This review concluded that cognitive behavioural therapy was usually more effective than usual treatment at improving return to work rates for individuals with mental health problems and/or physical injuries. Limitations of the included data and the risk of publication bias suggested that the findings should not be considered as reliable.

To assess the impact of psychological return-to-work interventions for individuals with mental health problems and/or physical injuries on work and health outcomes.

MEDLINE, EMBASE, CINAHL, PsycINFO, Cochrane Central Register of Controlled Trials and Cochrane Database of Systematic Reviews were searched from 1985 to 2000 for studies written in English or French and published in peer reviewed journals. Search terms were reported and a full search strategy was available from the authors

Studies that assessed return-to-work oriented interventions with the aim of improving the ability of employees on sick leave, with or without work-related physical injuries, to cope with or manage mental health problems were eligible for inclusion in the review. Interventions had to be carried out either in the workplace or a primary care setting. All eligible participants had to be 100% employed or have guaranteed employment and be absent from work due to work-related causes. Interventions with job-seeking components were excluded.

Included studies focused on mental health problems and/or physical injuries. Half of the studies were carried out in European countries. Most were in mixed gender populations. Cognitive behavioural therapy was often the main component of interventions. Other types of psychosocial interventions were included (such as coping strategies, problem-solving and belief/attitude adjustments). All of the studies considered outcomes on an individual participant level and almost half also considered other levels such as the organisation. Outcomes varied between studies, but included return-to-work status, sick leave duration and health-related quality of life.

Two reviewers reached a consensus decision about the eligibility of each study..

The authors stated that two independent reviewers assessed the quality of the studies using published criteria and reached a consensus decision about each study. However, the quality assessment consisted only of awarding each study a design rating (between one and five stars) based on the type of study design used (randomisation and use of a control group or another lesser design).

One reviewer extracted the study data and a second reviewer checked the accuracy of the data. Any discrepancies were discussed until a consensus was reached.

Studies were combined according to design and summarised in a narrative synthesis with accompanying data tables.

Fourteen studies were included in the review: four randomised controlled trials (RCTs) (n=796); two controlled studies (n=105); five uncontrolled studies (n=592); one evaluation (n=100); and two case studies (n=1 and n=100). Two out of four RCTs reported positive work outcomes and three reported positive mental health outcomes. One of the two controlled trials reported positive work outcomes, but neither trial reported positive mental health outcomes. Two thirds of participants in the five uncontrolled trials returned to work, although the studies used different outcome definitions and measured outcomes over different periods of time. Three out of four uncontrolled studies also reported positive mental health outcomes. One evaluation study reported that 84% of individuals returned to work at least partially after one year (73% returned fully). One case study reported that all participants returned to work.

Cognitive behavioural therapy was typically more effective than usual treatment at improving return-to-work rates.

This review answered a clear research question using a broad range of studies and interventions. A number of electronic databases and other sources were searched for relevant studies, but language and publication biases may be present (particularly given the topic area) as only French or English studies published in peer-reviewed journals were eligible for publication. The reviewers made appropriate attempts to reduce reviewer error and bias during the review process. They carried out an assessment of study quality, but this appeared to be crudely based on the overall type of study (such as randomised, controlled or uncontrolled), rather than a true assessment of bias within the individual studies. Most of the included evidence appeared to be of a poor design and likely to be at risk of bias. There also appeared to be variation between studies in terms of the types of participants, interventions and outcome definitions, which made it difficult to compare data between studies. The authors' narrative approach was appropriate, but their crude process of vote counting and combining return-to-work rates using differently defined outcomes was unlikely to be reliable. The conclusions also focused on a comparison with usual control, but less than half of the studies included a control group and these studies often reported mixed findings. Overall, the limitations of the included data and the risk of publication bias suggest that the findings should not be considered as reliable.

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

Practice: The authors suggested that future studies should combine several key components and integrate them into return-to-work interventions aimed at employees with mental health problems and/or physical injuries. Future interventions should engage employees emotionally or psychologically as soon as possible after mental/physical problems developed. Communication between employee and organisation was also important.

One of the review authors received support from the Michael Smith Foundation for Health Research and a New Investigator Award from the Canadian Institutes of Health Research.

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.