The Blind Leading the Not-So-Blind: A Meta-Analysis of Blinding in Pharmacological Trials for Chronic Pain.

Patient blinding is a critical feature of double-blind placebo-controlled randomized controlled trials (RCTs). Yet, very little is known about practices for assessing and reporting blinding in chronic pain trials. We examined the rates and predictors of assessing blinding and its success in pharmacological RCTs for chronic pain. Four-hundred eight trials (n = 103,983 participant) were identified via a systematic search between 2006 and 2016. Only 23 RCTs (5.6%) reported assessing patient blinding. Larger sample size, reference to a CONSORT statement, and pharmaceutical sponsorship were associated with lower rates of assessing blinding. A meta-analysis of the available data using Bang's Blinding Index indicated that blinding was not successful when combined across studies (g = 1.12, 95% confidence interval .92-2.01). Moderator analysis revealed that higher rates of adverse events and larger treatment effect sizes were associated with worse blinding outcomes, whereas including "don't know" responses seemed to improve blinding. Overall then, blinding is rarely reported and often fails in RCTs of pharmacological interventions for chronic pain. To address this finding, we recommend that all researchers conducting RCTs for chronic pain assess and report on the status of patient blinding when reporting the trial outcome. PERSPECTIVE: This meta-analysis examined patient blinding in pharmacological RCTs of chronic pain. The results indicated that blinding is rarely assessed and often fails. Some study characteristics were associated with lower rates of assessing blinding and its success, for example, pharmaceutical sponsorship and side effects. Implications and recommendations for chronic pain RCTs are discussed.