Summary of findings tables
| Summary of findings: exercise alone compared with standard care for knee replacement (long-term outcome only) | |||||
| Patient or population: adult patients scheduled for total knee replacement Setting: clinic-supervised exercise (five studies); home-based exercise (two studies) Intervention: 3–12 weeks’ exercise conducted in the presurgical period Comparison: standard care | |||||
| Outcome No of participants (studies) | Relative effect (95% CI) | Illustrative comparative risks (95% CI) | Certainty | Comments | |
|
Long-term pain
Assessed with: various scales (WOMAC*, KOOS†, HSSK‡) Follow-up: range 6–12 months No of participants in meta-analysis: 229 (6 RCTs) No of participants not in meta-analysis: 122 (1 RCT) |
Meta-analysis: SMD 0.2 higher (-0.06 to 0.47) Narrative ‘Non-significant difference’ | Mean KOOS scores in the untreated group ranged from 73.2 to 85.95 in the control group; mean HSSK scores in the control group were 27; mean WOMAC scores in control group were −2.3 to −4.8 | The mean level of pain after exercise was 0.20 SD lower (−0.06 to 0.47 lower). | ⨁⨁⨁◯ MODERATE§¶ | There was no clear evidence of a difference between exercise and standard care for long-term pain either from data from the meta-analysis or from a comparatively large study not included within the meta-analysis. We rescaled the SMD finding to a commonly used instrument in the field - the KOOS, using estimations based on the mean of a representative study within the analysis (Huber 2015) and our pooled SMD of 0.02 (−0.06 to 0.47 higher). This indicates that exercise led to a mean increase of 2.7 units/points on the KOOS (95% CI −0.8 to 6.29) in the treated group. Sensitivity analysis of two studies at low risk of bias suggested a smaller effect (mean increase of 1.46 points on the KOOS (95% CI-4.58 to 7.50) |
| Adverse events | See comment | See comment | See comment | ⨁⨁◯◯ LOW§** | Two studies reported treatment-related adverse events (increase in pain) separately from the main pain outcomes, which occurred during the period of intervention. Information on perioperative complications, for example, superficial and deep infections, were reported in all studies. Events were few and attribution to treatment status difficult |
|
GRADE: Working group grades of evidence
High certainty: We are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: We are moderately confident in the effect estimate: The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: Our confidence in the effect estimate is limited: The true effect may be substantially different from the estimate of the effect. Very low certainty: We have very little confidence in the effect estimate: The true effect is likely to be substantially different from the estimate of effect. | |||||
*WOMAC (Bellamy et al 1988).55 The original pain scale runs from 0 to 20 and a higher number indicates greater pain. There are several modifications. Some trialists also reverse the polarity and/or transform responses into a 0–100 scale where a high number indicates less pain. We standardised to this latter method as it was most common and was similar to the KOOS and the HSSK.
†KOOS (Roos et al 1998).56 The scale ranges from 0 to 100 and a higher number indicates less pain.
‡HSSK (Insall et al 1976).58 Pain is measured on a scale of 0–30, with a high number indicating less pain.
§Downgraded due to risk of bias associated with lack of blinding of both personnel and participants in all studies also concerns about sequence generation/allocation concealment in three studies.
¶We did not downgrade for imprecision as findings of the one large study not included in metaanalysis were very similar to the six studies included in meta analysis.
**We downgraded the quality of evidence for study limitations, in particular imprecision: estimate based on few events.
GRADE, Grading of Recommendations Assessment, Development and Evaluation; HSSK, Hospital Society Score; KOOS, Knee Injury and Osteoarthritis Outcome Score; MD, mean difference; SMD, standardised mean difference; WOMAC, Western Ontario and McMaster Universities Osteoarthritis Index.