Epidemiological Study And Prediction Models Of Food Poisoning In Hunan Province

Objectives:To describe the incidence, mortality, trends, analyze the epidemiological characteristics of food poisoning and explore the application of gay model(1,1), curve fitting, exponential smoothing, ARMA(auto-regressive moving average) model in the number of food poisoning prediction in Hunan Province, providing scientific basis for the prevention and control of food poisoning in Hunan Province.Methods:Collect the data of food poisoning and population from2000to2011, and use the descriptive analysis, circular distribution methods and the trend-test to describe the current situation and epidemiological characteristics of food poisoning in Hunan Province. Collect the number of food poisoning from2000to2010in Hunan Province for gay model(1,1), curve fitting, exponential smoothing, ARMA(auto-regressive moving average) model fitting, and use the number of food poisoning in2011to verify the effect of model prediction, then predict the number of food poisoning in2012. At last select some major food poisoning cases to analyze the characteristics, causes and results of food poisoning, and make recommendation for food poisoning.Results:1. The incidence, mortality and trends of food poisoning in Hunan Province:according to the study, the times of poisoning, incidence and mortality of poisoning in Hunan from2000to2011was41,1.668per100000and0.012per100000every year, respectively. The trend test indicated that the incidence, mortality, fatality rate of poisoning decreased significantly from2000to2011(U=-2.331, P<0.05; U=-3.429, P<0.05; U=-2.331,P<0.05, respectively).2. Regional distribution character:according to the study, the population of poisoning from Xiangxi Autonomous Prefecture was largest (0.136per100000population), when Chang de City was the least(0.033per100000population). And the incidence of poisoning from Zhang Jiajie City was the largest (6.173per100000population) when Chang de City was the least(1.057per100000population)3. Time distribution character:The circular distribution analysis showed that there was a central tendency of the distribution of food poisoning in Hunan Province, and the peak month of food poisoning was from May to September. In2000-2011the average peak time was at August6th and the average peak time zone was from May14th to October4th..4. Poisoning places and responsibilities of unit distribution character: Home was the main food poisoning place and responsible units in Hunan, and followed by is canteens. The fatality rate of food poisoning happened at home is the highest, and as to other places, responsible unit was the highest.5. Poisoning food, risk factors, cause of poisoning distribution character:The microbial food poisoning was the main food poisoning in Hunan Province, there were mainly fruits and vegetables, meat and meat products, cereal and cereal products and the leading cause of poisoning was misusing of toxic species.6. Prediction of the number of food poisoning results showed that the simple exponential smoothing method better fits the number of food poisoning from2000to2010, and three function prediction model of curve fitting better predict the number of food poisoning in2011.Conclusions:1. According to the study, the times of poisoning, incidence and mortality of poisoning in Hunan from2000to2011were41,1.668per100000population and0.012per100000population every year, respectively. And the incidence, mortality, fatality rate of poisoning decreased significantly from2000to2011.2. Xiangxi Autonomous Prefecture, Zhang Jiajie were the areas that had higher food poisoning in Hunan Province. It mainly happened in summer and autumn, and its main places are rural families, microbial food poisoning account for the major part.3. Prediction of the number of food poisoning results showed curve fitting, exponential smoothing, ARMA (auto-regressive moving average) model can be used in the prediction of the number of food poisoning. The simple exponential smoothing method better fits the number of food poisoning from2000to2010, and three function prediction model of curve fitting better predict the number of food poisoning in2011.if used for long-term forecasts, the appropriate forecasting model should be chosen according to the actual situation and the characteristics of long-term test data.