Potentially Modifiable Factors Associated with Death of Infants and Children with Severe Pneumonia Routinely Managed in District Hospitals in Malawi

Objective To investigate recognised co-morbidities and clinical management associated with inpatient pneumonia mortality in Malawian district hospitals. Methods Prospective cohort study, of patient records, carried out in Malawi between 1 st October 2000 and 30 th June 2003. The study included all children aged 0-59 months admitted to the paediatric wards in sixteen district hospitals throughout Malawi with severe and very severe pneumonia. We compared individual factors between those that survived (n = 14 076) and those that died (n = 1 633). Results From logistic regression analysis, predictors of death in hospital, adjusted for age, sex and severity grade included comorbid conditions of meningitis (OR =2.49, 95% CI 1.50-4.15), malnutrition (OR =2.37, 95% CI 1.94-2.88) and severe anaemia (OR =1.41, 95% CI 1.03-1.92). Requiring supplementary oxygen (OR =2.16, 95% CI 1.85-2.51) and intravenous fluids (OR =3.02, 95% CI 2.13-4.28) were associated with death while blood transfusion was no longer significant (OR =1.10, 95% CI 0.77-1.57) when the model included severe anaemia. Conclusions This study identified a number of challenges to improve outcome for Malawian infants and children hospitalised with pneumonia. These included improved assessment of co-morbidities and more rigorous application of standard case management.