Article Text

Download PDFPDF

Ethical and practical issues relating to the global use of therapeutic hypothermia for perinatal asphyxial encephalopathy
  1. D J Wilkinson1,2,
  2. S Thayyil3,
  3. N J Robertson3
  1. 1The Ethox Centre, Department of Public Health and Primary Health Care, University of Oxford, Oxford, UK
  2. 2Oxford Uehiro Centre for Practical Ethics, University of Oxford, Oxford, UK
  3. 3Neonatology, Institute for Women's Health, University College London, London, UK
  1. Correspondence to Dr Dominic J Wilkinson, The Ethox Centre, Department of Public Health and Primary Health Care, The University of Oxford, Badenoch Building, Headington, Oxford OX3 7LF, UK; dominic.wilkinson{at}


In intensive care settings in the developed world, therapeutic hypothermia is established as a therapy for term infants with moderate to severe neonatal encephalopathy due to perinatal asphyxia. Several preclinical, pilot and clinical trials conducted in such settings over the last decade have demonstrated that this therapy is safe and effective. The greatest burden of birth asphyxia falls, however, in low- and middle-income countries; it is still unclear whether therapeutic hypothermia is safe and effective in this context. In this paper, the issues around treatments that may be proven safe and effective in the developed world and the caution needed in translating these into different settings and populations are explored. It is argued that there are strong scientific and ethical reasons supporting the conduct of rigorous, randomised controlled trials of therapeutic hypothermia in middle-income settings. There also needs to be substantial and sustainable improvements in all facets of antenatal care and in the basic level of newborn resuscitation in low income countries. This will reduce the burden of disease and allow health workers to determine rapidly which infants are most eligible for potential neuroprotection.

Statistics from


In the UK and other developed world settings, moderate or severe encephalopathy following perinatal asphyxia affects 1–2/1000 term births.1 There are major consequences for families and for society: approximately 15% of cases will die in the neonatal unit, 10%–15% will develop cerebral palsy and up to 40% will have other disabilities including blindness, deafness, autism, global developmental delay or problems with cognition, memory, fine motor skills and behaviour.1 2 Following the publication of several clinical trials in the last 2–3 years,3,,6 therapeutic hypothermia (cooling by 3°C–4°C for 72 h started within 6 h of birth) is established as the first effective therapy for neonatal encephalopathy. Meta-analyses of these clinical trials provide conclusive evidence for the safety and efficacy of therapeutic hypothermia in improving neurological outcome following neonatal encephalopathy.7 8 The typical number needed to treat for survival without neurological abnormality is 8 (95% CI 6 to 18).7

In low-resource settings such as sub-Saharan Africa, neonatal encephalopathy related to perinatal asphyxia is about 10–20 times more common than in the developed world.9 Annually, almost one-quarter of the 4 million global neonatal deaths are caused by perinatal asphyxia.10 In parallel with preventive strategies, a safe and effective therapy for neonatal encephalopathy that is appropriate for developing country settings could benefit millions of infants and families. It is tempting to believe that hypothermia may be such a therapy. However, it is not clear that results from previous trials of hypothermia can be extrapolated to settings outside the developed world. In this paper, we explore the ethical and practical considerations involved in the global use of hypothermia for perinatal asphyxia.

What is already known on this topic

  • Therapeutic hypothermia reduces the risk of death or severe disability in newborn infants with birth asphyxia in developed countries.

  • The greatest burden of birth asphyxia lies in low- and middle-income countries.

What this study adds

  • There are important population differences and practical considerations such that previous trial results cannot be safely extrapolated to developing countries.

  • There are strong ethical arguments to support rigorous, randomised controlled trials of therapeutic hypothermia in low- and middle-income countries.

Can results from trials in the developed world be extrapolated to low- and middle-income settings?

There are population differences and practical considerations, which mean that results from previous cooling trials cannot be safely extrapolated to low- and middle-income countries.

Population differences

First, in developing countries, brain injury from perinatal asphyxia may be more established because of maternal malnutrition, intra-uterine growth retardation, more frequent obstructed labour and poor obstetric and neonatal care. More than 2/3 of deliveries happen outside hospital settings.11 The effectiveness of cooling may be reduced if there is a high incidence of pre-existing brain injury or if the therapeutic window has passed. By the time that infants are treated in the neonatal unit, the secondary wave of neuronal death associated with brain secondary energy failure, measured using magnetic resonance spectroscopy may have already occurred.12

Second, in many areas, maternal HIV infection and puerperal sepsis are common. Perinatal asphyxia may coexist with neonatal sepsis,13 and the two conditions may be difficult to differentiate at birth. Exclusion of infected infants is not likely to be practicable, and core temperature reduction has been associated with neutrophil compromise, which has the potential to worsen sepsis and pneumonia.14 This association may at least partly explain the association between hypothermia and neonatal death in low-resource settings.13 15 16

Third, previous trials included mostly babies with severe illness requiring multi-organ support.4 5 Such babies are unlikely to survive in a low-resource setting, raising the possibility that hypothermia will be less effective in a different case mix of patients. On the other hand, even babies with a milder encephalopathy may be at a high risk of adverse neurodevelopmental outcome in these settings.17 Furthermore, it is possible that mild disabilities may influence quality of life and survival more directly than in the developed world and lead to significant societal and economic burden. Consequently, the risk/benefits of cooling in such countries are likely to be different.

Fourth, natural or accidental hypothermia is observed in many asphyxiated infants in developing country settings because of a number of factors including home birth, basic neonatal care and lack of overhead radiant warmers.16 Accidental hypothermia (with no control of the core temperature, a lack of physiological and metabolic intensive care support and no sedation) must be differentiated from therapeutic hypothermia (where cooling is induced in a controlled manner to a core temperature >33.5°C for 72 h under intensive care conditions and sedation). There is, however, a concern that accidental hypothermia may dilute the benefit of therapeutic hypothermia in the standard care group in clinical trials. In a recent pilot study in Uganda, asphyxiated infants undergoing standard care were hypothermic for up to 15 h after birth.18 This is consistent with a pattern first described in the 1950s and thought to be secondary to a reduced metabolic rate, a possible endogenous neuroprotective response.19 Many infants with perinatal asphyxia in neonatal units from low-resource settings are hypothermic on admission.15 20,,22 Rewarming of such infants is done rapidly (<2 h) in settings where radiant warmers are available23 24 and much more slowly (approximately 14 h) when such facilities are not available.18 Again, it is unclear whether these approaches are beneficial or harmful.

Practical and economic issues

One major practical barrier to provision of therapeutic hypothermia is that most infants with perinatal asphyxia in developing countries do not have access to intensive care. Although therapeutic hypothermia is thought to be clinically safe,7 8 the intervention is associated with increased inotropic support, bradycardia and thrombocytopenia.7 The effect of therapeutic hypothermia on meconium aspiration syndrome and persistent pulmonary hypotension, conditions occurring more commonly in low-resource settings, is unclear. The use of therapeutic hypothermia without adequate facilities to provide ventilatory or cardiovascular support could result in more harm.

Furthermore, the equipment used for providing hypothermia in previous trials4 5 in developed countries is prohibitively expensive. It may be vulnerable to the frequent power outages in many developing countries. Although some developing world hospitals may have access to such equipment (eg, some private hospitals in India), there is a need for low-cost, sustainable and effective means of providing cooling to be developed.

Finally, a threshold standard of neonatal care is required before therapeutic hypothermia can be used safely. It may be inappropriate to use hypothermia in neonatal units where there is inadequate maternal-fetal medicine support, resuscitation, basic monitoring facilities or sufficient skilled staff. In low- and middle-income countries, many deliveries occur at home or in small hospitals where there are limited facilities for neonatal care.

Is it ethical to undertake controlled trials of cooling in low-resource settings?

If results from trials of hypothermia cannot be applied in developing countries, there may be a strong case for undertaking new trials in these settings, particularly to establish the safety of this treatment. However, controlled trials of hypothermia could give rise to potential ethical concern. Perinatal research in developing countries has previously generated significant controversy. One frequently cited example is of placebo-controlled trials of short courses of zidovudine to prevent perinatal HIV transmission.25,,27 It was argued that instead of using placebo, researchers should have compared short with longer courses of zidovudine.28 A more recent neonatal example was a proposed placebo-controlled trial of a synthetic surfactant (lucinactant) in Latin America in 2001.25 29 After review by the Food and Drug Administration, the trial was changed to include only active controls (beractant and colfosceril).25 30

There has been a recent debate about the ethics of randomising infants to normothermia,31,,34 and in developed countries, hypothermia has moved to become a standard of care in many neonatal units.35a 35b The Declaration of Helsinki, successor to the Nuremberg Code, has strongly supported the idea of a universal standard of care for research in its revised version of 2008. “The benefits, risks, burdens and effectiveness of a new intervention must be tested against those of the best current proven intervention.”36 This could be interpreted to prohibit controlled cooling trials in developing countries.

However, as noted in the previous section, the best current proven intervention for asphyxia in developed countries may not be effective in developing countries. A number of scientists and ethicists have argued that the Declaration of Helsinki is too rigid and that it would prevent research that has the potential to improve the health and well-being of patients in developing countries.27 37 The US Food and Drug Administration currently requires clinical trials undertaken outside the United States to comply with ‘good clinical practice’ standards, which are less restrictive of the use of placebo than the Declaration of Helsinki.38a 38b

Research in the developing world may involve different treatments from those that would be applied in developed countries, but healthcare in those countries also involves different standards of treatment. The mere presence of different standards does not mean that there are “double standards” in the pejorative sense; the question is whether there are morally relevant differences justifying a different treatment.39 Some of the reasons why cooling does not currently represent a standard of care in developing countries have already been outlined.

Instead of a universal standard of care, some have proposed that research participants should be offered as a minimum the best intervention currently available as part of a national public health system.39 No patient would be denied treatment that would be otherwise available to them. This would permit controlled cooling trials in low- and middle-income countries.

The other ethical concept that would support such trials is that of clinical equipoise.40 41 While some clinicians in developing countries are keen to use therapeutic hypothermia given its effectiveness in previous trials, the factors noted previously have led others to fear that hypothermia would be ineffective or even harmful. The most appropriate way to resolve this disagreement would be by performing definitive trials.

Finally, given the contribution of poor antenatal care to perinatal asphyxia, it might be argued that attention should be given to addressing the cause of asphyxia in developing countries rather than studying possible treatments. However, given that perinatal asphyxia continues to affect a large number of infants in low-resource settings, it would seem sensible to attempt to influence the cause and the consequences of asphyxia. If low-cost methods of cooling are effective in developing countries, they have the potential to dramatically improve global child mortality and morbidity.

Where to from here?

The revised Declaration of Helsinki reflects a widespread concern about the appropriateness of placebo or no-treatment trials when proven treatments already exist. It is important that we think twice about performing trials in developing countries that would not be acceptable in developed countries. But there are situations where this is appropriate. Therapeutic hypothermia for birth asphyxia provides such an example. As outlined above, there are important scientific and ethical reasons to be cautious about adopting therapeutic hypothermia and to support controlled trials of cooling in developing countries. Such trials should aim in particular to assess the safety of therapeutic hypothermia in this setting, include strict stopping rules and plan subgroup analysis in infants with growth restriction, sepsis and meconium aspiration.

On the other hand, it is important to recognise that settings differ. In transitional economies like India and China, neonatal units and populations vary from fully resourced in some healthcare settings to low resourced in other settings. Where patient characteristics, intrapartum management and intensive care facilities are similar to those in previous trials, as is the case in some neonatal units in private healthcare settings in India, for example, cooling may be offered in accordance with established protocols.

Several small pilot studies have been undertaken to examine the feasibility of cooling in low-resource settings.18 42 43 Horn et alused ice caps to induce therapeutic hypothermia in 20 newborns with neonatal encephalopathy in South Africa.42 Their trial was stopped midway because of wide fluctuation in temperatures. They have subsequently investigated the use of a servocontrolled fan for cooling; this resulted, however, in shivering in the majority of cooled infants.44 A pilot study from China on 62 babies using a locally made head cooling device was able to induce and maintain hypothermia in babies admitted with asphyxia.43 The mortality was 6% in the control and cooled arms. A recently concluded feasibility study of therapeutic hypothermia for perinatal asphyxia using water bottles in a neonatal unit in Uganda showed 33% mortality in the cooling arm versus 7% in control arm.18 While the study was not powered to look at this outcome, and the difference may have been related to unequal proportions of severely affected infants, it highlights the potential for harm and the importance of performing controlled trials. Although these studies suggest that less technologically sophisticated methods of cooling may be feasible, their safety has yet to be demonstrated. Furthermore, high ambient temperatures with considerable seasonal variation are likely to affect the ability of cooling equipment to achieve and maintain target temperatures and may render simpler techniques ineffective. There is no data on the safety of methods for inducing and maintaining hypothermia during transport.

One of the most significant hurdles is that birth asphyxia has not been a funding priority despite that the number of disability-adjusted life-years due to intrapartum asphyxia exceeds those due to all childhood conditions preventable by immunizations.9 Efforts should continue to be made to prevent birth asphyxia by improving antenatal and perinatal care in developing countries. However, there is also a need for carefully designed, rigorous, randomised controlled trials. If cooling is effective in developing countries, and if it can be administered effectively and economically in low-resource settings, it has the potential to save and improve the lives of hundreds of thousands of newborn infants. Policy makers, experts in therapeutic hypothermia and paediatricians in developing countries should join forces to help evaluate whether cooling is safe and effective in low-resource settings. A phased approach may be most effective with an initial focus of clinical trials in mid-resource settings while improving basic antenatal, perinatal and neonatal care in low resource settings.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 27.
  28. 28.
  29. 29.
  30. 30.
  31. 31.
  32. 32.
  33. 33.
  34. 34.
  35. 35a.
  36. 35b.
  37. 36.
  38. 37.
  39. 38a.
  40. 38b.
  41. 39.
  42. 40.
  43. 41.
  44. 42.
  45. 43.
  46. 44.


  • Funding DW is supported by an Oxford Nuffield Medical fellowship, an Eric Burnard fellowship and a Royal Australasian College of Physicians Astra-Zeneca medical fellowship. ST is funded by the Comprehensive Biomedical Research Centre (CBRC), University College London. This work was undertaken in part at UCL/UCH who received a proportion of funding form the UK Department of Health's NIHR Biomedical Research Centre's funding scheme. The funders had no involvement in this work.

  • Competing interests None.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.