Methods/design
The trial is a multi-centre randomised controlled group trial (RCT) to assess whether the speed of increasing milk feed volumes (faster increase [30 ml/kg/day] versus slower increase [18 ml/kg/day]) in very preterm (<32 weeks) or very low birth weight infants (<1500 g) infants has any effect on survival without moderate or severe disability at 24 months corrected age.
Two thousand eight hundred infants from approximately 30 neonatal units will be recruited within the UK and Ireland over 3 years.
Inclusion criteria
-
Gestational age at birth is <32 completed weeks, or birth weight <1500 g
-
Receiving ≤30 ml/kg/day of milk at randomisation
-
Written informed parental consent is obtained
Exclusion criteria
-
Severe congenital anomaly
-
In the opinion of the treating clinician, have no realistic chance of survival
-
Unlikely to be traceable for follow-up at 24 months of age (for example, infants of non-UK residents)
Primary outcome
The primary outcome will be the proportion of infants surviving without moderate or severe neurodevelopmental disability at 24 months of age corrected for prematurity. This composite outcome will be determined by confirming that the child is alive or dead using records held and maintained by The Health and Social Care Information Centre and other central UK NHS bodies. For live infants, a parent report questionnaire will be used to assess sensory and gross motor impairment and standardised measures to assess cognitive function in order to identify children with:
-
Moderate/severe visual impairment (reduced vision uncorrected with aids; or blind in one eye with good vision in the contralateral eye; or blind/perceives light only)
-
Moderate/severe hearing impairment (hearing loss corrected with aids; or some hearing loss but not corrected by aids; or deaf)
-
Moderate/severe gross motor impairment (unable to walk or sit independently)
-
Moderate/severe cognitive impairment assessed using the Parent Report of Children’s Abilities–Revised (PARCA–R), a parent report measure of non-verbal cognitive and language development. Total PARCA–R scores <44 will be used to identify children with moderate/severe cognitive impairment [
16]. This questionnaire has been shown to have at least 80% sensitivity and 80% specificity for identifying children with scores < −2SD on a Gold Standard development test [
16,
17]
A child who has any one or more of these impairments will be classified with a moderate/severe disability. Definitions for motor and sensory impairments described above are as defined in the report published by British Association of Perinatal Medicine (BAPM) in 2008 [
18].
Secondary outcomes
Secondary outcomes to be assessed when infants are discharged home for the first time are:
-
Survival to discharge home.
-
Microbiologically-confirmed or clinically suspected late-onset invasive infection from trial entry to discharge home.
-
NEC (Bell stage 2 or 3) from trial entry to discharge home.
-
Time taken to reach full milk feeds (tolerating 150 ml/kg/day for 3 consecutive days).
-
Growth (change in z score–weight and head circumference for gestational age) from birth to discharge home.
-
Duration of parenteral feeding.
-
Length of time in intensive care.
-
Length of hospital stay to discharge home.
In addition, the separate components of the composite primary outcome at 24 months of age corrected for prematurity will be analysed individually as secondary outcomes. The diagnosis of cerebral palsy by a doctor or other health professional will also be a secondary outcome assessed at this age.
Trial procedures
Written consent will be sought from parents only after they have been given a full verbal explanation and written description of the trial via a parent information leaflet. Consent will be obtained by means of dated parental signature on a study consent form and the signature of the person who obtained informed consent. Recruitment will be conducted by a health professional with delegated authority. Parents who do not speak English will only be approached if an interpreter is available.
Remuneration
Parents will not be given any financial or material incentive or compensation for enrolling their babies in this trial.
Randomisation and allocation
Randomisation will take place at the time the clinician is ready to start increasing the feed volume. This will be performed through a secure website hosted by the NPEU CTU with telephone back-up available 24/7, 365 days a year.
A minimisation algorithm will be used to ensure balance on important prognostic factors: hospital, multiple birth, gestational age ranges, and birth weight <10th centile for gestational age. Multiple births will be given the same allocation.
Blinding
This is an open trial; blinding of the clinicians, nursing staff, and parents is not possible. A blinded endpoint review committee will examine the relevant data collection forms and clinical notes of infants with possible sepsis and NEC and classify them systematically according to predefined criteria.
Stopping or modifying the trial intervention
Deviations from the scheduled speed of increase may be made at the discretion of the treating clinician if the infant appears unable to tolerate the allocated speed of milk feed increase.
Withdrawal
At all stages it will be made clear to the parents that they are free to withdraw their infant from the trial at any time, without the need to provide explanation. If parents choose to withdraw, they will be asked for permission to complete data collection and/or follow-up.
The attending clinician may also withdraw the infant from the allocated treatment if they consider this to be in the best interest of the infant’s health and well-being.
Data collection before discharge
All outcome data are routinely recorded clinical items that can be obtained from the clinical notes or local microbiological laboratory records. No additional blood or tissue samples are required for this trial. Clinical information will be collected using specially created data collection forms.
Data collection after discharge
A parent questionnaire will assess neurodevelopmental outcomes and health care costs when the infant is 24 months of age (corrected for prematurity).
Data collection and processing
Data will be processed using validated data management systems to ensure consistency, viability and quality of data. It will be stored in line with the Data Protection Act 1998.
Safety reporting
Adverse events are defined as serious if they:
-
Result in death
-
Are life-threatening
-
Require inpatient hospitalisation or prolongation of existing hospitalisation
-
Result in persistent or significant disability/incapacity, or
-
Are a congenital anomaly/birth defect
The term “life-threatening” refers to an event in which the child was at risk of death at the time of the event; it does not refer to an event which hypothetically might have caused death if it were more severe. SAEs are to be reported from randomisation until discharged home.
Safety reporting will be carried out in accordance with the NPEU Clinical Trials Unit standard operating procedures and clinical regulations.
Sample size and power
The primary comparison will be the difference in the proportion of infants surviving without moderate or severe disability at 24 months of age corrected for prematurity. Based on previous trials, it is estimated that 80% of the infants will survive to 2 years, and that 11% of these will have a moderate or severe disability [
19]. Hence it is estimated the proportion surviving without moderate or severe disability in the control group receiving the 18 ml/kg/day increment will be 71%. With a total sample size of 2500 and allowing for a response rate of 80%, there will be 90% power to detect an absolute difference of 6.3% (from 71.0% in the control group to 77.3%) in this proportion, with a two-sided 5% significance level.
With the same level of significance, a sample size of 2500 infants will have 90% power to detect an absolute risk difference of 5.4% (from 25.0% in control group to 19.6%) in the incidence of sepsis [
20] and an absolute risk difference of 3.5% (from 6.0% in control group to 9.5%) in the incidence of NEC (Bell stage 2 or 3) [
21‐
23].
An inflation factor of 1.12 was applied to the sample size to allow for multiple births, expected to have correlated outcomes, receiving the same allocation. This was based on an estimate of 25% for the proportion of multiple births and an intraclass correlation coefficient of 0.9 for the primary outcome at 2 years, based on a similar outcome in a preterm population [
24]. The total number of babies that will be recruited will be 2800.
Statistical analysis
Demographic factors, clinical characteristics and outcomes will be summarised with counts (percentages) for categorical variables, mean (standard deviation) for Normally distributed continuous variables, or median (interquartile or entire range) for other continuous variables. Infants will be analysed according to allocation regardless of the speed of milk feed increase they actually receive.
The two groups will be compared using generalised estimating equations, adjusting for the minimisation factors to account for the correlation between treatment groups. This method of analysis will also account for the correlation in outcomes between twins and siblings born in a subsequent pregnancy during the trial period. For the primary outcome, an adjusted risk ratio with 95% confidence will be calculated using log binomial regression, or log poisson regression with a robust variance estimator if the binomial model fails to converge [
25]. Linear regression will be used for Normally distributed outcomes, quantile regression for skewed continuous variables, and Cox regression for time to event outcomes. Ninety-nine percent confidence intervals will be calculated for all secondary outcomes.
The consistency of the effect of advancing milk feeds across specific subgroups of infants will be assessed using the statistical test of interaction. Pre-specified subgroup analyses include: (i) week of gestation at birth, (ii) birth weight (<10th centile for gestational age versus ≥ 10th centile and (iii) type of milk (breast milk only/formula only/mixed). Subgroup analysis will be performed on the primary outcome, and the incidence of sepsis and NEC.
Economic data collection
Relevant resource use data collection will be undertaken prospectively from centres participating in the trial. Private out of pocket costs to parents will be collected via a questionnaire sent at 24 months of age corrected for prematurity. Unit costs will be obtained from published sources and centres participating in the trial and applied to resource use. Published sources will include Unit Costs of Health and Social Care [
26] and NHS Reference Costs.
The main economic analysis will be in the form of a cost-effectiveness analysis from the perspective of the health care provider (National Health Service), based on an intermediate outcome of cost per neonatal sepsis avoided when discharged home and on the outcome of disability-free survival at 24 months of age corrected for prematurity (cost per additional survivor without disability at 24 months of age corrected for prematurity). A secondary analysis will extend the perspective to include private out of pocket costs to families associated with travel and time off work during the period of follow up.
The analyses will adopt an incremental approach in that data collection will concentrate on resource use and outcome differences between trial arms. A bootstrapping approach will be undertaken in order to calculate confidence intervals around the mean costs [
27,
28]. Costs and benefits will be discounted as per NICE guidelines at 3.5%.
Results will be presented using cost-effectiveness acceptability curves. The robustness of the results will be explored using deterministic and probabilistic sensitivity analysis.
Discussion
Preterm infants are at significant risk of poor long-term neurodevelopmental problems with almost 12% having moderate or severe disability [
19], with both sepsis and NEC dramatically increasing this risk [
29‐
34]. Achieving full milk feeding sooner is associated with significant cost savings through decreased use of intravenous nutrition, a reduction in time spent in a specialist tertiary neonatal unit, shortened total hospital stay (potentially saving £1000 per day), and reductions in societal costs due to improved long-term outcomes [
29‐
34]. Vacating tertiary level neonatal cots sooner will also improve the family’s experience and the infant’s safety by decreasing the need for transfer to other hospitals for intensive care.
Infection and NEC remain highly predictive factors for neurodevelopmental disability. Any reduction in either problem may therefore be expected to reduce long-term disability in this population.
Overall lifetime financial costs of disability are significant, and so preventing even a few cases and reducing cognitive problems at the population level would reduce the financial burden of long-term care for the NHS and society. No additional resources will be needed to implement the optimal feeding strategy, which, if successful, could be adopted rapidly across the NHS at low cost.
Acknowledgements
Not applicable.