Small Group Discussions

Break-Out Group Meetings Were Held on Four Topics:

Breakout Group 1 was chaired by Teresa Murguia de Sierra, Hospital Infantil de México Federico Gómez, Mexico City, Mexico and David Woods, Universityof Cape Town, Cape Town, South Africa

Group 1 was comprised of approximately 20 participants who were asked to ascertain the current state of knowledge about birth asphyxia in developing regions.

1. Definitions to be used.

After a long and passionate discussion it was agreed that the definition of “birth asphyxia” was very problematic and unsatisfactory and that no agreement had been achieved over many decades of debate about this term. There was no consensus on a single definition as this depended on the question being asked. Suggestions were made for a programmatic or outcomes-based definitions (e.g. no cry at birth) as well as process-based definitions (e.g. raised base deficit at birth).

All agreed that the most important aspect of “birth asphyxia” was tissue hypoxia. Hypoxia usually occurs during labour and delivery (intrapartum hypoxia) but could also occur during inadequate resuscitation after delivery or immediately before the onset of labour with placental abruption (peripartum hypoxia).

The term “acute peripartum event” was offered but not accepted by all as a replacement for “birth asphyxia”. A plea was made to no longer use the word “asphyxia” but rather define what is being assessed. Finally it was agreed that a description was preferable to a definition.

The clinical results of “birth asphyxia” (the burden) may be the best way of measuring the occurrence of this important but ill-defined condition. These outcomes consist of:

a. Acute intrapartum stillbirths

A fresh stillbirth is usually used as a proxy. Infants with gross birth defects incompatible with life (e.g. anencephaly) should be excluded. Some weight cut-off should also be used but no agreement could be reached. The predictive value of obstetric observations (e.g. foetal distress or placental abruption) for fresh stillbirth should be tested. Clear clinical guidelines are needed to recognise a “fresh” stillbirth. Good care during labour is critical to reduce intrapartum stillbirths.

b. Infants who are alive at birth but in poor clinical condition

The definition of these infants would depend on the clinical situation (e.g. poor cry, tone or breathing in a home setting, low Apgar score in a peripheral health facility or detailed biochemistry in a tertiary hospital). It is important to note that many infants will not have all markers (e.g. low Apgar but normal base deficit at birth or the reverse). Again, infants with gross birth defects should be excluded as well as small/immature infants. A cut-off birth weight of 2000 g was suggested. The primary cause of the poor condition in infants born below this weight would probably be immaturity rather than hypoxia. Good basic resuscitation would save most of these infants.

c. Infants who developed neonatal encephalopathy within the first 12 hours of life

Infants with severe physical trauma, gross birth defects or obvious severe infection should be excluded. The diagnosis of neonatal encephalopathy is usually reserved for term infants (37 weeks or more). Inherited metabolic disorders are rarely present in the first 12 hours of life. Neonatal encephalopathy is a syndromic diagnosis. Future research is urgently needed to relate simple clinical signs of abnormal neurological behaviour (e.g. poor suck, hypotonia and fits) to a “gold standard” of costly investigations (e.g. MRI and CT scanning and biochemical markers of multi-organ failure).

d. Neurodevelopmental disability at follow up

How and when this is measured is still hotly debated. No attempt was made to reach clear answers. However, screening at times of immunization (e.g. 14 weeks and about 1 year) and school entry would be most practical. Assessments could be based on examination or history (look or ask).

The total burden of death (mortality) plus disability (morbidity) following presumed hypoxia associated with an acute peripartum event must be measured. This would be an important marker for changes in both care of mothers in labour and infants after delivery.

2. Existing mortality data.

Mortality data can be collected in a number of different ways depending on the circumstances.

a. Vital registration

This is current practice in many well-resourced countries. It gives an accurate indication of the number of stillbirths, live births and neonatal deaths. The accuracy of cause of death is often questionable.

b. Hospital based data

This is a selective and biased method but helps to identify the causes of death as well as avoidable (modifiable) factors which can be addressed. This information may be more valuable than an accurate count of all deaths. Often only neonatal deaths that occur in hospital before the infants are discharged home are included. This may miss many deaths due to infection.

c. Sentinel sites in city, small town and rural sites

This gives a better overall view of causes of deaths, plus modifiable factors associated with these deaths, especially if each death is discussed at a mortality audit meeting.

3. Existing morbidity data.

This is far more difficult to collect than mortality data. Condition at birth, neonatal encephalopathy and neurodevelopmental disability would be obvious end-points.

4. Maternal health and birth asphyxia.

Many associated factors during pregnancy (e.g. pre-eclampsia) and labour (e.g. cephalopelvic disproportion) are well established. These would all be risk factors for an acute perinatal event resulting in mortality or morbidity of the infant (or foetus). Family and social factors also have an impact on maternal health (e.g. home support). Specific risk factors mentioned were the abuse of misoprostil to augment labour, the caesarean section rate and lack of a skilled attendant at birth.


1. A description rather than a definition of “birth asphyxia” is needed. The critical event is acute perinatal hypoxia. It is better to describe the specific event and that can lead to meaningful study of etiology and outcome which will be important for strategic planning to improve the situation

2. The total burden of associated mortality (fresh stillbirths and neonatal deaths following poor condition at birth and/or neonatal encephalopathy ) and morbidity (neurodevelopmental disability) should be used as the outcomes measure. This could be restricted to a specific weight category (e.g. 2000 g or more) to exclude deaths due to immaturity.

3. Limited data on mortality, and to a lesser extent morbidity, are available from developing countries but data collection systems require strengthening. Audits of causes and modifiable factors at selected sites may be more important than trying to count all infant births and deaths in a region.

4. Associated maternal health variables have been documented and should be used as risk factors rather than diagnostic markers.

5. The prevention, diagnosis and management of acute perinatal events that may damage the foetus and newborn infant remain major challenges which deserve urgent attention.

6. Research is needed to causally link associated pregnancy factors, acute events during labour, the infant’s condition at birth, neonatal encephalopathy and neurodevelopmental disability. Simple clinical signs for each stage are required to make a diagnosis in poor communities. These signs need to be validated against a “gold standard” such as base deficit (as a marker of tissue hypoxia). Steps to reduce the total burden of mortality and morbidity should be evaluated.

7. The term neonatal encephalopathy is preferable to the term hypoxic ischemic encephalopathy.

Breakout Group 2 was chaired by Zulfiqar Bhutta, Aga Khan University, Karachi, Pakistan and Ola Didrik Saugstad, National Hospital, Oslo, Norway. The 13 members of this group were asked to identify and prioritize research that is needed in the area of birth asphyxia.

The group emphasized that the diagnosis of birth asphyxia remains a central unsolved problem. Recently the World Health Association held a meeting to discuss the issue of definition and produced a report with recommendations for study of the problem.

The group felt that there is need for an operational definition of birth asphyxia. Studies could then be mounted in representative, “sentinel” settings where this question of the burden of the problem, and its significance in the context of neonatal mortality, morbidity, and outcomes in relation to handicap could be evaluated in a standardized fashion.

It is also possible, within these studies, to nest additional questions that pertain to issues of determinance and risk factors, including qualitative issues relating to care seeking, knowledge, attitudes, perceptions and community determinants. The methodology could be standardized and information could be compared across regions.

There is a need to consider related issues such as body size and micronutrient malnutrition, reproductive tract infections and systemic infections since all of these bear a relationship to birth asphyxia.

It is possible to identify prospective cohorts where a host of these factors could be assessed and broad burden estimates of birth asphyxia, intermediate and long-term outcomes, could be captured, but these would need to be set up prospectively. It was felt that there is opportunity to set this up in representative populations. Where are those representative populations or areas?

It is quite evident, from existing groups working in the regions, that in both South Asia and Africa there are already studies or interventions in place where cohorts could be developed and expanded to include additional information that would allow the capture of these data and, in perhaps 3-4 years, researchers could return with answers rather than “guesstimate” data from developing countries on how important this issue is, how it may matter to health systems and communities, and what is the real burden of short-term handicap that may be associated with this.

At least five sites were mentioned. In addition, some members of the group volunteered to find out from China and Latin America if there may be additional opportunities for including some sites there. So there is some possibility of complimentary studies.

The group then considered interventions, i.e. treatment of the asphyxiated child. There is sufficient evidence to conclude that perhaps where health systems have the capacity, there is a need to look at bag and mask ventilation. There was some question as to whether there is still an issue of the value of mouth-to-mouth resuscitation. In general, it was felt that perhaps there are sufficient barriers to this in HIV-endemic populations for this not to be a recommendation anymore and where the standard of care does not allow alternatives perhaps there could be some information captured on the impact of this in the hands of community health workers. In general the group felt that the evidence was strong enough to look at scaling up of bag and mask resuscitation in health systems in effectiveness trials to see its impact on reducing the burden of the problem or otherwise.

The group also felt that there are additional studies that are very important, but perhaps do not need to be conducted at scale in health systems. These studies pertain to important questions of whether or not suction is necessary; whether or not some better methods of circulatory support could be instituted; and whether or not blood pressure measurement as well as interventions could be done in a simplified manner. All of these questions, the group felt, could be explored through the opportunities that are available for collaborative research between developing countries and developed countries. So there is a possibility that these questions could be studied reasonably well in a collaborative fashion between developing and developed countries.

It was recommended that a catalogue of existing studies should be created. That would allow The Programme for Global Paediatric Research to draw up a list of opportunities for coordinating and communicating about the existing range of studies, to plan new studies, and to contact potential donors and funders to see what could be done in terms of incremental funding.

During the question and answer period that followed the breakout group sessions, two additional points about the topics discussed by Group 2 were raised.

Breakout Group 3 was chaired by Waldemar Carlo of The University of Alabama, Birmingham, U.S.A.and Vinod K. Paul of All India Institute of Medical Sciences, New Delhi, India. The 14 members of this group discussed resources that are needed for research on birth asphyxia in developing countries.

Training Tools

The group determined that training tools for mothers, families, communities, trained birth attendants (“TBA”s), midwives, and physicians need to be developed and evaluated. These tools need to cover late pregnancy care, intrapartum care, care of the neonate at birth and the first week, and follow-up outcomes. TBAs need simple tools that can be implemented at the community level. In many places, unskilled health care workers (e.g. nursing aides) are providing care. Simple training tools such as visual aides, pictures and other materials for illiterate birth attendants need to be developed and tested. Research is needed to identify the best ways to recognize and teach danger signs in the mother and baby associated with birth asphyxia.

WHO has training programs such as Making Pregnancy Safer and Pregnancy, Childbirth, Portpartum and Neonatal Care (“PCPNC”) Materials on essential maternal and newborn care have been developed for institutional care but they need to be adapted for training TBAs and other perinatal health care providers at the community level. Tools have to be standardized for the various levels of caretakers, including the most basic level (village/community) and should be adapted to the various regions, and countries. These tools need to be tested for efficacy and effectiveness. Research needs to be performed to develop and test simple, low maintenance and reusable resuscitation equipment that can be effectively used at institutional and home-birth level. In addition, tools for mothers, families, and communities that highlight danger signs and “essential” care are needed.

Members of human subjects/ethics review committees or institutional review boards need to be educated on resuscitation research. Tools need to be developed to assist researchers to perform asphyxia-related research. Training has to be practical and applicable to the developing world. The American Office for Human Research Protections (“OHRP”) and other research tools are available.

Pedagogic research needs to be performed to evaluate effectiveness of teaching and retention of knowledge and skills in prevention and treatment of birth asphyxia. Tools are needed to better determine the presence of birth asphyxia. Tools need to be applicable to institutional and community births in low to middle resource countries (Table).

Table: The availability of tools for measuring key research outcomes in asphyxia at the institutional and community levels in low to middle resource countries


Institutional Level

Community Level

 Birth asphyxia



 Stillbirth vs. neonatal death



 Asphyxia-specific death

± cause of death

?Verbal autopsy

 HIE assessment (term)



 HIE Assessment (preterm)



 Gestational age classification



 Long term outcomes/disability


?Ages and Stages Questionnaire

[ Tools generally available and used (+); not available (-); may be available but need improvement/adaptation (+); more information required to ascertain status (?)]

Tools to identify essential data variables are needed. Tools for quality care-seeking behaviour are needed. Tools are needed to evaluate the effect of birth asphyxia on neonatal death, on the mother/family and as a burden on society. Reliable data are needed on stillbirth rates, neonatal mortality rates, and birth asphyxia rates.

Research Units

Research units in developing countries need a major effort at building capacity that is sustainable. It is necessary to develop local experts who have the capacity to lead research. Capacity needs to be developed at the institutional and community levels for transfer of appropriate technologies. Countries and researchers need to identify priorities that research units should address. Partnership with organizations and institutions with research expertise frequently leads to completion of superior research, sustained funding, and long-term commitments. Partnerships work best when there is team work and mutual respect. Study protocols need to be developed jointly. Financial resources are needed to establish research units, but some successful research units have been established with limited local resources. Examples of successful collaborations to develop research units include multiple academic institutions and non governmental organizations, Saving Newborn Lives, the NICHD Global Network for Women’s and Children’s Health Research,  International Clinical Epidemiology Network (“Inclen”), and Pragmatic Randomized Controlled Trials in Health Care (“Practihc”).

Scientific sessions such as those held by the Pediatric Academic Societies raise the awareness of the need for research in asphyxia in the developing world. Scientific meetings need to be a source of novel ideas to reduce the burden of asphyxia in the developing world.

Collaboration between investigators should be promoted. Computer access is important for research in developing countries but is not readily available in many locations. A website portal for information would be helpful. The Programme for Global Paediatric Research (“PGPR”) should serve a catalytic role to promote and review research progress in neonatal health and survival research, including all aspects of research and implementation. PGPR should collaborate with the Society for Maternal and Fetal Medicine to promote the agenda of asphyxia research.

The group expressed the concern that there should be efforts to make sure the research agenda identified is promoted. PGPR should do advocacy work. It was proposed that PGPR continues the focus on asphyxia during the next five years or so.

Breakout Group 4 was chaired by Gary Darmstadt of Johns Hopkins University, Baltimore, U.S.A.and Alvin Zipursky of The Hospital for Sick Children, Toronto, Canada. The 18 members of this group were asked to determine how to further the development of international collaborations for the study of birth asphyxia in developing regions.

The group felt that whenever possible international research programs should try to work through national networks or specific existing programs; however, most existing national networks are in high income countries. An exception is India, but national networks need to be developed in other low income countries.

Networks brings great strength (resources and expertise), but they also present potential pitfalls. They may affect the independence of investigators; favour clinical and institutional elites and international programs rather than national researchers; and divert resources from national capacity building. These potential dangers must be recognized so that network planning proceeds as effectively as possible. 

The Programme for Global Paediatric Research (“PGPR”)TM needs to determine how it can best assist in the development of international communication and collaborations between researchers studying birth asphyxia.

Workshops, email communication and an interactive PGPR web site are essential for ongoing communication and the development of programmes. It was suggested that experts involved in PGPR could help in the development of specific research studies being developed by nationals, based on local priorities, by providing advice about plans and funds for those studies. PGPR itself can continue to act as an intermediary, connecting experts for consultations.

A formal plan for the study of the global crisis of birth asphyxia is needed. This plan is necessary as a guide and as a form of advocacy. The best chance of achieving a response from governments is by issuing a real plan. This notion is supported by the response to the Lancet Neonatal Survival Series which was published in March. Already, several African governments have expressed intentions to move ahead with programs identified in that series.

PGPR should distribute this formal plan for the study of birth asphyxia, as well as the proceedings from its meetings so that information is shared throughout the world. It was also suggested that besides issuing reports and statements on research needs, small groups could be created within PGPR to implement specific recommendations.

Although it is very important for those at the grass roots level to identify their own needs and call for action on the priority areas they identify, expertise often exists in tertiary centres. Those experts should have a major role in bringing programs to the grass roots level.

It is important to develop “champions” in countries. These are highly trained individuals who can assume responsibility for programs and advocate for resources.

The expertise in high income countries can contribute to the creation of research capacity in the mid and low income countries through: