Feasibility and acceptability of bubble continuous positive airway pressure oxygen therapy for the treatment of childhood severe pneumonia and hypoxemia in Bangladeshi children

Background Effective management of hypoxemia is key to re-ducing pneumonia deaths in children. In an intensive care setting within a tertiary hospital in Bangladesh, bubble continuous posi-tive airway pressure (bCPAP) oxygen therapy was beneficial in re-ducing deaths in this population. To inform a future trial, we in-vestigated the feasibility of introducing bCPAP in this population in non-tertiary/district hospitals in Bangladesh.Methods We conducted a qualitative assessment using a descrip-tive phenomenological approach to understand the structural and functional capacity of the non-tertiary hospitals (Institute of Child and Mother Health and Kushtia General Hospital) for the clinical use of bCPAP. We conducted interviews and focus group discus-sions (23 nurses, seven physicians, 14 parents). We retrospectively (12 months) and prospectively (three months) measured the prev-alence of severe pneumonia and hypoxemia in children attending the two study sites. For the feasibility phase, we enrolled 20 pa-tients with severe pneumonia (age two to 24 months) to receive bCPAP, putting in place safeguards to identify risk.Results Retrospectively, while 747 of 3012 (24.8%) children had a diagnosis of severe pneumonia, no pulse oxygen saturation in-formation was available. Of 3008 children prospectively assessed with pulse oximetry when attending the two sites, 81 (3.7%) had severe pneumonia and hypoxaemia. The main structural challeng-es to implementation were the inadequate number of pulse oxim-eters, lack of power generator backup, high patient load with an inadequate number of hospital staff, and inadequate and non-func-tioning oxygen flow meters. Functional challenges were the rapid turnover of trained clinicians in the hospitals, limited post -admis-sion routine care for in-patients by hospital clinicians due to their extreme workload (particularly after official hours). The study im-plemented a minimum of four hourly clinical reviews and provided oxygen concentrators (with backup oxygen cylinders), and auto-matic power generator backup. Twenty children with a mean age of 6.7 (standard deviation (SD) = 5.0)) months with severe pneumonia and hypoxemia (median (md) SpO2 = 87% in room air, interquar-tile range (IQR) = 85-88)) with cough (100%) and severe respiratory difficulties (100%) received bCPAP oxygen therapy for a median of 16 hours (IQR = 6-16). There were no treatment failures or deaths. Conclusions Implementation of low-cost bCPAP oxygen therapy is feasible in non-tertiary/district hos-pitals when additional training and resources are allocated.