CPAP Improving Mortality for Pneumonia in African Children Trial

Despite laudable reductions in global childhood mortality rates, pneumonia remains the second most frequent killer of children less than five years old worldwide. Nearly one million children succumbed to pneumonia in 2013, with greater than half of these deaths in Africa. In Malawi, which has a high prevalence of malnutrition and Human Immunodeficiency Virus (HIV) infection, pneumonia is a major cause of pediatric mortality. In patients with World Health Organization (WHO)-defined severe pneumonia, malnutrition, HIV-infection, and hypoxemia are the primary drivers of poor outcomes. In a recent analysis of 2001-2012 child pneumonia outcomes in Malawi the overall case fatality rate decreased from 15% to 4% except in children with severe malnutrition. The pneumonia mortality rate in malnourished children remained elevated at 15% despite antibiotics and increased access to supportive interventions like low-flow supplemental oxygen. In Malawian children with HIV-infection, WHO very severe pneumonia and severe malnutrition were the strongest predictors of death. Severe hypoxemic pneumonia may be as common in children as HIV-affected or severely malnourished cases, and may also have higher mortality than non HIV-affected, non-severely malnourished cases without severe hypoxemia. Non-invasive ventilation, already routinely used in industrialised countries, may provide an advanced treatment solution for certain patient populations such as children with WHO severe pneumonia complicated by severe malnutrition and/or HIV-infection or -exposure or severe hypoxemia. Bubble continuous positive airway pressure (bCPAP) is non-invasive and is widely used for preterm neonatal respiratory failure in industrialised countries. Along with a flow generator, bCPAP uses a water column to deliver continuous positive pressure to a spontaneously breathing child. bCPAP is relatively inexpensive and requires little technical expertise compared with mechanical ventilation, but there is limited experience of bCPAP in resource-poor settings. Recently, small studies have explored its use in preterm neonates in Malawi. However, few studies have described its use in older infants and children, none of which included mortality as a primary endpoint and specially focused on the main drivers of poor pediatric pneumonia outcomes in southern Africa, HIV, malnutrition, and hypoxemia. The investigators data using bCPAP in Malawian children with severe pneumonia suggest feasibility for implementation. The investigators have previously reported that using a bCPAP system derived from locally available, relatively inexpensive supplies has shown promise in the management of hospitalized HIV-infected children with pneumonia in Malawi. The investigators observational case series further delineates the outcomes of 77 Malawian children hospitalized at a tertiary referral facility with severe pneumonia who were treated with bCPAP. Nearly half were infants either infected or exposed to HIV or were severely malnourished. Although the mortality of this series of patients was 50.0%, bCPAP was initiated in this cohort only when patients were found to be failing standard treatment. The investigators estimated that more than 75% of these children would have been eligible for mechanical ventilation. In this proposed study the investigators will be initiating bCPAP earlier in the hospitalization prior to treatment failure. Unlike previous studies conducted at referral hospitals, the investigators will perform this study at the district hospital level where 80% of hospitalized child pneumonia cases are cared for in Malawi. Although bCPAP is relatively inexpensive, scale-up in countries like Malawi with significant pneumonia burden and high HIV prevalence will require substantial resources to meet expected needs. In order to appropriately allocate precious resources and provide practical clinical guidance for healthcare providers who may use bCPAP, it is paramount to fully understand the utility of bCPAP treatment in this setting. To the investigators knowledge no bCPAP data using a control group with mortality as the primary outcome has been reporting in a similar generalized HIV epidemic African patient population 1-59 months of age. Data generated from this research will be additionally critical for formulating future studies that may include bCPAP refinements or exploration of other feasible modalities like high-flow nasal cannula or bi-level positive airway pressure. Therefore, the more rigorous methodology proposed here is warranted and supported by the Malawi Ministry of Health. If bCPAP proves an effective treatment modality for children hospitalized with WHO severe pneumonia, it is a simple technology that could be operationalized to help thousands of children with life-threatening pneumonia. The investigators propose to address this critical evidence gap by conducting a randomized controlled study determining bCPAP outcomes, compared to the currently recommended standard of care endorsed by the WHO and Malawi Ministry of Health, in hospitalized Malawian children with WHO-defined severe pneumonia complicated by malnutrition and/or HIV-infection or -exposure, or severe hypoxemia. RATIONALE Quality randomized studies comparing bCPAP versus a standard-of-care control group that includes low-flow oxygen therapy and using a primary endpoint of mortality are not available in low-resource settings including high prevalence HIV countries like Malawi for children 1-59 months of age with severe pneumonia. Demonstrating a mortality benefit with bCPAP is needed to support further investment and scale up of bCPAP in the care of older Malawian children 1-59 months of age with World Health Organization (WHO) severe pneumonia complicated by HIV and/or malnutrition, or severe hypoxemia. STUDY HYPOTHESIS AND OBJECTIVES • Study Hypotheses The investigators hypothesize that bCPAP, compared to standard care, will reduce the mortality of Malawian children with WHO-defined severe pneumonia complicated by a severe co-morbidity (HIV-infection or HIV-exposure and/or severe malnutrition), or severe hypoxemia without a severe co-morbidity. • Study Objectives The broad objective of this study is to provide scientific evidence assessing the effectiveness of treatment with bCPAP for WHO severe childhood pneumonia for children 1-59 months of age in Malawi, Africa. * Primary Objective 1 • Determine the pneumonia mortality rate for bCPAP treatment, compared to standard of care, for children with WHO severe pneumonia. * Primary Objective 2 • Determine the pneumonia mortality rate for bCPAP treatment, compared to standard of care, for children with WHO severe hypoxemic pneumonia without co-morbidity (i.e., no HIV infection, no HIV-exposure, no severe malnutrition). * Primary Objective 3 • Determine the pneumonia mortality rate for bCPAP treatment, compared to standard of care, for children with WHO severe pneumonia and co-morbidity (i.e., HIV-infection or HIV-exposure and/or severe malnutrition). * Secondary Objectives * Determine the pneumonia day 14 treatment failure rate for bCPAP treatment, compared to standard of care, for HIV-infected children with WHO severe pneumonia. * Determine the pneumonia day 14 treatment failure rate for bCPAP treatment, compared to standard of care, for HIV-exposed, uninfected children with WHO severe pneumonia. * Determine the pneumonia day 14 treatment failure rate for bCPAP treatment, compared to standard of care, for severely malnourished children with WHO severe pneumonia. * Determine the pneumonia day 14 treatment failure rate for bCPAP treatment, compared to standard of care, for severely hypoxemic children with WHO severe pneumonia. * Determine the proportion of children alive at day 30 phone follow up. * To investigate whether there may be differential treatment responses in certain baseline characteristics including but not limited to children with severe anemia, in those who test positive for malaria, in those with wheeze at baseline, in those with altered mental status, in those with digital auscultation-defined lung disease, and whether there is a differential treatment response by age. * To determine whether intervention arms have equivalent rates of adverse events as control arms.