Speed of Increasing milk Feeds Trial (SIFT)

• Design: Multi-centre pragmatic randomised controlled parallel-group trial. • Setting: 30 tertiary level neonatal units and their associated continuing care & data collection sites within the UK and Ireland using established neonatal clinical and trial networks. Participating centres are within established neonatal networks including the UK Neonatal Infection Surveillance Network, the Northern, Yorkshire, Trent, Midlands, South Central South Coast, and Norfolk, Suffolk and Cambridgeshire Neonatal Networks, and managed clinical and research networks in Northern Ireland and Scotland. • Target Population: Infants born before 32 completed weeks' gestation or <1500g birthweight, and receiving <30ml/kg/day of milk and where the clinical team are uncertain over the best rate of increase in feeds. Consent will be obtained in the first 48 hours and randomisation will take place at the time the clinician is ready to start increasing the feed volume. Infants with severe congenital anomalies, those not likely to survive or be contactable at follow up will not be eligible. • Health Technologies Being Assessed: Based on survey data indicating acceptable limits of current clinical practice (1), we will compare two rates of increasing the daily volume of milk feeds: 18ml/kg/day increment (slow) and 30 ml/kg/day (fast). If milk is increased as intended, full milk feeds (tolerating 150ml/kg/day) will be reached 4 days sooner in the fast group. • Primary outcome - The primary outcome is the proportion of infants surviving without moderate or severe neurodevelopmental disability at 24 months of age corrected for prematurity. In order to report the proportion of infants surviving without disability a composite parent report questionnaire will be used to identify children with (i) moderate/severe vision impairment, (ii) moderate/severe hearing impairment, and (iii) moderate/severe motor impairment and moderate/severe cognitive impairment defined as PARCA-R PRC scores <44 respectively (2). A child who has any one or more of these will be classified with disability. Definitions for motor and sensory impairments will be as defined in the report published by BAPM in 2008. • Secondary outcomes: incidence of late onset sepsis (after 72 hours of age) before discharge. Sepsis is defined as microbiologically-confirmed or clinically strongly-suspected infection, as defined by the NEONIN dataset; incidence of necrotising enterocolitis; time to reach full milk feeds (defined as 150ml/kg/day for 72 hours); growth to discharge (measured by weight and head circumference); duration of parenteral feeding before discharge; length of time in intensive care; length of hospital stay and survival without serious disability at 24 months corrected age. • Cost analysis: The health service resources used and the costs to the family during the infant’s hospital stay will be identified, measured, valued and combined with clinical effectiveness data and used to generate an incremental cost per additional benefit. • Proposed Sample Size: The event rate for the primary outcome is expected to be 25%. To detect an absolute risk reduction (ARR) of 5.4% from 25% to 19.6% with 90% power, assuming a 2-sided 5% significance level will require ~2500 infants (a minimum of 1236 per arm) (3). This sample size would also have 90% power to detect an increase in the incidence of stage II or III NEC from 6.3% to 9.8% with a 2-sided 5% significance level. Based on previous trials, we anticipate attrition of no more than 20% at two years, reducing the sample evaluated at 2 years to ~2000 babies. We estimate the proportion of infants born before 32 weeks' gestation surviving to 2 years with no or mild disability in the control group to be 69%. The trial will have 90% power to detect an (absolute) increase in survival with no or mild disability of 6% from 69% to 75%. An inflation factor of 1.12 was applied to the sample size to allow for multiple births, expected to have correlated outco