Treatment guided by cerebral oximetry in mechanically ventilated newborns: a statistical analysis plan for step one of the SafeBoosC-IIIv randomised clinical trial

Background: Newborns requiring invasive mechanical ventilation are at high risk of neurodevelopmental impairment, prolonged hospitalisation, and increased mortality. Treatment guided by cerebral oximetry monitoring has been proposed to reduce morbidity and mortality. Methods: The SafeBoosC-IIIv trial is a multicentre, parallel-group, randomised clinical trial. The trial will be conducted in two steps. This is a statistical analysis plan for step one. The objective of step one is to assess whether treatment guided by cerebral oximetry monitoring, compared with usual care, increases the number of hospital-free days in newborns receiving invasive mechanical ventilation. Inclusion criteria are gestational age ≥ 28 + 0 weeks, postnatal age less than 28 days, expected to receive invasive mechanical ventilation (intubation) for at least 24 h, and a cerebral oximeter available so monitoring can be started within 6 h after initiation of invasive mechanical ventilation. Exclusion criteria are suspicion of or confirmed brain injury or congenital heart malformation likely to require surgery. A total of 1610 participants will be randomised 1:1 to treatment guided by cerebral oximetry monitoring or usual care. The primary outcome will be hospital-free days within 90 days of randomisation, which will be analysed with the van Elteren test stratified by 'centre'. This statistical analysis plan provides a detailed description of the planned analyses, including methods for handling missing data and assessing statistical assumptions. Analyses will follow the intention-to-treat principle and will be performed independently by two statisticians. Conclusion: This statistical analysis plan describes the planned statistical analyses in detail for step one of the SafeBoosC-IIIv trial. Trial registration: ClinicalTrials.gov NCT05907317. First submitted on 8 June 2023, https://clinicaltrials.gov/study/NCT05907317.