Nutritional management in newborn babies receiving therapeutic hypothermia: two retrospective observational studies using propensity score matching.

BACKGROUND: Therapeutic hypothermia is standard of care for babies with moderate to severe hypoxic-ischaemic encephalopathy. There is limited evidence to inform provision of nutrition during hypothermia. OBJECTIVES: To assess the association during therapeutic hypothermia between (1) enteral feeding and outcomes, such as necrotising enterocolitis and (2) parenteral nutrition and outcomes, such as late-onset bloodstream infection. DESIGN: A retrospective cohort study using data held in the National Neonatal Research Database and applying propensity score methodology to form matched groups for analysis. SETTING: NHS neonatal units in England, Wales and Scotland. PARTICIPANTS: Babies born at ≥ 36 gestational weeks between 1 January 2010 and 31 December 2017 who received therapeutic hypothermia for 72 hours or who died during treatment. INTERVENTIONS: Enteral feeding analysis - babies who were enterally fed during therapeutic hypothermia (intervention) compared with babies who received no enteral feeds during therapeutic hypothermia (control). Parenteral nutrition analysis - babies who received parenteral nutrition during therapeutic hypothermia (intervention) compared with babies who received no parenteral nutrition during therapeutic hypothermia (control). OUTCOME MEASURES: Primary outcomes were severe and pragmatically defined necrotising enterocolitis (enteral feeding analysis) and late-onset bloodstream infection (parenteral nutrition analysis). Secondary outcomes were survival at neonatal discharge, length of neonatal stay, breastfeeding at discharge, onset of breastfeeding, time to first maternal breast milk, hypoglycaemia, number of days with a central line in situ, duration of parenteral nutrition, time to full enteral feeds and growth. RESULTS: A total of 6030 babies received therapeutic hypothermia. Thirty-one per cent of babies received enteral feeds and 25% received parenteral nutrition. Seven babies (0.1%) were diagnosed with severe necrotising enterocolitis, and further comparative analyses were not conducted on this outcome. A total of 3236 babies were included in the matched enteral feeding analysis. Pragmatically defined necrotising enterocolitis was rare in both groups (0.5% vs. 1.1%) and was lower in babies who were fed during hypothermia (rate difference -0.5%, 95% confidence interval -1.0% to -0.1%; p = 0.03). Higher survival to discharge (96.0% vs. 90.8%, rate difference 5.2%, 95% confidence interval 3.9% to 6.6%; p < 0.001) and higher breastfeeding at discharge (54.6% vs. 46.7%, rate difference 8.0%, 95% confidence interval 5.1% to 10.8%; p < 0.001) rates were observed in enterally fed babies who also had a shorter neonatal stay (mean difference -2.2 days, 95% confidence interval -3.0 to -1.2 days). A total of 2480 babies were included in the matched parenteral nutrition analysis. Higher levels of late-onset bloodstream infection were seen in babies who received parenteral nutrition (0.3% vs. 0.9%, rate difference 0.6%, 95% confidence interval 0.1% to 1.2%; p = 0.03). Survival was lower in babies who did not receive parenteral nutrition (90.0% vs. 93.1%, rate difference 3.1%, 95% confidence interval 1.5% to 4.7%; p < 0.001). LIMITATIONS: Propensity score methodology can address imbalances in observed confounders only. Residual confounding by unmeasured or poorly recorded variables cannot be ruled out. We did not analyse by type or volume of enteral or parenteral nutrition. CONCLUSIONS: Necrotising enterocolitis is rare in babies receiving therapeutic hypothermia, and the introduction of enteral feeding is associated with a lower risk of pragmatically defined necrotising enterocolitis and other beneficial outcomes, including rates of higher survival and breastfeeding at discharge. Receipt of parenteral nutrition during therapeutic hypothermia is associated with a higher rate of late-onset infection but lower mortality. These results support introduction of enteral feeding during therapeutic hypothermia. FUTURE WORK: Randomised trials to assess parenteral nutrition during therapeutic hypothermia. TRIAL REGISTRATION: Current Controlled Trials ISRCTN474042962. FUNDING: This project was funded by the National Institute for Health Research (NIHR) Health Technology Assessment programme and will be published in full in Health Technology Assessment; Vol. 25, No. 36. See the NIHR Journals Library website for further project information.

Every year, approximately 1200 babies in the UK suffer a lack of oxygen to the brain around birth. This is called hypoxic­ischaemic encephalopathy and can lead to brain injury or death. To treat hypoxic­ischaemic encephalopathy, babies receive cooling treatment in which their body temperature is lowered. Doctors do not know the best way to give nutrition to babies receiving cooling treatment. Babies can either be fed milk into their stomach (enteral nutrition) or be given nutrients through their veins (parenteral nutrition). We compared babies who were fed milk while they were being cooled with babies from whom milk was withheld while they were being cooled to see if there was a difference in the frequency of necrotising enterocolitis, a severe gut disease. In addition, we compared babies who received parenteral nutrition while they were being cooled with babies who did not to see if there was a difference in infections. Finally, we looked at other outcomes, including survival and breastfeeding. We used the National Neonatal Research Database, which holds de-identified (i.e. no baby can be identified) information on all babies who have received NHS neonatal care. We used a statistical approach to match babies in each group (i.e. fed babies and not fed babies) as closely as possible so that any difference in outcomes was because of different nutrition and not because of other differences. We included > 6000 babies with hypoxic­ischaemic encephalopathy. Approximately one in three babies received milk feeds and one in four babies received parenteral nutrition during cooling. Necrotising enterocolitis was very rare. More babies who were fed milk during cooling had good outcomes (e.g. being breastfed at discharge) and fewer had necrotising enterocolitis. Most of these babies received only a small amount of milk in the first 3 days. More babies given parenteral nutrition had infections, but also more survived. This suggests that it is probably safe and may be beneficial to feed babies milk during cooling. More research should look at milk feeding and parenteral nutrition during cooling.

Study protocol: optimising newborn nutrition during and after neonatal therapeutic hypothermia in the United Kingdom: observational study of routinely collected data using propensity matching.

INTRODUCTION: Therapeutic hypothermia is standard of care for infants born ≥36 weeks gestation with hypoxic ischaemic encephalopathy (HIE); consensus on optimum nutrition during therapeutic hypothermia is lacking. This results in variation in enteral feeding and parenteral nutrition (PN) for these infants. In this study, we aim to determine the optimum enteral nutrition and PN strategy for newborns with HIE during therapeutic hypothermia. METHODS AND ANALYSIS: We will undertake a retrospective cohort study using routinely recorded electronic patient data held on the United Kingdom (UK) National Neonatal Research Database (NNRD). We will extract data from infants born ≥36 weeks gestational age between 1 January 2008 and 31 December 2016, who received therapeutic hypothermia for at least 72 hours or died during therapeutic hypothermia, in neonatal units in England, Wales and Scotland. We will form matched groups in order to perform two comparisons examining: (1) the risk of NEC between infants enterally fed and infants not enterally fed, during therapeutic hypothermia; (2) the risk of late-onset blood stream infections between infants who received intravenous dextrose without any PN and infants who received PN, during therapeutic hypothermia. The following secondary outcomes will also be examined: survival, length of stay, breast feeding at discharge, hypoglycaemia, time to full enteral feeds and growth. Comparison groups will be matched on demographic, maternal, infant and organisational factors using propensity score matching. ETHICS AND DISSEMINATION: In this study, we will use deidentifed data held in the NNRD, an established national population database; parents can opt out of their baby's data being held in the NNRD. This study holds study-specific Research Ethics Committee approval (East Midlands Leicester Central, 17/EM/0307). These results will help inform optimum nutritional management in infants with HIE receiving therapeutic hypothermia; results will be disseminated through conferences, scientific publications and parent-centred information produced in partnership with parents. TRIAL REGISTRATION NUMBER: NCT03278847; pre-results, ISRCTN47404296; pre-results.