The authors concluded that the integration of substance abuse and mental health treatments appeared to improve a variety of outcomes, but further research is required since the included studies had methodological flaws. Overall, despite the limitations of this review, the authors' cautious conclusions appear appropriate in view of the reported methodological limitations of the included studies.

To evaluate the effectiveness of residential programmes for people with 'dual disorders', defined as severe mental illness with co-existing substance use disorders.

MEDLINE, PsycLIT, the Cochrane Library and Project CORK were searched. In addition, unpublished reports were sought using relevant specified national institutes and health departments of mental health, alcoholism, and drug and substance abuse.

Study designs of evaluations included in the review

Controlled studies were eligible for inclusion in the review. Randomised controlled trials (RCTs) and quasi-experimental studies were included.

Specific interventions included in the review

Studies of residential programmes were eligible for inclusion. The included studies evaluated a range of residential programmes: programmes with varying levels of demand, therapeutic communities, integrated out-patient programmes, dual- and single-disorder programmes and usual treatment. In the review, studies were classified as short term (average stay 6 months or less) or long term (average stay more than 6 months).

Participants included in the review

Inclusion criteria were not specified in terms of the participants, but it was clear that studies of people with severe mental illness and co-existing substance use disorders (dual disorders) were eligible for inclusion. The included studies involved homeless and incarcerated people; some studies included both men and women, others involved only men, and some involved people with psychotic disorders.

Outcomes assessed in the review

Inclusion criteria were not specified in terms of the outcomes. The review focused on the assessment of rates of retention, housing and substance abuse outcomes. In the included studies, outcomes were assessed from 1 month post-discharge to 2 years after baseline.

How were decisions on the relevance of primary studies made?

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

The authors reported methodological limitations of the included studies with reference to selection of participants, attrition, baseline comparability of treatment groups, sample size, description and implementation of treatment programmes, and representativeness of sample. The authors did not state how the validity assessment was performed.

The authors did not state how the data were extracted for the review, or how many reviewers performed the data extraction. The percentages of participants with the outcome of interest were reported for treatment group within each study.

How were the studies combined?

The studies were grouped by time period of outcome assessment (short and long term) and combined in a narrative with accompanying tables.

How were differences between studies investigated?

Differences between the studies were apparent from the text and tables.

Ten controlled trials (n=4,457) were included. The authors stated that although three of these were randomised controlled trials (RCTs), two were analysed as quasi-experimental studies because of methodological issues identified in the review. The other studies were described as quasi-experimental.

Methodological limitations of the studies included high rejection rates of referred people, high rates of attrition often due to self-selection, treatment groups not comparable at baseline, small sample sizes, samples not representative and poor definition of interventions.

Short-term residential programmes (5 studies).

One study reported that an integrated mental health and abuse programme was associated with higher rates of programme completion than a substance abuse programme, but reported no difference between treatments in substance abuse.

One study reported that an integrated lower-demand rehabilitation programme was associated with higher rates of programme retention and housing and lower rates of substance abuse relapse than a higher-demand therapeutic programme. One study reported that an integrated residential programme was associated with higher retention rates at 3 months than an out-patient programme. The retention rates were low in both treatment groups (24% versus 8%). There were no reported differences between treatment groups in terms of housing and substance abuse outcomes. One study reported that eight residential programmes classified as having a 'good dual diagnosis treatment climate'(analysed together) were associated with higher levels of coping and abstinence than seven residential programmes not classified as having a 'good dual treatment climate'. One study reported that dual and primary substance abuse programmes were associated with the same low rates of 'successful discharges'.

Long-term residential programmes (5 studies).

One study reported that a low-demand integrated dual treatment residential programme was associated with increased rates of retention at 2 months, 'successful discharge' and abstinence, compared with two modified therapeutic communities. One study (RCT) reported that a low-demand dual treatment community residential programme was associated with increased rates of retention when compared with a dual treatment modified therapeutic community, but there were no treatment differences in substance abuse outcomes. Both programmes rejected many referred people.

One study reported that a low-demand therapeutic community was associated with increased rates of retention at 1 year and improved outcomes (less substance abuse, criminal activity, lower HIV risk behaviour and higher employment) at 1 and 2 years compared with a high-demand therapeutic community. Both therapeutic community treatment programmes improved outcomes in comparison with usual out-patient treatment.

One study in incarcerated people reported no difference between a dual treatment programme and a therapeutic programme in re-incarceration rates or substance-related crime 1 year after release. Participation in a modified therapeutic community after release was associated with less re-incarceration, less criminal activity and less substance abuse-related criminal activity than non-participation. One study reported that a long-term integrated residential programme(unlimited length of stay) was associated with increased rates of treatment retention at 3 months and increased rates of abstinence, and reduced homelessness at 6 months compared with a short-term (intended stay 3 months) integrated programme.

A cost-analysis of one of the included studies reported that residential treatment costs were similar to out-patient treatment costs (due to increased costs of hospital attendance in the out-patient group).

Findings suggested that the integration of substance abuse and mental health treatments appeared to improve outcomes, but studies had methodological flaws and further research is required.

The review addressed a clear question in terms of the participants, intervention and study design; inclusion criteria were not specified for the outcomes of interest. Several relevant sources were searched and attempts were made to locate unpublished studies, thus limiting the potential for publication bias. The search strategy was not described in full; dates searched and search terms were not reported and no attempts to minimise language bias were reported. The methods used to select studies, assess validity and extract the data were not described, so it is not known whether any efforts were made to reduce reviewer errors and bias. Although methodological limitations of the included studies were summarised, the quality of the individual studies was not reported. Some aspects of study quality, such as the validity of methods used to assess outcomes (self-report or objective measures), were not reported.

The results were reported without levels of statistical significance and this made it difficult to interpret the evidence. Given the differences between the studies, a narrative synthesis was appropriate. However, the individual studies were described rather than the evidence being synthesised. Most of the included studies were described as quasi-experimental, which means that baseline differences between treatment groups may have contributed to differences in the outcomes. There were limitations to this review but, overall, the authors' cautious conclusions appear appropriate in view of the reported methodological limitations of the included studies, and the stated need for further research appears justified.

Practice: The authors stated that residential programmes should be flexible, supportive, low intensity and offered long term.

Research: The authors stated the need for further research to evaluate the effectiveness of residential programmes; to examine methods to increase rates of participation and retention; to identify the characteristics of people who may obtain most benefit from these programmes; and to determine the optimal duration of stay and discharge process. An RCT is greatly needed to evaluate a standardised programme in a well-defined population, with blind assessment of the outcomes.

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.