The Epidemiological Characters, The Construction And Evaluation Of Prediction Model On Hand-foot-mouth Disease In Changsha, China

OBJECTIVES:To investigate the epidemiological characters of hand-foot-mouth disease (HFMD) in Changsha, and to analyze the correlation between the incidence of HFMD and meteorological factors, and to establish and assess a model to predict the incidence rate of hand-foot-mouth disease. The aim of this study is to provide some scientific basic evidences for public health resource allocation and HFMD surveillance and control in Changsha.METHODS:The surveillance data of HFMD epidemic was collected and digital maps at the township (1:4000000) levels and epidemiological data were collected. The epidemiological characters of HFMD were described by using the indicators of prevalence and constituent ratio. The annual incidences of HMFD in each district of Changsha were calculated and mapped from2008to2013respectively. Each region was marked with different color on the county-level difital map. The Chi-Square test analysis, Spearman correlation analysis, Cross correlation analysis and time-series analysis were used to process data. Soft wares used in this study included SPSS18.0, SAS9.1.RESULTS:1. Reported cumulative incidences of HFMD from2008to2013in Changsha were109.86/100000,118.21/100000,337.40/100000,287.60/100000,363.56/100000,276.17/100000respec-tively. Among these,147were severe cases with26deaths.2. A large summer peak of HFMD occurred between April and July with a second smaller peak between October and December. The severely and death cases were most frequently seen between April and July.3. The patients under6years of age were at the highest risk, with higher incidence of severe cases and death cases. 4. There was a male predominance of HFMD cases, with a male-to-female ratio between1.44:1and2.06:1.5. The children who stadyed at home and in kindergarten were the dominant, especially scattered children (between65.5%and81.0%of the cases were reported from2008to2013).6. The highest incidence of HFMD was distributed in Yuhua district. The annual incidences were between202.11/100000to528.28/100000from2008to2013.7. From lab confirmed cases, the predominant pathogen from2010to2012in Changsha was EV71, in2013was other enterovirus. There were113severe cases,56.5%of them had serological evidence of EV71. There were25death cases,92.0%of them had serological evidence of EV71.8. The monthly incidence of HFMD was positive correlated with current monthly maximum temperature, mean temperature, minmum temperatre, cumulative rainfall and mean vapour pressure respectively, was negative correlated with current monthly mean air pressure and mean wind speed respectively, was positive correlated with maximum temperature and cumulative rainfall of previous month respectively, was negative correlated with mean air pressure of previous month. There were no significant relation between monthly incidence and cumulative sunshine hours and mean relative humidity.9. SARIMA (1,0,0)(0,1,1)12model was the best model to forecast the HFMD incidence. The prediction model of HFMD epidemic in Changsha is (1-0.701B)(1-B12)Zt=(1-0.372B12)μt. Where B was backward shift operator, Z, was arithmetic square root of monthly incidence of HFMD,μ was random disturbance.10. This model fitted the incidence of the previous monthly incidence from January2008to December2012, and the predicted monthly incidence in2013by the model was consistent with the actual incidence. The predicted annual incidence of HFMD in2013was335.59/100000, the actual annual incidence of HFMD in2013was276.16/100000, and the relative error was21.52%.CONCLUSIONS:The epidemic situation of HFMD in Changsha is relatively serious, and more attention should be payed during April to July and October to December. Scatter children and preschool children younger than6years old are the focus of prevention and control. Yuhua district are the main surveillance area. The progressive disease of cases of EV71infection should be paid more attention. Moreover, time series model can be used to fit the changing trends of incidence of HMFD at different time. It can be used to forecast the incidence of HMFD in the future, and to provide service for prevention and control of HMFD.