Abstract: Data analysis is undoubtedly a major part in any general epidemic preparedness and control. Several statistical methods have been developed over the years for better and accurate analysis of epidemic data. Key among such methods is the use of stochastic epidemic models. This study used Bayesian technique to do inference for stochastic epidemics given data from CSM outbreak. The data for the analysis was a 2010 CSM outbreak data from Jirapa township in Ghana with a population of 6139 people and an ultimately infected number of 354 people. Conjugate and uninformative prior selections methods were considered for the study of which the former was seen to present better results based on the diagnosis analysis of results from both methods. In estimating the summaries of these parameters from the joint conditional distribution, the study adopted MCMC algorithms which was implemented in R using a combination of packages. In presenting the results, the study considered 2 parameters out of the several that can be estimated with the model. The results of the sampling of the exposure period E shows that the outbreak population has a very high exposure period with E having a median of 8.8965 days. This is indeed very high as compared to the WHO standard of between 2 and 10 days. Also, the rate of infection β had a median of 0.7334. The two parameter estimates are some of the factors that lead to the ultimate extinction of the outbreak before it could spread to all the entire township.
Keywords: Epidemics, CSM, stochastic, MCMC algorithm, random graph, Bayesian theorem, social network.
Title: MODELLING OF CEREBRAL SPINAL MENINGITIS (CSM) OUTBREAK USING STOCHASTIC EPIDEMIC MODELS WITH BAYESIAN PRIORS
Author: ABDUL BADI SAYIBU
International Journal of Mathematics and Physical Sciences Research
ISSN 2348-5736 (Online)
Research Publish Journals
