A review of statistical models for early detection of outbreaks of disease epidemics and application

Abstract

Statistical models provide a unique description to available data from public health surveillance systems, which can provide meaningful measures of population risks for disease, disability and death. Analysis and evaluation of these data help public health practitioners react to important health events in a timely manner both locally and nationally. Different methods exist for monitoring health surveillance data, and no method is universally superior. This paper discusses some methods commonly used for detection of outbreaks of disease epidemics and demonstrates the applicability two of these models on hospital data. it compares the performance of EWMA and CUSUM models for detection of outbreak of disease epidemics and finds that EWMA chart is slightly superior to CUSUM models for early detection of outbreaks of malaria and measles.