Bayesian survival analysis of under-five pneumonia patients in tercha general hospital, dawro zone, snnpr, south west Ethiopia

Abstract

Pneumonia is the number one largest infectious cause of death in children worldwide. More than 150 million cases of pneumonia occur in each year and it kills about 2,500 children every day. It is most prevalent in South Asia and sub-Saharan Africa. In Ethiopia, pneumonia is a leading single disease killing under-five children. Pneumonia is ranked as first cause of morbidity and mortality of children in Dawro zone and its prevalence is 27.81%. The aim of this study was to investigate the survival rate of under-five pneumonia patients in Tercha General Hospital using Bayesian and classical survival analysis. Methodology: Retrospective study was conducted in Tercha General Hospital from September 2016 up to August 2017. Children whose age greater than 29 days and less than five year were included in the study and Patients with insufficient information were excluded from the study. Bayesian Survival analysis is a statistical method for data analysis of time to event data by introducing external information in terms of the prior distribution. The Semi-parametric, classical parametric models and Bayesian parametric models are used for the analysis. Result: The Weibull Accelerated failure time model is good model compared to lognormal and Log-logistic models in both Classical and Bayesian approach based on AIC and DIC evidence respectively. The results implied that patients whose residence were urban and patient nurse ratio (PNR) were prolong timing death of under-five pneumonia patients, while season of diagnosis were Spring and summer, patients with comorbidity and patients with severe acute malnutrition (SAM) were statistically significantly shorten timing of death of under-five pneumonia in Tercha General Hospital in both Classical and Bayesian approach analysis. Conclusion: Finally, the results from both classical and Bayesian approach analysis showed that sex, residence, season of diagnosis, comorbidity, severe acute malnutrition (SAM), patients refer status and patient nurse ratio (PNR) were found to be significant predictors for survival time of patients in Tercha General Hospital. The researchers who are interested to investigate on similar area recommended applying Bayesian analysis by introducing frailty modelling.