Send to

Choose Destination
Lancet Respir Med. 2019 Nov;7(11):964-974. doi: 10.1016/S2213-2600(19)30243-7. Epub 2019 Sep 24.

Bubble continuous positive airway pressure for children with high-risk conditions and severe pneumonia in Malawi: an open label, randomised, controlled trial.

Author information

Eudowood Division of Pediatric Respiratory Sciences, Department of Pediatrics, School of Medicine, Johns Hopkins University, Baltimore, MD, USA; Department of International Health, Bloomberg School of Public Health, Johns Hopkins University, Baltimore, MD, USA. Electronic address:
University of North Carolina Project Malawi, Lilongwe, Malawi; Department of Pediatrics, School of Medicine, University of North Carolina at Chapel Hill, NC, USA.
Division of Emergency Medicine, Cincinnati Children's Hospital Medical Center, Cincinnati, OH, USA; Department of Pediatrics, University of Cincinnati College of Medicine, Cincinnati, OH, USA.
University of Utah School of Medicine, Salt Lake City, UT, USA.
University of North Carolina Project Malawi, Lilongwe, Malawi.
BEAD Core, Johns Hopkins University, Baltimore, MD, USA.
Malawi Ministry of Heath, Lilongwe, Malawi.
University of North Carolina Project Malawi, Lilongwe, Malawi; Division of Infectious Disease, University of North Carolina at Chapel Hill, NC, USA.



Pneumonia is the leading cause of death among children globally. Most pneumonia deaths in low-income and middle-income countries (LMICs) occur among children with HIV infection or exposure, severe malnutrition, or hypoxaemia despite antibiotics and oxygen. Non-invasive bubble continuous positive airway pressure (bCPAP) is considered a safe ventilation modality that might improve child pneumonia survival. bCPAP outcomes for high-risk African children with severe pneumonia are unknown. Since most child pneumonia hospitalisations in Africa occur in non-tertiary district hospitals without daily physician oversight, we aimed to examine whether bCPAP improves severe pneumonia mortality in such settings.


This open-label, randomised, controlled trial was done in the general paediatric ward of Salima District Hospital, Malawi. We enrolled children aged 1-59 months old with WHO-defined severe pneumonia and either HIV infection or exposure, severe malnutrition, or an oxygen saturation of less than 90%. Children were randomly assigned 1:1 to low-flow nasal cannula oxygen or nasal bCPAP. Non-physicians administered care; the primary outcome was hospital survival. Primary analyses were by intention-to-treat and interim and adverse events analyses per protocol. This trial is registered with, number NCT02484183, and is closed.


We screened 1712 children for eligibility between June 23, 2015, and March 21, 2018. The data safety and monitoring board stopped the trial for futility after 644 of the intended 900 participants were enrolled. 323 children were randomly assigned to oxygen and 321 to bCPAP. 35 (11%) of 323 children who received oxygen died in hospital, as did 53 (17%) of 321 who received bCPAP (relative risk 1·52; 95% CI 1·02-2·27; p=0·036). 13 oxygen and 17 bCPAP patients lacked hospital outcomes and were considered lost to follow-up. Suspected adverse events related to treatment occurred in 11 (3%) of 321 children receiving bCPAP and 1 (<1%) of 323 children receiving oxygen. Four bCPAP and one oxygen group deaths were classified as probable aspiration episodes, one bCPAP death as probable pneumothorax, and six non-death bCPAP events included skin breakdown around the nares.


bCPAP treatment in a paediatric ward without daily physician supervision did not reduce hospital mortality among high-risk Malawian children with severe pneumonia, compared with oxygen. The use of bCPAP within certain patient populations and non-intensive care settings might carry risk that was not previously recognised. bCPAP in LMICs needs further evaluation before wider implementation for child pneumonia care.


Bill & Melinda Gates Foundation, International AIDS Society, Health Empowering Humanity.

Supplemental Content

Full text links

Icon for Elsevier Science Icon for PubMed Central
Loading ...
Support Center