| Although brilliant achievements have been made in the prevention and control of schistosomiasis japonica in People’s Republic of China after great efforts for more than half a century, there is still a public health problem of schistosomiasis in the lake and marshland region of P.R. China. Schistosomiasis control in the lake and marshland region is still a major and difficult task in the national schistosomiasis control program in P.R. China. The transmission process of schistosomiasis japonica includes infection sources, intermediate host and definitive hosts, and are impacted by many factors, such as natural environmental, cultural and socio-economic factors. Therefore, the key issue to effectively control schistosomiasis transmission is to understand comprehensively the relationship and function among elements of humans, cattle and snail in the transmission process of schistosomiasis japonica, and to understand the complex relationship among those impact factors. This investigation aims at (1) to analyze and clarify the interrelated roles among humans, cattle and snail in the transmission process of schistosomiasis japonica,(2) to explore the relationship between impact factors of anthropogenic and socio-economic factors and human infection/re-infection with Schistosoma japonicum in the lake and marshland region,(3) to identify main impact factors thus to provide more scientific information for further research on the transmission dynamics of schistosomiasis, and for formulation of control measures of schistosomiasis.Firstly, data of40monitoring sites in the lake and marshland region where the transmission of schistosomiasis was still undergoing was extracted from80monitoring sites in the national surveillance system of schistosomiasis from2005to2010. The data from40monitoring sites shows that the overall trend of human infection rate with S. japonicum was decreasing in each site, and so was the cattle infection rate with only a slight fluctuation in few sites. Much more fluctuation in the snail infection rate was observed. The infection rates in humans, cattle and snail were employed to develop back propagation (BP) neural networks, in which the input variables included the infection rates of cattle and snail, and the output variable was the human infection rate. The relationship among infection rates of humans, cattle and snail was simulated and analyzed on the basis of the optimal network selected based on the evaluation indicators that were MSE=0.032724, Gradient=0.035683, RTesting=0.54955. The results indicated that the higher the infection rates in cattle and snails were, the higher the human infection rate was; with the rise of the infection rates of cattle and snail, the human infection rate increased gradually and progressively approached to a specific value, and the higher the infection rate of cattle, the shorter the first platform of human infection rate with the variation of snail infection rate, but the longer the second platform, which illustrated that the relationship among infection rates of human, cattle and snail were approximately presented as "S" shape,.cattle was still playing an important role in schistosomiasis transmission. So, it’s concluded that back propagation artificial neural networks can be applied to the study on the transmission dynamics of schistosomiasis.Secondly, investigation on transmission impact factor was undertaken in Jiangling county, Hubei province. The cluster stratification sampling was adopted to sample villagers aged from6to65years in6villages to carry out the household survey. A total of2339villagers from1247households were investigated. The data of household survey were collected as follows: basic information of respondents, family health and economic status, history of infection with S. japonicum, history of contact with infested water. Based on above mentioned survey data, logistic regression model and BP artificial neural network were built. Results of those two models were compared and showed that best fitness for the logistic regression model was the model with R2=0.5984, and other indicators were AIC=1793.805, SC=1897.440,-2LogL=1757.805. Results indicated that the main risk factors were history of infection with S. japonicum, history of contact with infested water, the extent of which varied with village, age, education background, family economic status. The evaluation indexes of the optimal BP artificial neural network were as follows:MSE=0.0734, Gradient=0.0019082, Validation checks=0and R=0.65361. According to the size of the absolute value of MIV, the first16factors were in turn as follows:the interaction between history of infection with S. japonicum and history of contact with infested water, the family with post infection with S. japonicum, history of contact with infested water, history of infection with S. japonicum, times of infection with S. japonicum, age, type of the latest infection, education, drugs, local prevalence level, courtyard, village, family latrines, lifestyles of contact with infested water, family economic status and entertainment of contact with infested water (Table2.14). Results from both of logistic regression model and BP neural networks indicated that the high risk factors of schistosomiasis transmission were human behavior, and children under age of15were still higher risk population for schistosome infection, along with the fact that the better the family economic conditions were and/or the higher the education level was, the lower the human infection rate were. So, the control measures of schistosomiasis need to be formulated tailored to local settings which varied from time to time, and from place to place.All in all, the relationship among infection rates of humans, cattle and snail presented a biologically nonlinear ("S" type) pattern. The cattle is still an important infection source of schistosome infections, so that control of cattle infection is a major measure to achieve the goal of schistosomiasis elimination, which is not only able to cut the transmission cycle of the snails, livestock, humans, but also able to protect the natural environment. Human behavior is one of the main risk factors for schistosomiasis transmission in the lake and marshland region. With sustained implementation of control measures, the schistosomiasis endemicity will be at a relative lower level, which may create a platform stage for the control effort. Therefore, strengthening health education to change human behavior, especially the population aged less than15, is one of the major tasks to reduce the probability of the human infection, which is also one of the important control measures for sustained keeping the schistosomiasis prevalence at a low level. |
PDF Full Text Request