Assessing Uncertainty in Spatial Exposure Models for Air Pollution Health Effects Assessment.

Background: Although numerous epidemiologic studies now use models of intra-urban exposure, there has been little systematic evaluation of the performance of different models. Objectives: In this paper we proposed a modeling framework for assessing exposure model performance and the role of spatial autocorrelation in health effects estimation. Methods: Data was obtained from an exposure measurement substudy of subjects from the Southern California Children's Health Study. We examine how the addition of spatial correlations to a previously described unified exposure and health outcome modeling framework affects estimates of exposure-response relationships using the substudy data. The methods proposed build upon the authors' previous work, which developed measurement-error techniques to estimate long-term nitrogen dioxide (NO₂) exposure and its effect on lung function in children. This paper further develops these methods by introducing between and within-community spatial autocorrelation error terms to evaluate effects of air pollution on forced vital capacity (FVC). The analytical methods developed are set in a Bayesian framework where multi-stage models are fitted jointly, properly incorporating parameter estimation uncertainty at all levels of the modeling process. Results: Findings suggest that the inclusion of residual spatial error terms improves the prediction of health effects. The results also demonstrate how residual spatial error may be used as a diagnostic for comparing exposure model performance. [ABSTRACT FROM AUTHOR]