Extracting knowledge from data: Combining environmental measurements and field observations in statistical models of infectious disease ecology.

摘要

The public health burden of infectious diseases worldwide remains a high priority for intervention; the ability to understand when and where transmission will or will not occur can improve the efficiency and efficacy of these interventions. A useful tool for increasing our understanding of these issues is statistical modeling of disease processes. The number of factors involved in disease transmission, however, make model building a significant challenge. The difficulty increases substantially when the cycle includes more than just humans and the pathogen; nonhuman reservoirs or vectors are involved in many disease transmission pathways. One complication can be a lack of suitable population data which is critical to estimating the local prevalence of a pathogen in a reservoir population. Here a model of Black Creek Canal virus exposure in Sigmodon hispidus is developed that incorporates factors at multiple scales in order to ensure that all relevant information is taken into consideration. Estimation of the parameters and predicted values for this model was done using empirical Bayesian Markov Chain Monte Carlo (MCMC) methods. Another challenge to understanding infectious disease transmission when nonhuman species are involved is the influence of climate. The connection between meteorology and vector populations is exploited here to develop models of mosquito population dynamics. The model building process is used to identify the key weather conditions from the off-season that impact several mosquito species during their overwintering period. These models are also demonstrated to be efficacious for predicting monthly and annual totals of mosquito populations. Overall, the models presented here provide several insights into the ecology of specific disease systems and suggest possible directions for handling the complexity of understanding infectious disease ecology in general.