Analysis of distribution characteristics and prediction model of hepatitis A incidence based on spatiotemporal big data

Objective To understand the spatiotemporal distribution of hepatitis A incidence in China, and to assess the feasibility of ARIMA product season model in the prediction of hepatitis A incidence. Methods Data about hepatitis A incidence and demography from 2006 to 2017 were collected, spatiotemporal distribution diagram of hepatitis A incidence was drawn by using MapInfo 11.0, and ARIMA model for the 2006-2016 incidence series was established by using SPSS 23.0. Data of 2017 incidence were used to evaluate the prediction accuracy of the model. Results In 2006-2016, a total of 430 962 cases of hepatitis A were reported nationwide, the incidence declined with each passing year, and the western was significantly higher than the eastern region. The optimal model ARIMA(0, 2, 2)(0, 1, 1)₁₂ was established after data smoothing, leveling, parameter estimation and model test.The prediction results of hepatitis A incidence were roughly consistent with the observations in 2017, and the relative error was between 2.0% and 39.7%. Conclusions The incidence of hepatitis A has been decreasing by years and shown significant differences between the eastern and western regions in China. The ARIMA product season model is suitable for forecasting the hepatitis Aincidence in short-term.