Quality of reporting matched interventions for non-specific neck pain in randomised controlled trials and its association with trial outcomes: a secondary analysis of a systematic review

Background

Exercise and manual therapy are recommended for managing non-specific neck pain (NSNP), but most randomised controlled trials (RCT) have not identified specific subgroups that might respond best to each intervention. Although matched exercise was found to be superior to unmatched interventions in the short term, high heterogeneity in effectiveness estimates and variable levels of intervention fidelity and quality of delivery may contribute to inconsistent findings on the effectiveness of matched or stratified care.

Objectives

To assess the quality of intervention reporting and its relationship to treatment effect estimates in RCTs comparing matched versus unmatched exercise or manual therapy interventions for NSNP.

Design

Secondary analysis of a systematic review with meta-analysis.

Methods

For all included RCTs, independent reviewers rated the quality of intervention reporting using the TIDieR checklist. Meta-regression was used to assess the direction and magnitude of association between TIDieR scores and treatment effect estimates (standardised mean difference of change scores) for short-term pain and disability outcomes.

Results

The items related to intervention fidelity were not adequately reported. Materials used, intervention provider, location and modification were inconsistently reported. Poorer quality of intervention reporting was associated with larger treatment effect estimates in short-term pain and disability in favour of matched interventions.

Conclusions

Trials with inadequate intervention reporting may overestimate the benefits of matched exercise or manual therapy treatments for NSNP. The lack of SUFFICIENT information to judge intervention fidelity in this field reinforces the need for better guidance on this specific aspect of study design, conduct, and reporting.

Contribution of the Paper

• •
  Reporting of intervention fidelity in trials comparing matched versus unmatched exercise or manual therapy in non-specific neck pain is sub-optimal.
• •
  Trials with poorer completeness of intervention reporting tend to favour matched interventions