Background
Objectives
Methods/design
Design and setting
Study protocol development and conduct
Randomisation and data collection
Primary outcome
Secondary outcomes
Safety
Statistical methods specified in the study protocol
Sample size calculation
Originally proposed analyses
Interim analyses and safety reporting
-
Mortality was ≥ 10% higher in the HC group than in the placebo group
-
Combined NEC Bell stage III and SIP ≥ 10% higher in the HC group than in the placebo group
-
ROP stage III or higher ≥ 10% higher in the HC group than in the placebo group
-
PVL stage I or higher ≥ 5% higher in the HC group than in the placebo group
Statistical analysis plan
Overall principles
Handling of missing data
Definition of analysis sets
Analysis population | HC treatment group | Placebo treatment group |
---|---|---|
Intention to treat ‘as randomised’ | HC randomisation: • Including all protocol deviations (e.g., in eligibility criteria) | Placebo randomisation: • Including all protocol deviations (e.g., in eligibility criteria) |
As treated ‘actual treatment’ |
Patient received HC (at least one dose), regardless of allocated treatment at randomisation:
• Including those infants who received rescue open-label corticosteroids as described according to protocol • Including those infants who received rescue open-label corticosteroids not following protocol (protocol deviation) • Including those infants with other protocol deviations (e.g., in eligibility criteria) |
Patient received placebo (at least one dose), regardless of allocated treatment at randomisation:
• Including those infants not receiving any corticosteroid dose for pulmonary reasons • Excluding those infants who received rescue open-label corticosteroids following study protocol • Excluding those infants who received rescue open-label corticosteroids not following study protocol • Including those infants with other protocol deviations (e.g., in eligibility criteria) |
Per protocol | HC randomisation and treated according to study protocol: • Including those infants receiving rescue open-label corticosteroids as described by the study protocol • Excluding those infants receiving rescue open-label corticosteroids not according to study protocol (protocol deviation) • Excluding other protocol deviations (e.g., in eligibility criteria) | Placebo randomisation and treated according to study protocol: • Including those infants receiving rescue open-label corticosteroids following study protocol • Excluding those infants receiving rescue open-label corticosteroids not following study protocol (protocol deviation) • Excluding other protocol deviations (e.g., in eligibility criteria) |
Statistical analyses
Patient flow
Protocol deviations
Baseline characteristics
Short-term outcomes until initial hospital discharge (study phase I)
Variable/outcome | Type of outcome | Statistical analysis |
---|---|---|
Study phase I: short-term outcomes until initial hospital discharge
| ||
BPD-free survival at 36 weeks PMA | Primary | Logistic regression with correction for stratification factors plus sensitivity and sub-group analyses |
Death at 28 days PNA, at 36 weeks PMA and at hospital discharge BPD at 36 weeks PMA Failure to extubate 3, 7, 14 and 21 days after start of trial medication Time to extubation, censored at death Time to supplemental oxygen independence Time to hospital discharge Necrotising enterocolitis Gastrointestinal bleeding Spontaneous intestinal perforation Intraventricular haemorrhage and/or periventricular leucomalacia Retinopathy of prematurity Hypertension Hyperglycaemia requiring insulin therapy Nosocomial infection, including clinical or culture-proven sepsis, meningitis, pneumonia Patent ductus arteriosus needing medical intervention or surgical ligation Weight, head circumference, length at 36 weeks PMA Use of open-label HC treatment | Short-term secondary | Linear, logistic or Cox regression or competing risk model, as appropriate |
SUSARs, SAEs | Short-term secondary | Descriptive statistics |
Study phase II: long-term follow-up at 2 years CA
| ||
Survival without neurodevelopmental impairment | Key long-term secondary | Logistic regression with correction for stratification factors, sensitivity and sub-group analyses |
Mortality Number of hospital readmissions since first discharge to home Cerebral palsy and its severity Hearing loss requiring hearing aids Blindness Behaviour problems (Child Behavior Checklist)Growth (weight, length, head circumference) | Long-term secondary | Linear, logistic or Cox regression or competing risk model, as appropriate |