This review concluded that evidence supports the use of exercise and rest breaks in reducing musculoskeletal discomfort in computer tasks. Due to several limitations with the included studies, particularly the poor quality of the studies, and the sometimes poor reporting of the review, the authors' conclusions should be interpreted with caution as they may not be reliable.

To examine the evidence on the effects of exercise and rest breaks on musculoskeletal discomfort during computer tasks and compare the results with current clinical guidelines.

PubMed/MEDLINE and CINAHL were searched up to April 2005. DARE, Cochrane Database of Systematic Reviews, ACP Journal Club and the Cochrane Central Register of Controlled Trials were searched for articles in English. References of relevant retrieved articles were searched. Conference proceedings and articles published before 1990 were included if they were referenced in two or more relevant articles. Search terms were reported.

Articles analysing rest breaks and exercise independent of other variables in individuals with musculoskeletal discomfort (defined as a constellation of signs and symptoms that included muscle tension, paresthesia, pain and/or physical strain in soft tissues and bones) during computer tasks were eligible for inclusion. Eligible studies were required to describe the parameters of rest breaks. Studies examining multiple occupations that did not include computer use and studies involving exercise lasting more than 10 minutes were excluded. Review articles other than meta-analysis or systematic reviews were also excluded.

Included studies assessed exercise breaks, rest breaks or both types of breaks of varying regimens and durations. Some studies included a control group. The main outcomes reported were musculoskeletal discomfort and productivity, with musculoskeletal discomfort reported subjectively by participants. Some studies reported other outcomes such as mood state and eyestrain.

The authors stated neither how many reviewers selected the studies nor how discrepancies were resolved.

Included studies were assessed on validity using previously published criteria, including items on blinding, dropout/withdrawals, power calculations and study design. Studies received a validity score: six or seven represented strong internal validity (well conducted study); four or five depicted moderate internal validity (fairly conducted); and three or less represented weak internal validity (poorly conducted). One reviewer assessed the included studies, which were checked by a second reviewer. Disagreements resolved by consensus.

The authors stated neither how data were extracted nor how discrepancies were resolved.

Data were presented as a narrative synthesis and in tables, grouped by type of intervention and stratified by outcome result (beneficial or non-beneficial).

Fifteen randomised and non-randomised controlled trials were included in the review: three looked at exercise alone; seven looked at rest breaks alone; and five studies considered both. The total number of participants was not reported. Studies reporting sample sizes usually included less than 100 participants. The majority of studies were rated as having moderate or weak internal validity, with only two studies rated as strong. Some studies reported high drop-out rates.

Eight of nine study arms reported reduced musculoskeletal discomfort in computer tasks when exercise breaks were introduced. Ten of 13 study arms reported that rest breaks were effective in reducing musculoskeletal discomfort in computer tasks.

Some studies reported on productivity levels after exercise and/or rest breaks, but the findings appeared to be inconsistent. Some studies assessed other outcomes, which were reported in the review. Comparisons of findings with existing clinical guidelines were also reported in the review.

The evidence supported use of exercise and rest breaks in reducing musculoskeletal discomfort in computer tasks, with no additional benefits gained with exercise compared to rest breaks alone.

The objectives of the review were clear. The supporting inclusion criteria were fairly broad. Appropriate sources were searched for relevant articles, but it did not appear that unpublished papers were searched. Articles were restricted by language. It is possible that publication and language biases may have been introduced. Studies were assessed on validity, but quality of the studies was generally weak to moderate, which impacts on the reliability of the conclusions. The process for assessment of validity was clear, but study selection and data extraction methods were not, so reviewer error and bias cannot be ruled out. Considering the methodological differences between the included studies, a narrative synthesis was appropriate. However, data on study characteristics were limited and few details on patient characteristics, such as age and co-morbidities, were reported. It is, therefore, unclear whether patients were comparable. The limitations with the included studies, particularly the poor quality of the studies and the small sample sizes, and the sometimes poor reporting of the review, meant that the authors' conclusions should be interpreted with caution and may not be reliable.

The authors did not state any implications for practice or research.

Not stated.

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.