Effect of Early Hydrocortisone on Risk of Gastrointestinal Perforations in Extremely Preterm Infants

Effect of Early Hydrocortisone on Risk of Gastrointestinal Perforations in Extremely Preterm Infants: A Protocol for a Retrospective Cohort Study Using Routinely Collected Data

A large, randomised control trial, the PREMILOC trial, has established that giving low dose hydrocortisone prophylactically in the first ten days of life reduces the risk of bronchopulmonary dysplasia in babies born before 32 weeks' gestation. However, the PREMILOC trial was underpowered to investigate rarer side effects, such as gastrointestinal perforation. This study aims to establish whether the odds of gastrointestinal perforation increase when extremely preterm infants are given prophylactic hydrocortisone in the first ten days of life.

This retrospective cohort study will use routinely collected data from the U.K. National Neonatal Research Database. The investigators will examine the records of all infants born before 28 weeks' gestation and cared for in English and Welsh neonatal units between 2016 and 2023. Infants will be considered exposed if they received hydrocortisone for at least eight consecutive days, beginning on postnatal day 1 or 2. The primary outcome will be gastrointestinal perforation, as recorded in the infant's neonatal unit record. This outcome will be validated with the original care teams for a sample of babies. Data will be analysed using a propensity score matched approach to reduce the impact of confounding.

Inclusion Criteria:

Infants who:

• were admitted to a neonatal unit between the 1st January 2016 and 31st March 2023 and
• received any of their care in a NHS neonatal unit in England and Wales (part of UK Neonatal Collaborative and therefore contributing data to the NNRD) and
• were born before 28 weeks' gestation.

Exclusion Criteria:

• They have missing data for principal background variables (gestational age at birth, birth weight, year of birth and date of death for those that died).
• Their recorded birthweight absolute value z score exceeds 4 or is missing.
• They died on postnatal day 1 or 2 .