Cooling for newborns with hypoxic ischaemic encephalopathy

Therapeutic hypothermia for term newborn infants with moderate-to-severe hypoxic ischaemic encephalopathy reduces the combined outcome of death or long-term neurodevelopmental disability at 18 months. Trials included in this review found that the benefits of cooling far outweighed the reported side-effects, and there was no increase in the risk of neurodevelopmental disability among the survivors who had received cooling therapy.

RHL Commentary by Ballot DE


In developed countries, perinatal asphyxia (lack of oxygen in the brain around the time of birth) affects 3–5 infants per 1000 live births, with 0.5–1 infants per 1000 live births developing brain damage in the form of hypoxic ischaemic encephalopathy (HIE) (1). Up to 60% of infants with HIE die and 25% of survivors are left with a significant handicap (2). HIE prevalence data for developing countries are not available. However, my own observation is that perinatal asphyxia poses an even greater problem in under-resourced settings – both as a major cause of neonatal deaths and of significant disability in survivors. In such settings, a significant proportion of mortality and morbidity due to asphyxia is caused likely by the lack of trained health-care staff and appropriate facilities.

Until recently there was no effective treatment for HIE; once an infant developed asphyxia, there was no way to stop brain damage or death from occurring. Moderate cooling of the brain (to between 32 and 34 degrees Celsius) has been proposed as a possible way of preventing brain damage and death in asphyxiated newborn infants. This Cochrane review (3) evaluated the effects on mortality and long-term neurodevelopmental disability, as well as clinically important side-effects, of moderate brain cooling in newborn infants with HIE.


This systematic review followed the Cochrane Collaboration methodology according to the guidelines of the Cochrane Neonatal Review Group. The review authors searched the the Oxford Database of Perinatal Trials, the Cochrane Central Register of Controlled Trials, MEDLINE, abstracts and conference proceedings for randomized and quasi-randomized trials that had compared the effects of cooling (either whole body or head only) versus no cooling in infants with HIE. Three authors independently identified studies to be included, assessed their quality and extracted the data. The quality of each trial was assessed according to blinding of randomization, blinding of the intervention, completeness of follow-up, and blinding of the outcome measurement.

Twenty trials were identified, out of which eight randomized controlled trials were included in this review. Nine trails were excluded as these did not meet the inclusion criteria, and three trials were still ongoing. The research methods employed in the eight included trials were judged by the review authors to be of high quality.

The study participants were full-term newborn babies with evidence of perinatal asphyxia and HIE, without any major congenital abnormalities. The main outcome measure was a combination of death or major neurodevelopmental handicap at 18 months of age; effects of cooling on death and neurodevelopmental outcome alone were also determined. Neurodevelopmental handicap at 18 months was defined as a combination of cerebral palsy, developmental delay or intellectual impairment, blindness and sensorineural deafness requiring hearing aids. Secondary outcome measures were the effects of cooling on each of the individual neurodevelopmental components, incidence of the side-effects of cooling and effects on neurological function within the first 3 months of life. Potential side-effects of cooling that were evaluated included effects on heart rate and rhythm, blood pressure, full blood count (including platelet count) and bleeding tendency, electrolyte levels, and kidney function. Early neurological function included staging of HIE, seizures, degree of electroencephalogram (EEG) abnormality, magnetic resonance imaging (MRI) changes and number of days until the infant was able to breastfeed properly.

Results from all the studies included in the review were combined and a meta-analysis was done to determine the relative risk (RR) and risk difference (RD); 95% confidence intervals (CI) were reported for all the results. The number needed to treat (NNT) was also calculated for significant results. Babies with missing outcome results were not included in the analysis.


A total of 630 infants close to term with moderate-to-severe HIE and no obvious congenital abnormalities were included in the analysis.

3.1 Death or major neurodevelopmental disability

First, the combined outcome of death/major disability was considered. Four studies (506 babies; 336 survivors) that reported neurodevelopmental follow-up were included in this analysis and showed that moderate cooling significantly reduced death/major disability in newborns with moderate-to-severe HIE. The typical RR was 0.76 (95% CI 0.65–0.89) and RD was −0.15 (95% CI −0.24 to −0.07), NNT 7 (95% CI 4–14). Allowance for missing data did not change the significance of these findings.

The benefit of cooling remained when death and major disability were considered separately as outcome variables. Meta analysis of all eight trials showed that occurrence of death was significantly reduced in the asphyxiated babies who had been cooled (RR 0.74; 95% CI 0.58–0.94; RD −0.09; 95% CI −0.16 to −0.02). In the four trials that had reported neurodevelopmental outcome, cooling also showed a significant reduction in major neurodevelopmental disability (RR 0.68; 95% CI 0.51–0.92; RD −0.13, 95% CI −0.23 to −0.03).

The reviewers then analysed outcomes according to the method of cooling (whole body versus head cooling alone). Whole-body cooling was associated with a significant reduction in the combination of death and major disability (RR 0.69, 95% CI 0.55–0.86; RD −0.21, 95% CI −0.33 to −0.09). It was also associated with significantly reduced number of deaths alone (RR 0.66, 95% CI 0.47–0.93; RD −0.13, 95% CI −0.23 to −0.02; NNT 8, 95% CI 4–50) and the number of survivors with major neurodevelopmental disability (RR 0.60, 95% CI 0.40–0.92; RD −0.17, 95% CI −0.31 to −0.03). Isolated cooling of the head did not show any benefit in terms of reduction of rates of death or major neurodevelopmental disability.

Although major neurodevelopmental disability as a whole was reduced by cooling, there was no benefit of cooling with regard to separate outcomes of cerebral palsy, neuromotor delay, developmental delay, blindness or sensorineural deafness requiring hearing aids.

3.2 Adverse effects of cooling

Cooling was safe and did not result in serious side-effects, which included a slightly lower baseline heart rate (RR 5.96, 95% CI 2.15–16.49; RD 0.07, 95% CI 0.04–0.11), a marginally significant increase in the need for blood pressure support (RR 1.17, 95% CI 1.00–1.38; RD 0.08, 95% CI 0.00–0.17), and more babies with a platelet count below 150 X 109 /litre (RR 1.55, 95% CI 1.14–2.11; RD 0.09, 95% CI 0.03–0.15). <sup>9</sup>

Cooling did not cause any abnormal heart rhythms and had no effect on the number of infants receiving blood transfusion, low white cell count, bleeding tendency, low blood sugar, low potassium level, reduced urine output, or incidence of sepsis.

3.3 Short-term neurological function

Cooling did not have any effect on seizures within the first 3 days of life. Other outcomes, including MRI findings, standardized neurological assessment and the time to start taking feeds by sucking, could not be analysed because these results were not reported. Subgroup analysis, for example by degree and duration of cooling, was not done as this information was not available.

3.4 Degree of encephalopathy

The reviewer authors then analysed the effects of cooling according to the initial severity of encephalopathy in asphyxiated babies. Cooling showed a significant reduction in the combined outcome of death/disability in babies with severe encephalopathy (RR 0.80, 95% CI 0.68–0.94; RD −0.18, 95% CI −0.31 to −0.05; NNT 6, 95% CI 3–20). When the outcomes were considered separately in this group of infants, cooling had no benefit on neurodevelopmental disability alone, but death rate was significantly lower (RR 0.72, 95% CI 0.56–0.94; RD −0.20, 95% CI −0.35 to −0.04; NNT 5, 95% CI 3–25). In babies with moderate encephalopathy cooling caused a slight reduction in the combined outcome of death and disability (RR 0.76, 95% CI 0.58–1.00; RD −0.12, 95% CI −0.23 to 0.00), but had no effect on either death or disability alone.



The reviewers conclude that therapeutic hypothermia for term newborn infants with moderate-to-severe hypoxic ischaemic encephalopathy reduces the combined outcome of death or long-term neurodevelopmental disability at 18 months. These results are both significant and clinically important, with an absolute risk reduction of 15% (NNT = 7). The benefits of cooling were still evident when the outcomes were considered separately; cooling reduced the chance of death and, if the baby survived, it decreased the risk of developing a neurodevelopmental disability. Furthermore, cooling was safe with only minor side-effects; the benefits of cooling far outweighed the side-effects reported. Also, there was no increase in the chance of neurodevelopmental disability among the surviving infants who had received cooling therapy. The quality of the trials was good and, although the caregivers were not blinded, there was no suggestion of bias in the data. The reviewers do include a cautionary note that results of 829 infants known to be randomized into ongoing cooling trials are not included in this review and their findings may produce different results in the future.

The reported trials were conducted in developed countries, often in centres of excellence with much experience in cooling babies and usually according to strict protocols. However, the findings would be applicable in under-resourced settings as there is no biological reason to believe that infants in such settings would react differently to cooling. Moreover, cost-effective means of cooling, such as fans and ice-packs, are likely to be available in institutional settings in developing countries


Cooling programmes should initially be implemented in academic centers using strict guidelines and adequate supervision. Cooling should be undertaken in a neonatal intensive care unit with proper monitoring and should involve staff that are experienced and trained in neonatal care. Cultural, religious and traditional beliefs are unlikely to affect cooling any cooling therapy programmes. The danger of excessive hypothermia, especially in low-birth-weight infants, must always be remembered.

The methods for cooling babies in this review ranged from using ice-packs and turning warmers off to servo-controlled mattresses and cooling caps. A cooling programme in an under-resourced setting should use a simple, low-cost method. Horn et al. (4) describe a cerebral cooling method with a servo controlled fan blowing room air. Suitable candidates for applying cooling therapy can be identified clinically; an amplitude-integrated EEG is not required (5).


Further topics for research include: development of low-cost methods of cooling appropriate for under-resourced settings; direct comparison of effects of whole body versus selective head cooling on outcomes; the effects of early initiation of cooling (directly after birth) as opposed to cooling at 6 hours; different methods and duration of re-warming of the baby; methods appropriate for under-resourced settings for identifying suitable babies for applying the cooling intervention; and the role of other therapies combined with cerebral cooling, e.g. phenobarbitone.

Acknowledgements: I would like to thank Prof. Haroon Saloojee for recommending me to The WHO Reproductive Health Library as a potential reviewer.


  • Levene MI, Sands C, Grindulis H, Moore JR. Comparison of two methods of predicting outcome in perinatal asphyxia. The Lancet 1986;8472:67-69.
  • Vannucci RC. Current and potentially new management strategies for perinatal hypoxic ischaemic encephalopathy. Paediatrics 1999;85:961-968.
  • Jacobs SE, Hunt R, Tarnow-Mordi WO, Inder TE, Davis PG. Cooling for newborns with hypoxic ischaemic encephalopathy. Cochrane database of systemic reviews 2007;Issue 4. Art No CD003311; D01: 10.1002/14651858 CD003311 pub 2.
  • Horn A, Thompson C, Woods D, Nel A, Bekker A, Rhoda N, Pieper C. Induced hypothermia for infants with hypoxic ischaemic encephalopathy using a servo controlled fan: an exploratory pilot study. Pediatrics 2009;123: e1090- e1098.
  • Sarkar S, Barks JD, Donn SM. Should amplitude integrated electroencephalography be used to identify infants suitable for hypothermic neuroprotection? Journal of Perinatology 2008; 28: 117-122.

This document should be cited as: Ballot DE. Cooling for newborns with hypoxic ischaemic encephalopathy: RHL commentary (last revised: 1 October 2010). The WHO Reproductive Health Library; Geneva: World Health Organization.


Related documents

About the author