Analysis of prediction for malaria cases by time series fitting in Jiangsu province,1980-2023
Objective To investigate the quantitative changes of malaria cases in Jiangsu province from 1980 to 2023,to predict malaria cases by time series fitting.Methods The monthly bulletins about malaria cases from Public Health Science Data Center and Jiangsu Commission of Health were retrieved.The quantitative changes of malaria cases were investigated.The suitable autoregressive integrated moving average(ARIMA)model and exponential smoothing model were sifted with criterion of parameters.The prediction effects were compared by relative errors.Results Since 1980,there was a significant decreasing tendency in malaria cases in Jiangsu province.The yearly cases peak was 296 844 in 1980,whereas the minimum yearly value was 37 in 2021 and 2022.The monthly cases peak was 77 198 in September 1980,whereas there was no reported case in July 2020,October 2021,February and March of 2022.Most of the malaria cases were concentrated from June to October before 2012.The absolute values of relative errors of monthly predicted values in 1989 with the exponential smoothing models were less than the ARIMA models,hence the prediction effects were better for the exponential smoothing models.The absolute values of relative errors of monthly predicted values in 1999 and 2011 with the ARIMA models were less than the exponential smoothing models,hence the prediction effects were better for the ARIMA models.The absolute values of relative errors of monthly predicted values in January,September and November of 2023 with the ARIMA models were less than the exponential smoothing models,whereas the absolute values of relative errors of monthly predicted values in February to August,October and December of 2023 with the exponential smoothing models were less than the ARIMA models.Conclusion The time series models were helpful in surveillance and early warning for malaria The comprehensive measures should be carried out to prevent and control of malaria after elimination of the indigenous malaria cases.
MalariaCaseTime seriesAutoregressive integrated moving average modelExponential smoothing model
NETL
NSTL
万方数据
