To assess the efficacy, effectiveness and safety of exercise for the treatment of knee osteoarthritis (OA).

MEDLINE and CINAHL were searched from 1979 to 1999 using the search terms 'osteoarthritis', 'exercise' and 'clinical trial'. Additional papers were identified by searching through referenced papers. The search was restricted to English language publications.

Study designs of evaluations included in the review

No a priori study designs were stated by the authors, although one of the search terms was 'clinical trials'. The studies reported were RCTs, non-controlled trials, and case studies. Several studies performed a follow-up assessment; these were undertaken at various times in the different studies, i.e. 2, 3, 6, 8, 9, 12, and 18 months post-treatment.

Specific interventions included in the review

Hospital-based non-functional exercises such as inner range quadriceps contractions over a small block, to exercises mimicking function such as sitting to standing exercises. A number of the included studies provided walking programmes for participants. The randomised controlled trials (RCTs) included comparisons of educational programmes, sham treatments, home exercises, electrotherapy, phone contact, or another unstated control. The exercise programmes ranged from 4 to 16 weeks in duration. The average treatment duration was 18.7 hours (standard deviation 13.7) over 8.8 weeks (standard deviation 3.5). In Europe and Australia, all of the exercise programmes were administered by physiotherapists. In the USA, programmes were delivered by physical therapists or athletic trainers.

Participants included in the review

Patients with OA of the knee, as demonstrated by radiological evidence. The majority of the trials were confined to patients with primary OA, where screening blood tests were commonly performed to exclude inflammatory joint disease. Some of the included studies required participants to have had joint pain for more than 4 months, to be taking non-steroidal anti-inflammatory medication, and to have evidence of OA in other joints. Patients were excluded if their general health was a contraindication to exercise therapy, if they had recently received an intra-articular steroid injection, or if they had recently participated in a similar type of exercise programme.

Outcomes assessed in the review

No a priori outcomes were specified by the authors. The outcomes reported in the review were:

reduction in dependency;

reductions in functional, pain and strength scores;

improvement in aerobic capacity, balance ability, and sensorimotor function at the knee;

increase in walking speed, stair climbing ability, and quadricep strength;

improvement in pain scores; and

improvement in the Arthritis Impact Measurement Scale scores.

Exacerbation of pain was also reported.

How were decisions on the relevance of primary studies made?

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

The authors do not state that they assessed validity.

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

Data were extracted for the following categories: study identification; sample size; treatment duration; whether home exercises were given; whether weight bearing exercises were given; outcomes; follow-up assessments; and comments of the review's authors.

How were the studies combined?

A narrative synthesis was undertaken, where studies were grouped according to study design and subgrouped according to type of exercise.

How were differences between studies investigated?

The authors do not state a method for assessing any differences between the studies.

Nineteen studies with a total of 1,069 participants were included in the review: 11 RCTs, 6 uncontrolled trials, and 2 case studies.

RCTs (n=11).

Non-weight bearing, non-functional exercises are, in the worst-case scenario, ineffective, and at best, may lead to improvements in muscle strength and reduction in pain scores, but not necessarily improvements in functional capacity.

Weight bearing exercise trials showed mixed results, although when compared with the group receiving the health education programme, the largest trial in this group showed significant improvements in functional capacity and pain scores for both the resistance exercise and aerobic exercise groups. At 18-months follow-up, the significant improvements were still evident, although the degree to which improvement had been maintained was related to the degree to which patients were compliant with home exercising.

Uncontrolled trials (n=6).

Benefits were found, but the conclusions that could be drawn were limited by limitations in the study designs. The benefits found included: improvement in muscle strength, functional capacity, and reduction in pain with the use of a progressive resistance exercise programme in one series of studies; and improvement in pain scores and function in a second study. Two single case studies demonstrated improvements in both muscle strength and functional capacity.

The authors stated that exercise programmes of at least 5 weeks' duration produced greater improvement than those of a shorter duration. Exercise programmes with treatment times of over 4 hours also produced greater improvement than those with shorter times. The addition of a home exercise element to an exercise programme led to greater improvement. Exercise programmes that provided weight-bearing exercise, such as walking or programmes that required the participants to undertake functional mimicking exercises, were more effective than those that employed non-functional, non-weight bearing exercise.

Exacerbations.

Only a very small percentage of participants were reported as experiencing exacerbations of their pain.

The authors state that exercise treatment when provided by physiotherapists in a hospital setting, is safe, reduces pain, improves function, and increases strength. The exercise classes are well attended and do not exacerbate symptoms in the vast majority of patients. There was no evidence to suggest that exercise programmes cause acceleration of the disease process.

The authors stated the research question, and the inclusion and exclusion criteria were quite thorough. The literature search was limited. The search was restricted to English language publications and there was no attempt to find unpublished or grey literature. It is possible that additional relevant studies may have been missed. There were no tests for publication bias.

The quality of the included studies was not formally assessed. In addition, the authors did not report how the articles were selected, or who performed the selection and data extraction.

The extracted data were reported in tabular format, then summarised and discussed in the text of the review. The studies were not statistically combined. Heterogeneity was discussed, in that no single trial was directly comparable with another.

The authors' conclusions appeared to follow from the results, but should be viewed with caution because of the limitations in the quality of the review process.

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

Research: The authors state that many of the trials in the literature are flawed and more work is required to establish the optimum types, frequency, duration and intensity of exercise programmes for patients with OA of the knee.

NHS R&D HTA Programme.

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.