Placebo run-in periods in anticholinergic trials are not associated with treatment effect size or risk of attrition: an empirical evaluation.

Abstract Objective We aimed to explore the impact of run-in periods on the magnitude of treatment effect and risk of attrition in a sample of randomised trials. Study design and setting We identified randomised trials from a published systematic review examining the effects of anticholinergics for the treatment of overactive bladders. We fitted meta-analytic mixed-effects models to assess whether the type of run-in (placebo run-in vs no run-in) was associated with the magnitude of the effect estimates for the following outcomes: the number of voids per day, number of leakages per day, presence of dry mouth, cure/improvement, patient withdrawal from the trial, compliance with the trial protocol and/or adherence to study drug. We adjusted for potential confounders. Results A total of 96 trials met the eligibility criteria; 59 trials had no run-in (which included those with a screening or withdrawal period), 37 trials had a placebo run-in, and no trials had a drug run-in. The magnitude of the effect estimates for all outcomes did not importantly differ between trials with a placebo run-in and trials with no run-in. Adjustment for the confounding variables did not materially change the estimates. Conclusions The hypothesised benefits of placebo run-in periods were not observed in our sample of anticholinergic randomised trials for the treatment of overactive bladders. Designing future trials of anticholinergics with more pragmatic intentions is likely to result in evidence that more directly informs clinical decision making.