Paediatrics (RCH) - Theses

Permanent URI for this collection

Search Results

Now showing 1 - 1 of 1
  • Item
    Thumbnail Image
    Improving oxygen therapy for children and newborns in Nigerian hospitals
    Graham, Hamish Robert ( 2018)
    Oxygen is a long-established medical therapy that can be life-saving for severely ill children admitted to hospital. Effective provision of oxygen to patients requires a multifaceted system that involves technical, clinical, supply chain, financing, and other managerial and policy elements. However, this system is currently difficult to achieve in many low-resource settings resulting in poor access to, and use of, oxygen and subsequent excess mortality. Previous work has shown that improved hospital oxygen systems can improve patient access to oxygen therapy and reduce inpatient case fatality rates from childhood pneumonia. However, studies have shown variable impact in different settings due to contextual factors that are not well understood. In addition, the burden of hypoxaemia and impact of oxygen on conditions other than pneumonia is unclear – particularly in the large neonatal population. This study aimed to understand how to improve oxygen systems in particular contexts, and generate evidence to support the scale-up of effective oxygen systems in Nigeria and globally. My realist review of past oxygen projects developed a theoretical framework describing how improved oxygen systems could improve clinical outcomes in particular contexts – highlighting the interaction between efforts to improve oxygen access and the clinical use of oxygen. My oxygen needs assessment in 12 secondary-level Nigerian hospitals provided detailed data on existing oxygen systems – highlighting the impact of poor power supplies, weak maintenance systems, and lack of pulse oximetry. My prospective cohort study provided new data on the epidemiology of hypoxaemia in the Nigerian context. This study showed high prevalence of hypoxaemia among admitted children and neonates with a range of conditions, highlighted the strong association between hypoxaemia and death, and demonstrated poor accuracy of clinical signs to predict hypoxaemia (particularly for children with non-respiratory conditions). My mixed-methods realist evaluation identified how pulse oximetry could be most effectively adopted into routine paediatric and neonatal care – highlighting the role of key influencers to model behaviour, practical training and ongoing encouragement, personal experience of benefit, and the reasons why nurses valued pulse oximetry. My stepped wedge trial evaluated the effect of our intervention (improved oxygen system) on clinical outcomes and care practices. This study demonstrated mortality benefit for children admitted with pneumonia, and suggested that the introduction of pulse oximetry generated most of this benefit by stimulating better use of existing oxygen supplies. We found no mortality benefit for children with other conditions or neonates, and detected an unexpected trend towards higher mortality in the “full oxygen system” period compared to the “pulse oximetry period” for neonates. Post-hoc analysis tested some potential explanatory theories for these findings – highlighting the effects of pre-existing oxygen access and external factors. In conclusion, improving oxygen systems is complex. The studies contained in this thesis have helped fill evidence gaps that are hindering oxygen policy and planning decision-making in Nigeria and globally. They have directly informed national and global policies and program planning.