MICA: Children's Oxygen Administration Strategies Trial

The Children's Oxygen Administration Strategies Trial (COAST) is a multicentre randomised controlled trial aimed at identifying which children admitted to hospital with suspected pneumonia with a low level of oxygen in their blood (called hypoxia) would benefit from oxygen and whether oxygen is best delivered by low flow (routine care) or by high flow. Although oxygen is a basic element of hospital care, there are no relevant studies to guide which level of oxygen saturation should be target for its use and what is the best method of how to administer it (low flow or high flow) to improve outcome. In practise many children in low-income countries do not receive oxygen, despite being recommended, owing to the lack of its availability due to the high cost, or supplies that are unpredictable (erratic delivery of cylinders or electricity to power oxygen concentrators) resulting in mismatch between supply and demand.

We will enrol 4200 African children, aged 2 months to 12 years, at admission to hospital with respiratory distress complicated by low oxygen levels (defined as a blood oxygen saturation, SaO2 level below 92%) over 30 months in 4 hospital in 2 countries (Uganda and Niger) and follow the children up over a period of 28 days.

COAST trial will evaluate two related components of management.

1/ Who to give oxygen too?
Children who are admitted to hospital in Africa with hypoxia have a poorer in-hospital outcome than children with normal oxygen levels. For children with hypoxia (oxygen saturations between 80% and 92%) between 9-10% will die in hospital, similarly 26-30% will die if they have severe hypoxia (oxygen saturations <80%). COAST will examine whether or not oxygen improves outcome in hypoxia and the best threshold for giving oxygen in children without severe hypoxia (oxygen saturations 80% - 92%). Children with severe hypoxia (SaO2 <80%) will all receive oxygen as we less unsure about the benefits of oxygen in this group. For the children with SaO2 between 80 and 92% we are not certain which is the best level to provide oxygen and whether this will result in a better outcome - so half of the children will receive oxygen the other half will not receive oxygen.

2. How best to give oxygen?
We will compare whether giving oxygen through a tube with two small prongs into the nose low flow (standard of care) to high flow in all children receiving oxygen. High flow oxygen provides extra pressure to the airways to prevent them from collapsing after every expiratory breath. High flow is safe and well tolerated in children and babies - as it helps reduce effort of breathing in critically sick children, which is substantial when lungs are congested with infection, that often leads to respiratory exhaustion and ultimately respiratory failure in the children who cannot access mechanical ventilation (the majority of hospitals in Africa).
The major aim (or outcome) is to reduce shorter-term mortality at 48-hours (primary endpoint) and longer-term morbidity and mortality to 28 days.

A trial demonstrating that oxygen is an important life saving treatment will provide important new evidence for which level of oxygen saturation to target oxygen therapy that is both clinically beneficial and cost-effective for health services. This would lead to substantial refinements to treatment recommendations and can be used to put pressure on health services for wider implementation - allowing policymakers to make decisions on how best to allocate scarce health resources.

COAST is a randomised controlled trial designed to provide the evidence for the most clinically-effective and cost-efficient targeted use of oxygen as a life-saving treatment with respect to the optimal oxygen saturation threshold for treatment and mode of delivery in 4200 African children, aged 2 months to 12 years, presenting to hospital with respiratory distress complicated by hypoxia (oxygen saturation (SaO2)< 92%).
Children will be enrolled over 30 months on admission to hospital at 4 sites in 2 countries (Uganda and Niger) and followed for 28 days (a sub-sample to 90 days). The trial has a pragmatic design, to ensure this encompasses a spectrum of high-risk children, identified largely by clinical criteria, so that the results are applicable to health services in Africa, with limited-resources.
The trial will have two randomisations:
R1: For children with a baseline SaO2<92% but >or=80%, randomisation to liberal oxygenation versus a more conservative strategy allowing permissive hypoxia (routine care)
R2: For children with a baseline SaO2<80% and for those randomised to liberal oxygenation for R1, high flow versus low flow delivery (routine care)
R1 addresses the question of what threshold to give oxygen in children with hypoxia (above a threshold of SaO2>or=80%); R2 addresses how best to give oxygen all receiving to oxygen. For R1, a cut off of <92% assesses a range of saturations that would be consistent with international recommendations for use of oxygen in sick children to provide additional data about those at a higher threshold than currently recommended by WHO. Children randomised to permissive hypoxia will receive low flow oxygen if they ever develop severe hypoxia (SaO2<80%). All randomisations will be open.
Primary outcome is mortality at 48 hours Secondary outcomes include survival to 28 days; time to hypoxia resolution; duration of respiratory support to 28 days; severe respiratory failure at 28 days; neurocognitive sequelae at 28 days.

The direct beneficiaries will be African children hospitalised with severe illness and hypoxia. The trial will also contribute new knowledge to the relevant academic disciplines as well as new opportunities for additional sub-studies to increase our knowledge and understanding of host response to the oxygenation strategies. In addition to generating new knowledge these will provide opportunities for more junior members of the trial teams to participate or lead and in particular to develop research outlines for higher degrees and gain experience in writing proposals for funding. COAST will therefore provide substantial intellectual and academic capacity development within Africa.

In order for benefits to be realised, we plan to engage with (1) national policymakers (Ministry of Health); (2) international policymakers (WHO, UNICEF); (3) nursing and paediatric associations, and NGOs involved in providing treatment or advocacy, in order to disseminate the outputs from the trial. These stakeholders will benefit from our research, which aims to provide clear evidence on optimal treatment of children with hypoxia, which is currently lacking. The trial interventions may provide substantial refinements to current treatment recommendations. If the targeted use of oxygen reduces mortality and recurrent morbidity, and is efficient and cost-effective for health services, this could greatly improve overstretched acute child health services. Whilst a single guideline is anticipated, refinement of this will be possible for particular sub-groups where response is not uniform.

In order to influence relevant policymakers and practitioners, we will also need to engage with academic audiences (who influence WHO policy) and trainers. The involvement of ISARIC will ensure we are engaging with key audiences throughout the duration of the trial, to ensure they are aware of our research and when results will be available. Results from the trial may spur further research and academic debate, helping to catalyse further research on (cost-)effective interventions and improvement of acute care for children in low-income settings.

A key element of our pathway to impact is ensuring the results are available in formats that are accessible and appropriate to our key audiences. We will publish the results in prestigious, open access, peer reviewed journals, and presenting them at relevant international conferences to reach the academic audience and provide the quality assurance needed by guideline developers. We will also present the results in other formats, such as policy briefs aimed at national policymakers and provide clear interpretation of the results that can feed into training programmes. Following the success of the FEAST trial video in communicating the results of that trial, we will produce a video that can be used for communicating the results to communities and health professionals.

The first strategic approach for achieving impact has already started: ensuring the trial addresses priority issues for the target audiences and will provide results that are generalisable. This has been central to the design of the trial. The question was identified as a research priority by International Severe Acute Respiratory Infection Consortium (ISARIC) - which is a global consortium of researchers of severe respiratory illness including representation from policy makers (http://isaric.tghn.org/). The questions that the COAST trial addresses have been identified as key gaps in the current evidence by this group and also in systematic reviews and by policy makers. The trial has a pragmatic design, to ensure results are applicable to health services in Africa, with limited access to health technologies. The health economics component is also vital to ensuring that evidence is provided not just on clinical effectiveness, but also cost-effectiveness, to allow policymakers to make decisions on how best to allocate scarce health resources.