Fig. 3From: Evaluating the efficacy of dexamethasone in the treatment of patients with persistent acute respiratory distress syndrome: study protocol for a randomized controlled trialEstimation of samples size calculations based on expected ventilator-free days and mortality. Power analysis has been performed according to Schoenfeld et al [32]. Since we need three probabilities (see Table 1, text highlighted in yellow), we have selected three probabilities compatible with our hypothesis (using Mann-Whitney test). We hypothesized that patients in the experimental treatment group have a greater probability of being weaned earlier from the ventilator (0.0641 vs. 0.481) and a lower probability of dying (0.0345 vs. 0.05) for each day of the follow-up periodBack to article page