compared Interpersonal Psychotherapy for Depression (IPT-D) with other treatment (Brief Supportive Therapy (BST) or Psychoeducation). The other three studies compared different types or combinations of psychological treatments. No studies compared psychological interventions with no treat ment or treatment as usual control conditions. The studies included a diverse range of participants (e.g. veterans, prisoners, community adults and adolescents). All studies were at high risk of performance bias, other main sources were selection, outcome detection and attrition bias. Due to heterogeneity between studies only two meta-analyses were conducted. The first meta-analysis focused on two studies (296 participants) comparing ICBT to TSF. Very low-quality evidence revealed that while the TSF group had lower depression scores than the ICBT group at post-treatment (mean difference (MD) 4.05, 95% confidence interval (Cl) 1.43 to 6.66; 212 participants), there was no difference between groups in depression symptoms (MD 1.53, 95% Cl-1.73 to 4.79; 181 participants) at six- to 12-month follow-up. At post-treatment there was no difference between groups in proportion of days abstinent (MD-2.84, 95% Cl-8.04 to 235; 220 participants), however, the ICBT group had a greater proportion of days abstinent than the TSF group at the six- to 12-month follow-up (MD 10.76, 95% Cl 3.10 to 18.42; 189 participants). There were no differences between the groups in treatment attendance (MD-1,27, 95% Cl-6,10 to 3.56; 270 participants) or treatment retention (RR 0.95, 95% Cl 0.72 to 1.25; 296 participants). The second meta-analysis was conducted with two studies (64 participants) comparing IPT-D with other treatment (Brief Supportive Psychotherapy/Psychoeducation). Very low-quality evidence indicated IPT-D resulted in significantly lower depressive symptoms at posttreatment (MD-0,54, 95% Cl-1.04 to S-0,04; 64 participants), but this effect was not maintained at three-month follow-up (MD 3.80, 95% Cl -3.83 to 11.43) in the one study reporting follow-up outcomes (38 participants; IPT-D versus Psychoeducation). Substance use was examined separately in each study, due to heterogeneity in outcomes. Both studies found very low-quality evidence of no significant differences in substance use outcomes at post-treatment (percentage of days abstinent, IPD versus Brief Supportive Psychotherapy; MD-2.70, 95% Cl -28.74 to 23.34; 26 participants) or at three-month follow-up (relative risk of relapse, IPT-D versus Psychoeducation; RR 0.67, 95% Cl 0.30 to 1.50; 38 participants). There was also very low-quality evidence for no significant differences between groups in treatment retention (RR 1.00, 95% Cl 0.81 to 1.23; 64 participants). No adverse events were reported in any study. Authors' conclusions The conclusions of this review are limited due to the low number and very poor quality of included studies. No conclusions can be made about the efficacy of psychological interventions (delivered alone or in combination with pharmacotherapy) for the treatment of comorbid depression and substance use disorders, as they are yet to be compared with no treatment or treatment as usual in this population. In terms of differences between psychotherapies, although some significant effects were found, the effects were too inconsistent and small, and the evidence of too poor quality, to be of relevance to practice.
