A review of some of the latest low-resources emergency medicine evidence

High flow vs low flow oxygen vs permissive hypoxia in children with pneumonia, an RCT.

Randomised controlled trial of oxygen therapy and high-flow nasal therapy in African children with pneumonia

Maitland, K., Kiguli, S., Olupot-Olupot, P. et al.  Intensive Care Med 47, 566–576 (2021). https://doi.org/10.1007/s00134-021-06385-3


To establish if high flow nasal oxygen can prevent mortality in children with pneumonia and to establish if any oxygen administration is beneficial in children with pneumonia and moderate hypoxia (80-91%)


Pneumonia is a common killer in patients under 5. It is unclear whether oxygen can prevent acute deaths in these patients. It is also unclear to which degree of hypoxia is dangerous. Oxygen is a scarce resource in many hospitals and clinics in Africa and it would be helpful to know how best to use it. 


Open label, 2 centred, 2 stratum randomised trial.


6 hospitals in Kenya and Uganda


Children aged 28 days to 12 years and fulfilled one of the WHO criteria for severe pneumonia (cough +Difficult breathing and either chest indrawing, grunting, nasal flaring). Excluded; uncorrected cyanotic heart disease, chronic lung disease (not asthma) 1842 children enrolled.


2 Stratrums. Those with severe hypoxia (<80%) were randomised to either high flow(optiflow) or low flow (standard oxygen therapy).

Those with moderate hypoxia were randomised to high flow(optiflow) or low flow (standard oxygen therapy) or permissive hypoxia (no oxygen)

Common standard treatment; antibiotics, antimalarials, antipyretics, anticonvulsants and transfusion as per standard guidelines .

Per protocol analysis undertaken.

Primary outcome

Death at 48 hours AND to 28 days.

Secondary outcomes

Treatment failure at 48 hours( sats< 92 and respiratory distress), time to resolution of hypoxia, duration of oxygen, neurocognitive sequelae, readmission and anthropomorphic status.


In the hypoxaemia stratum, 4 (1.1%), 9 (2.5%) and 10 (1.4%) children in HFNT vs LFO vs permissive hypoxaemia had died.

No significant survival benefit for liberal versus permissive hypoxia. By 48 h in the severe hypoxaemia stratum, 18 (9.3%) in HFNT versus 26 (13.4%) in LFO groups had died.

No significant difference in high vs low flow oxygen was detected. Neurocognitive sequelae largely transient.

Author's Discussion

They could not achieve selected aims as study terminated early and also as on survival benefit demonstrated from high flow oxygen, but this could be due to insufficient numbers. However they noted that this study does question the liberal oxygen approach and raised the possibility of using high flow devices as an oxygen conserving strategy and in potentially reducing mortality


Reasonable to be open-label Ambitious study answering important questions in a low-resource setting.

Hard, patient-orientated outcome relevant to emergency medicine.  

Very large study (massive in terms of paediatric low-resource settings.

High adherence to protocol. Very low loss to follow up.

Resourced very standard for a typical moderately resourced African hospital

Broad inclusion criteria make it very generalisable in low-resource settings.


Terminated early due to concerns of the Ugandan team, therefore unfortunately underpowered.

Difference in expected morality from power calculations.

Approximately 70% had radiological pneumonia, it is possible hypoxia would have very different effects based on underlying pathology

My humble opinion

Very well conducted and informative study despite limitations of early termination. It is reassuring that if you do have oxygen to provide this may be less harmful than generally thought, however not enough evidence to stop providing oxygen if it is able, and usually standard oxygen therapy should be sufficient (and usually this is all that is available. Interestingly, hypoxaemia could be terminated by high flow air alone; this is an interesting avenue to explore for devices that can blow air without oxygen available or to minimise oxygen use. I will continue to target sats in the low 90’s when oxygen is available.

Written by Daniel Roberts

Any questions or comments @notdocdan on twitter, or by email me directly.