Methods A placebo-controlled, double blind, randomised, two-period crossover study was designed with interim analyses for futility and to allow possible sample size adjustment during the study.