A total Gefitinib nmr of 10 participants would provide an 80% probability of detecting

a difference of 10 cmH2O in maximal inspiratory pressure at a two-sided 5% significance level. We anticipated that a substantial proportion of these critically ill participants would die or receive a tracheostomy. We therefore increased the recruited sample to 20 participants per group to allow for this. All participants with follow-up data were analysed according to their group allocation, ie, using the intentionto-treat principle. Statistical significance was considered as p < 0.05, therefore mean between-group differences and 95% confidence intervals are presented for maximal inspiratory pressure, the index of Tobin, and weaning time. The Kappa test was used to evaluate the agreement between the evaluators of maximal inspiratory pressure. Total intubation time was analysed using a Kaplan-Meier curve. In the event of death, tracheostomy, or self-extubation, participants were excluded from the independent t-tests of between-group differences see more and were treated as censored cases in the survival analysis. During the recruitment period, 198 patients were screened, of whom 67 were eligible and monitored daily to assess readiness

to start weaning. Of the 67, 20 were tracheostomised, 5 died, and 1 was transferred to another centre before the start of weaning. The remaining 41 were randomised: 21 to the experimental group and 20 to the control group. The baseline characteristics, ie, on the day weaning started, of the two groups are presented in Table 1 and in the first two columns of Table 2. Four participants in each group died before extubation. Three participants

in the experimental group and two in the control group were tracheostomised before extubation. The intensive care unit those had a total of 24 beds, with 8 of these dedicated to postoperative patients. The physiotherapy team comprised 11 physiotherapists working in three shifts, all with expertise in intensive care, of which two have doctoral and six have masters qualifications. Consistency between the physiotherapists for the assessment of maximal inspiratory pressure was good, with a Kappa value of 0.68. Participants in the experimental group underwent training on all days during their weaning period. The average training load of the participants in the experimental group increased from 3 cmH2O initially to 20 cmH2O at the end of the weaning period. Group data for all outcomes at the start of weaning and at extubation for the experimental and control groups are presented in Table 2 while individual data are presented in Table 3 (see eAddenda for Table 3). Maximal inspiratory pressure increased significantly more in the treatment group than the control group (MD 7.6 cmH2O, 95% CI 5.8 to 9.4). The index of Tobin increased (ie, worsened) in both groups over the weaning period, but the increase was attenuated significantly by the inspiratory muscle training (MD 8.3 br/min/L, 95% CI 2.9 to 13.7).