Successful Blinding or Wishful Thinking? A Secondary Analysis of a Randomised Controlled Trial of Manual Therapy.

BACKGROUND

Blinding is rarely assessed, estimated, or reported in randomised controlled trials (RCTs). To provide contextualised understanding of participant blinding in a placebo-controlled RCT of spinal manipulative therapy (SMT) for chronic primary low back pain (NCT05162924), we aimed to: (i) reassess blinding, (ii) explore treatment effects and adverse events by levels of beliefs, and (iii) examine whether blinding varied by treatment response.

METHODS

Blinding was assessed immediately after the first intervention session, at the two- and four-week follow-ups (primary endpoint). The sum Bang blinding index (BI), ranging from -2 to 2, was the primary blinding outcome for this secondary analysis. Treatment effects and adverse events were explored in blinded and unblinded participants. BIs were estimated by levels of treatment response after randomisation.

RESULTS

Of 102 randomised participants, 98 received their intervention as allocated and were included in this secondary analysis. Correct beliefs about intervention assignment in the active SMT group were balanced by incorrect beliefs in the placebo SMT group ('wishful thinking'), suggesting adequate blinding at the study level (sum Bang BI: 0.13 [95% CI, -0.22 to 0.48]). Treatment effects were markedly larger in unblinded participants, while adverse events were comparable among blinded and unblinded participants. BIs varied by treatment response.

CONCLUSION

Study-level blinding of participants was adequate in this randomised placebo-controlled trial of SMT. While treatment effects were markedly larger in unblinded participants, adverse events were comparable in blinded and unblinded participants. Blinding varied by treatment response, providing insights for the design and conduct of future trials.

SIGNIFICANCE STATEMENT: This secondary analysis of blinding in a placebo-controlled RCT of SMT found that blinding at the study level was adequate. Treatment effects were markedly larger in unblinded participants, highlighting the importance of blinding procedures in RCTs. Blinding findings by responder status enhance ongoing discussions regarding the feasibility of blinding in this area. Overall, insights from this secondary analysis encourage the thoughtful consideration of blinding procedures in RCTs of SMT and other physical interventions for pain.