The review concluded that integrated programmes for women with substance abuse issues and their children were associated with positive impacts on child development, growth and emotional and behavioural functioning. The authors' conclusion may not be reliable due to the limitations in the available evidence. The recommendation for future research was appropriate.

To examine the impact and effects of integrated programmes for women with substance abuse issues and their children.

PsycINFO, PubMed, Web of science, EMBASE, ProQuest Dissertations, Sociological Abstracts and CINAHL were searched from 1990 to May 2011 for studies published in English. Search terms were reported. Reference lists of retrieved articles were scanned and relevant journals were handsearched. Grey literature and conference proceedings papers were searched. Experts in the field were contacted.

Randomised and quasi-randomised controlled trials and cohort studies of substance abuse treatment programmes on women who were pregnant or parenting were eligible for inclusion. Treatment programmes had to include at least one specific substance use treatment and at least one parenting or child (<16 years) treatment service. Eligible studies had to report quantitative data on child outcomes or other outcomes.

The included studies comprised various combinations of group and individual addiction treatment programmes that addressed maternal mental health, parenting education and counselling, life skills training, prenatal education, medical and nutrition services, education and employment assistance, obstetrical and paediatric care, child care, children’s services and aftercare. The studies used various measurement tools for child development, growth and emotional and behavioural functioning outcomes. Assessed times were variable. Where reported, follow-up was for six or 12 months. Children were mostly assessed at ages in the range of three to 24 months; one study reported on older children.

Most studies were conducted in North America. The average age of participants ranged from 27 to 31 years. Most participants had experienced trauma such as childhood maltreatment, had mental health problems, were unemployed, single parents and were typically polysubstance users.

Two reviewers independently screened titles and abstracts of the studies identified in the searches. Any discrepancies were resolved by consensus.

Study quality of randomised trials was assessed using the Jadad scale of randomisation, blinding, withdrawals and drop-outs (maximum score 5). Non-randomised studies were assessed using the Newcastle-Ottawa Scale for study group selection, group comparability and outcome ascertainment (maximum score 9).

Two reviewers coded the study quality. Any discrepancies were resolved through consensus.

Data were extracted to calculate effect sizes (standardised mean difference; Cohen's d) which were used to express differential changes in pre and post studies or changes between intervention and control groups. Study authors were contacted for additional information.

One researcher extracted data. A second reviewer extracted data from 20% of studies and agreement was compared.

Studies were discussed in a narrative. Ranges of effect sizes within or between studies was reported. An effect size of 0.20 or less was considered small, 0.50 was medium and 0.80 was large.

Thirteen studies (eight cohort studies, three quasi experimental studies and two randomised controlled trials) were included in the review (775 participants). Two randomised controlled trials were rated moderate quality on the Jadad scale due to lack of blinding. The three quasi-experimental studies were rated low to moderate quality and the eight cohort studies were rated low on Newcastle-Ottawa Scale.

The number of participants in the pre-post studies ranged from nine to 80. There were small to large effect sizes for child development (d=0.007 to 1.132; two studies) and there were large effect sizes for emotional and behavioural functioning (d=0.652 to 1.132; two studies).

Three quasi-experimental studies compared integrated programmes to no treatment. Two of the studies appeared to have overlapping participants and neither had sufficient data to calculate effect sizes and statistical comparisons between groups were not available. In the remaining study, which appeared to have 19 participants, the effect sizes for development and most growth parameters (length, weight and head circumference) ranged from 1.16 to 2.48.

Two RCTs and one quasi-experimental study compared integrated to non-integrated programmes. One study found a small effect size for emotional and behavioural functioning (d=0.22 to 0.45) in favour of integrated programmes. A second study found effect sizes from 0.32 to 0.52 across a range of outcomes.

Several studies did not report sufficient data to enable calculation of effect sizes.

Integrated programmes for women with substance abuse issues and their children were associated with positive impacts on child development, growth and emotional and behavioural functioning.

The review addressed a clear question and was supported by broad inclusion criteria. Several sources were searched and attempts were also made to identify unpublished studies in English. The authors excluded 148 studies due to a lack of child outcomes data despite meeting all other inclusion criteria. It seemed unlikely that none of these trials assessed child outcomes: it was possible that such outcomes were gathered in at least some studies and not reported due to no significant effect of the intervention; therefore, there was a risk of outcome reporting bias in the group of trials under consideration in this review.

More than one reviewer was involved in study selection, quality assessment and data extraction which reduced the risk of error and bias. Included studies were quality assessed and most studies were of poor quality. The synthesis was limited by the small number of studies. Descriptions of the range of effect sizes in some instances referred to outcomes from a single study or across different lengths of follow-up.

The authors conclusion may not be reliable due to the limitations of the evidence available. The recommendation for future research was appropriate.

Practice: The authors did not state any implications for practice.

Research: The authors stated a need for more high quality studies to compare non-integrated to integrated programmes and examine various child outcomes with longer follow-up (at least two years). Moderator analysis could be used to identify the effective components of intervention. There should be full descriptions of the target population and intervention. The propensity score method should be used to address the potential problem of baseline differences between groups.

Canadian Institutes for Health Research (CIHR).

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.