A stochastic model for early identification of infectious disease epidemics with application to measles cases in Bangladesh.

Abstract

In this article, a stochastic modeling approach was employed for the detection of epidemics in advance that was based on a negative binomial model with 2 components: an endemic component and an epidemic component. This study used monthly measles cases from January 2000 to August 2009 collected from the Expanded Program on Immunization, Bangladesh. General optimization routines provided the maximum likelihood estimates with corresponding standard errors. The negative binomial model with both seasonal endemic and epidemic components was shown to provide adequate fit with no measles epidemic during September 2008 to August 2009.