Table 4 Sample size determination and interim analyses
CONSORT | ProACT |
|---|---|
Sample size | 1664 |
 Determination |  H1o: Procalcitonin guideline implementation does not reduce antibiotic exposure by Day 30. (superiority)  H2o: Procalcitonin guideline implementation increases the proportion of subjects who experience a composite endpoint of adverse outcomes by Day 30, by ≥4.5%. (non-inferiority)   ▪ Sample size is driven by H2o   ▪ 4.5% non-inferiority margin   ▪ ≥ 80% power, 1-sided alpha of 0.05   ▪ Lost to follow up and composite endpoint rates at 2nd interim data safety monitoring board meeting at ~2/3 accrual (April 2017) |
 Interim analyses and stopping rules |  Two interim analyses and one final analysis, approximately evenly spaced  O-Brien and Fleming stopping boundaries |