Pneumonia in rural Malawians under five years old: Treatment outcomes and clinical predictors of death on admission

BACKGROUND: High mortality and disability due to pneumonia occur worldwide. The introduction of the Integrated Management of Childhood Illness strategy in Malawi brought with it hope of an improvement in the outcome of pneumonia. However, the risk of death and treatment outcomes remain unknown in many districts. METHOD: The medical records of 466 consecutive patients admitted to the Mchinji District Hospital from January 2004 to January 2006 whose disease met the World Health Organization criteria for pneumonia were reviewed. Data were collected from forms that had been filled out and different treatment outcomes and determinants of death were analysed using logistic regression. RESULTS: Of the 466 patients, 62.7% completed treatment, 15.9% had unknown outcomes, 12.9% died, 8.4% were lost to follow-up, 0.8% failed to improve with treatment, and 0.4% were transferred to other facilities. Independent predictors of death were: age less than 2 years, female sex, history of pneumonia, chest retractions, type of pneumonia, and central cyanosis. CONCLUSION: A high proportion of deaths and unknown outcomes occurred among participants. Young age, female sex, history of pneumonia, chest retractions and central cyanosis were associated with death. Mortality from pneumonia may be reduced by close monitoring of these risk factors and by improving health education programmes and communicating these findings to parents and health workers. Further investigations of local reasons for high rates of unknown/unreported outcomes are welcomed.