To assess the structure, effectiveness, and efficiency of existing modified-work programmes.

The databases searched included MEDLINE, PsycINFO and ABI. The searches were limited to empirical studies and reviews published in English from 1975 to March 1997. The title, abstract, and keyword fields of articles indexed were searched for the phrase 'return to work' combined with any of the following terms: 'modified', 'alternate', 'alternative', 'light', 'rehabilitation', and 'reemployment'. The references of retrieved articles and articles in the first author's personal library were also examined.

Study designs of evaluations included in the review

Aside from case reports, any reference with some empirical data on modified-work programmes was included in the descriptive review. However, only publications with a quality rating of 2.5 or higher were included in the evaluation of the effectiveness of modified-work programmes.

Specific interventions included in the review

The specific interventions were all forms of modified work programmes. These included light duty (19 studies), graded work exposure (4 studies), work trial (6 studies), supported employment (2 studies), and sheltered workshop as well as specific employer accommodations in the absence of any formal programme (4 studies).

Participants included in the review

The participants were hospital workers (6 studies), workers in industrial settings (10 studies), armed service workers (one study) and injured workers of unspecified worksites (12 studies).

Eleven studies included workers who experienced any type of injury, 11 dealt with back injuries, 4 concerned other musculoskeletal injuries, and 3 were limited to traumatic injuries of the brain or head.

Outcomes assessed in the review

Return to work was assessed.

How were decisions on the relevance of primary studies made?

Each article was reviewed independently by three researchers. Any discrepancies in these evaluations were discussed and consensus was reached.

The studies were rated on a scale of 0 to 5, according to an approach used by Kristensen (see Other Publications of Related Interest). The assessment was based on 5 methodological criteria: temporality; selection; measurement of exposure and outcome variables; confounding; and study design and statistical analysis.Each article was reviewed independently by three researchers.

The authors do not state how the data were extracted for the review, or how many of the authors performed the data extraction.

How were the studies combined?

Eleven studies (13 reports), which were reported to be of a moderate to high quality, were synthesised qualitatively. The return-to-work improvement factor, which refers to the ratio of return-to-work outcome measures in the intervention and comparison groups, was presented for 10 studies. This factor may represent the odds ratio or relative risk of return-to-work rates, or the ratio of average days lost, or the average injury costs. Some estimates were presented inversely so that a factor greater than one always signified improvement.

How were differences between studies investigated?

Each study was discussed in detail. A 'best' estimate of the programme effectiveness was derived from the results of the highest ranking studies, i.e. those meeting 4 out of the 5 quality criteria. A range of estimates of the programme effectiveness was then derived from the point estimate of the remaining higher-quality studies.

Twenty-nine articles, which reported data from 24 different study populations, were included in the overview to assess the structure of modified-work programmes. Eleven studies (13 references) were included in the evaluation of the effectiveness and efficiency of modified-work programmes. The review comprised a minimum of 12,905 participants.

The best quality studies reported a doubling of the return-to-work rates and/or the number of days worked when modified-work programmes were offered to injured workers. The range of the magnitude of the effect estimates across the 11 higher-quality studies was narrow, with a point estimate of 0.7 at the lower end and a factor of 4.5 at the higher end. The quality rating scores for these studies ranged from 2.5 to 4.0, with a mean score of 3.

Yes. The available data on the direct cost-savings ranged from 8% when taking the costs of the programme into account, to 90% when not taking them into account.

The main finding of this review was that modified-work programmes facilitated return to work for temporarily and permanently disabled workers. Specifically, studies rated high in methodological quality showed that employees with access to modified work, return to work after a disabling injury about twice as often as those without access to modified duty. The number of work days lost per disabling injury was also cut in half when companies implemented modified-work programmes. The available evidence also suggested that modified-work programmes were cost-effective.

Nearly all 29 studies reviewed showed positive results for the effectiveness of modified-work programmes in returning injured workers to the workplace. However, a number of methodological shortcomings were identified that need to be addressed in future research. In particular, the standardisation and qualification of modified-work programmes needs to be a goal. The reviewed programmes represented very different applications of modified work, and not all studies provided interpretable quantification of their result, either in terms of monetary or return-to-work variables. For those with interpretable quantified outcome measures, less than half tested the results statistically.

The literature search was fairly thorough, although it was limited to English language publications and may, therefore, have missed some important information. There were no clear inclusion criteria. Thus, the included studies varied considerably in terms of the type of modified work-programme investigated, the outcomes being measured, and the type of participants involved, which meant that pooling the data would be inappropriate. There was also a great variation in the type of study design included, with many studies not having a concurrent control and some no control at all. Information on how the quality scores were assigned to the individual studies was very limited. In addition, the return-to-work factor reported, which was used as a summary of the effect, was based on different types of statistical estimates and should be interpreted with caution.

The authors note that to thoroughly evaluate the efficiency of modified-work programmes, additional studies are needed that take all direct and indirect injury and programme costs into account in a comprehensive cost-benefit analysis. In addition, because modified work is considered a key element in the rehabilitation of temporarily or permanently disabled workers by most experts in the field, future research should also consider a needs assessment for educating employees, employers, and treating physicians about existing modified-work programmes. The authors suggested several improvements in research design, such as the use of concurrent control groups, measurement of confounding factors, longer follow-up time and more outcomes).

Industrial Medical Council of the State of California; California Commission on Health and Safety and Workers' Compensation.

Kristensen TS. Cardiovascular diseases and the work environment. A critical review of the epidemiologic literature on chemical factors. Scand J Work Environ Health 1989;15:245-64.

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.