Table 5.

Randomization, blinding, and statistical methods

CONSORT	ProACT
Randomization
Sequence generation	Patient-level, permuted block design Stratified by center, race, age Randomized equally to each study arm
Allocation concealment	Central Web-based randomization, accessible 24 h/day
Implementation	Local center staff enroll patients via Web-based randomization system Web-based system then assigns patients to trial arm, based on computer generated allocation sequence
Blinding	Statistical analysis and post-discharge outcome assessment staff are blinded to study arm By arm outcome data restricted to unblinded statistician and data safety monitoring board
Statistical methods	Intention-to-treat, as per pre-established analysis plan (primary analysis) Per-protocol analysis, where procalcitonin guideline is followed