Use this URL to cite or link to this record in EThOS:
Title: An assessment of the clinical effectiveness of World Health Organization guidelines for the management of pneumonia among hospitalised children in Kenya
Author: Agweyu, Ambrose
ISNI:       0000 0004 7224 5015
Awarding Body: University of Warwick
Current Institution: University of Warwick
Date of Award: 2017
Availability of Full Text:
Access from EThOS:
Access from Institution:
Background: The effectiveness of World Health Organization (WHO) guidelines for pneumonia case management in sub-Saharan Africa has been contested. This thesis aims to determine the clinical effectiveness of these guidelines among children admitted to Kenyan hospitals in a period after the introduction of the pneumococcal and Haemophilus influenzae type B (Hib) conjugate vaccines. The studies focus on the treatment of children with chest indrawing pneumonia, who were previously regarded to be at high-risk requiring inpatient treatment but were reclassified as lowrisk (non-severe) in the WHO guidelines updated in 2013. Methods: This thesis consists of: (i) A systematic review linked to a national guidelinedevelopment exercise appraising the evidence for the WHO pneumonia guidelines, (ii) a prospective observational study evaluating adherence to, and effectiveness of the pneumonia guidelines in the national referral hospital in Kenya, (iii) a multi-centre pragmatic randomised controlled trial (RCT) comparing amoxicillin versus benzyl penicillin for chest indrawing pneumonia (iv) a cohort study comparing treatment effects among children enrolled in the antibiotic RCT with a similar group who received routine care, and (v) a multi-centre retrospective cohort study of children hospitalised with pneumonia describing factors associated with mortality, focusing on characteristics that increase risk of death among children who would, under current guidance, be assigned a non-severe classification. Results: Although evidence from clinical trials supported the adoption of oral amoxicillin for severe pneumonia over benzyl penicillin (the standard treatment) for chest indrawing pneumonia, a Kenyan guideline development panel raised concerns of generalizability citing the limited data from sub-Saharan African populations in whom mortality was argued to be high. This concern was explored using prospectively collected observational data from 385 children. Treatment failure and mortality were infrequent (< 2%) for chest indrawing pneumonia where strict definitions requiring documented evidence of clinical deterioration were applied. In comparison, high rates of treatment failure (21.4%) and mortality (10.5%) were observed for severe pneumonia (formerly very severe pneumonia). Using propensity scores to model treatment effects comparing guideline recommended regimens with more costly, broad-spectrum alternatives, similar risks of treatment failure were observed in both groups. Amoxicillin was compared with benzyl penicillin in a pragmatic clinical trial of 527 children that also revealed low and comparable risks of treatment failure (8%) and mortality (< 1%) for the two treatments. Consistent results were observed in an observational cohort of children hospitalised at the same health facilities over the period the trial was conducted. However, analyses of data from > 16000 children suggested that the presence of commonly-occurring clinical signs may be associated with increased risk among children with non-severe pneumonia. Specifically, very low weight-for-age Z score (WAZ) or pallor in children with non-severe pneumonia were shown to be associated with absolute risks of mortality as high as those for severe pneumonia. Conclusions: Findings from locally-conducted observational studies and a clinical trial indicate low risks of treatment failure and mortality among children with chest indrawing pneumonia following treatment with benzyl penicillin monotherapy or amoxicillin. In contrast, mortality for severe pneumonia was greater than 10 percent. These results are consistent with the updated WHO recommendations and have more recently informed the revision of the national policy for pneumonia case management in Kenya. However, these guidelines may apply to sub-populations of children with non-severe pneumonia and either very low WAZ or pallor. This evidence is expected to contribute to ongoing debates on the adoption of WHO guidance for pneumonia case management in similar settings across sub-Saharan Africa where coverage of the Hib and pneumococcal conjugate vaccines is high.
Supervisor: Not available Sponsor: Wellcome Trust
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available
Keywords: RJ Pediatrics