Abstract
Background
Exercise is a complex intervention, and often has more than one treatment target. Results from randomised controlled trials (RCTs) of exercise interventions for chronic low back pain (CLBP) typically show small to moderate effect sizes, but these may differ where outcomes better reflect the targets of interventions. This review aimed to describe what treatment targets, outcome domains and primary outcome measures are used in exercise RCTs, and examined how well the selected outcome domains match the treatment targets used in each RCT.
Methods and Results
A computer-aided literature search was performed in eight databases, from inception to August 2018. Inclusion criteria: RCTs in CLBP, exercise compared to a non-exercise arm, sample size >60 per arm. Title and abstract review, subsequent full text review, data extraction and risk of bias assessment were independently undertaken by pairs of reviewers.
Of 18251 initially identified titles and abstracts, 23 trials were included in the review. 30 treatment targets were extracted, and 6 primary outcome domains identified. A logic model of the treatment targets and outcomes demonstrated diverse relationships. Only 5 RCTs matched their primary outcome domain to the identified treatment targets, 12 used primary outcomes that did not match the reported treatment targets, and 6 were partially matched.
Conclusion
The majority of included trials did not match the primary outcome domain to the treatment targets of the exercise intervention. Further research will explore if better matching between treatment targets and outcomes may change the conclusions of exercise trials in CLBP.
No conflicts of interest
Sources of Funding: This PhD is funded by the Research Institute for Primary Care and Health Sciences, Keele University. Prof NE Foster is a UK National Institute for Health Research Senior Investigator, and was supported by a UK National Institute for Health Research Professorship (NIHR-RP-011-015). The views expressed in this publication are those of the authors and not necessarily those of the NHS, the NIHR or the Department of Health.
