Abstract

*Jiro E. Jay, Shuo Chang wu and Jason Ying-Jeou

This study is in order to explore the effect of cumulative sum (CUSUM) in the early warning of infectious diseases. The infection percentile of chickenpox in 2005 to 2009 was set as baseline. Using CUSUM for the early warning of chickenpox infection in 2010, the CUSUM parameter was determined in Wuhan City. CUSUM showed high specificity for the warning of a special time point; its low sensitivity can be compensated by reducing baseline or t value. The percentile method can promote the effect of CUSUM on the early warning of prevalent diseases. In Wuhan, when Ci is less than 10, the disease is not considered prevalent, but when Ci is more than 20, the disease is likely to be prevalent. In combination with the percentile method, CUSUM has high sensitivity and specificity for the early warning of infectious diseases and is useful for the prediction of the prevalence of diseases.