Abstract
Purpose of the study and background
A 2011 Cochrane review concluded that spinal manipulative therapy (SMT) is no better than other interventions for reducing pain and improving function in chronic low back pain (CLBP). Using individual participant data (IPD) from trials has advantages, among others: a more precise estimate of the effect and the potential to identify moderators. Our objective was to assess the effect of SMT in adults with CLBP and to identify relevant moderators.
Methods
All trials from the 2011 Cochrane review were included in this IPD. We updated the search (April 2016) IPD from eligible studies was requested. Primary outcomes were pain intensity (VAS/NRS) and back-specific function (RMDQ). Risk of bias was assessed. For the treatment effect, an one-stage approach (mixed model technique, intention-to-treat principle) was used; a second-stage approach was conducted as confirmation. For the moderator analyses, one-stage approach was conducted for 19 variables.
Results
We obtained IPD from 21 RCTs (n= 4638). Most studies were pragmatic comparing the SMT effect to recommended therapies (e.g. exercise). Effects for SMT were slightly, but not clinically better compared to other therapies for primary outcomes.
For the moderator analysis, 8 moderators showed a statistical significant effect at a specific time point (e.g. for SMT vs recommended therapy: age, both pain and function at 26 weeks). But for none of the moderators a consistent effect was identified across time points.
Conclusion
Findings of the Cochrane review were confirmed. No clinically-relevant moderators were identified for SMT compared to other therapies for pain and function at any follow-up measure.
No conflicts of interest
Sources of funding: European Chiropractic Research fund
