The Relationship Between Blinding Integrity and Treatment Efficacy in Randomised-Controlled Trials in Patients with Anxiety Disorders: A Systematic Review and Meta-Analysis
Year: 2024
Type: preprint
Study designs: Systematic Review, Meta-Analysis
Abstract: Background: Blinding of patients and assessors is thought to minimise expectancy effects and biases in double-blind randomised-controlled trials (RCT’s). However, whether blinding integrity should be assessed and reported remains debated. Furthermore, it is unknown whether blinding failure influences the outcome of RCT’s in anxiety disorders. We aimed to understand whether blinding integrity is assessed and reported in anxiolytic RCT’s. A secondary aim was to explore whether blinding integrity is associated with anxiolytic treatment efficacy.Methods: We systematically searched for placebo-controlled randomised trials in adults with generalised and social anxiety disorders, and in panic disorder, from 1980 to present day. We extracted data regarding blinding integrity and treatment efficacy. Where assessments of blinding integrity were not reported, attempts were made to acquire them from authors. Where possible, we subsequently calculated Bang’s Blinding Index, and assessed the association between blinding integrity and effect size of treatment compared with placebo through meta-regression.Results: Of the 248 RCT’s that met inclusion criteria, we were able to obtain assessments of blinding integrity from nine (3.63%). Overall, blinding failed in five of these trials (55.56%), but blinding was intact in 80% of placebo arms. We found a significant association between reduced blinding integrity among assessors and increased treatment effect size (beta’s < -6.50, p’s < 0.001), but this analysis involved only four studies. In patients, we saw a non-significant trend where reduced blinding integrity in the placebo groups was associated with increased treatment efficacy, which was not present in active medication arms.Conclusions: Consistent with work in other psychiatric disorders, blinding integrity is rarely reported in anxiolytic RCT’s. Where it is reported, blinding appears to often fail. We found signals that suggest unblinding of clinician assessors and of patients in placebo arms might be associated with larger treatment effect sizes. These analyses were based on limited data. We recommend that data regarding blinding integrity, along with the reasons patients and assessors offer for their beliefs regarding group allocation, are systematically collected in RCT’s of anxiolytic treatments.
Authors: Ruqayyah Haq, Laura Molteni, Nathan T M Huneke
Institutions: University of Southampton, National Health Service
Language: English
Subfield: Statistics, Probability and Uncertainty
Field: Decision Sciences
Domain: Social Sciences
Open Access status: gold
OpenAlex ID: W4393928383