Analysis And Prediction Of The Epidemic Situation Of Schistosomiasis In Qianjiang City

Objective To investigate the epidemic situation of schistosomiasis in Qianjiang city in recent years, so as to provide evidences for the future control strategies. To verify the feasibility of forecasting the infection of schistosomiasis in short term by using autoregressive integrated moving average (ARIMA) model, a method of time series analysis, so as to provide methodological basis on early warning and forecasting the incidences of schistosomiasis.Methods The investigative data of the infection of Schistosomiasis Japonica in residents and cattle and the infection of cercaria in oncomelania in four monitoring sites of Qianjiang city were collected and analyzed to describe the epidemic situation of schistosomiasis. Data of the infection incidence of schistosomiasis in Qianjiang city from 1959 to 2005 was used to set up ARIMA model, of which, parameters were estimated through maximum likelihood method and the order was confirmed through Akaike Information Criterion (AIC) and Beyes Information Criterion (BIC) and the precision was determined by the average relative error. The optimal constructed model was used to predict the incidences of schistosomiasis from 2009 to 2013 in Qianjiang city.Results There were 60.63 million people living in the endemic area of schistosomiasis in Qianjiang city in 2008. From 2006 to 2008, the incidence of infection in residents which was higher in male and older age group (40-60 years) comparing with female and the other age groups decreased year by year (3.35%, 2.97% and 1.77%, respectively, P<0.01) in four surveillance sites for schistosomiasis in Qianjiang city, where there were not new acute infected cases and advanced schistosomiasis cases during these three years and health education activities especially distributing and using personal protective articles decreased year by year. The incidence of infection in cattle also decreased year by year (10.75%, 6.83% and 5.18%, respectively, P<0.05) and the infection rates of cercaria in oncomelania fluctuated from 0.21% to 0.39% during this time. ARIMA((4),1,0) model can appropriately fit the time series trend of the schistosomiasis year incidence in Qianjiang city. The AIC was 219.849 and the optimal constructed model was x_t-x_t-1 =1/(1 +0.331505B⁴)ε_t. The average relative error between the predictive rates and actual rates was 7.37%. The predictive incidences of schistosomiasis from 2009 to 2013 were 2.11%, 2.76%, 3.22%, 3.44% and 3.33%.Conclusion The current epidemic situation of schistosomiasis in Qianjiang city had been controlled effectively, but the number of infected residents was still large. However, there are repeated infection in human and cattle and the situation of snail is unstable. ARIMA model can be used for dynamic analysis and short-term forecasting of schistosomiasis infection. The schistosomiasis incidence may rise in a few years showed in the reconstructed model indicated the necessary for the relevant department to take effective measures to prevent epidemic situation of schistosomiasis promptly.