In environmental health researches and practices, spatial analysis became an important approach to estimation of environmental exposure of human subjects under concern. A typical situation in this kind of application is that the data of pollution are available only at certain locations, and thus, inference is needed to convert a limited number of values at discrete locations into a continuous surface. This paper intends to clarify the distinction among three methods that can be used to achieve this conversion, namely interpolation, kernel density estimation and snapshotting. Due to the apparent similarity of the three, they may cause confusions that lead to misuses. We compare and contrast the three methods, in terms of nature of the input data, mathematical process of the inference, and essential meaning of the output. For each method, we suggest appropriate applications within the context of estimation of environmental exposure.

Traditional epidemic models of communicable diseases focus on the temporal dimension. We propose a framework for the epidemic forest approach to spatializing epidemic modeling. An epidemic forest is formed by epidemic trees, each using the tree structure to represent an epidemic starting from a primary case. Each node on the tree represents an individual case, and each link represents a parent–child transmission linkage and can be modeled with the spatiotemporal and other information about the cases. Structural, spatiotemporal, and epidemiological information can be extracted from constructed trees to characterize the epidemic they represent and to correspond to environmental data. When multiple primary cases can be determined, the forest is ready to build. We applied this method to the 2013 dengue fever epidemic in Guangzhou City, China. With the constructed forest, we particularly calculated the case reproduction ratio (R_t), an index widely used for characterizing epidemics, at the global, tree-wise, and pixel-wise scales. We further calculated correlation coefficients between these R_ts and climate variables. R_ts at different scales, as well as their associations with climate variables, offer information at different levels that is all important in epidemiological studies and disease control practices. Through this study, we explored and demonstrated how to spatialize epidemic modeling, what information can be extracted from this spatialization, and then how to use the extracted information. We also point out that spatialization of R_t is the essential process of mapping a communicable disease, corresponding to the spatialization of incidence or prevalence in mapping a chronic disease.

The spatial scan statistic method has been widely used for detecting disease clusters. Its results may be affected by scales, including the aggregation level of the input data and the population threshold used in the detection. Previous studies offered inconsistent findings, and few had considered both types of scales at the same time. Using 24 simulated datasets and two real disease datasets, we investigated the method’s sensitivity to the two types of scales. We aggregated the individual-level data into areal units of three levels, including county, town, and a 900 m grid. We detected clusters with three population thresholds, including 10%, 25%, and 50%. We used two measurements, distance between cluster centres and the Jaccard index, to quantify the consistency of clusters detected with different scale settings. We find: (1) the method is not greatly sensitive to the data aggregation level when the cluster is strong and in a place with high population density; (2) the method’s sensitivity to the population threshold is determined by the actual size of the true cluster; and (3) a regular grid with fine resolution is advantageous over the subjectively defined areal units. The process and findings may have broader meanings to similar spatial analyses.

Trichloroethylene (TCE) is a well-documented kidney carcinogen based on a substantial body of evidence including mechanistic and animal studies, as well as reports from occupational settings. However, the cancer risks for those in residential exposures such as TCE contamination in groundwater are much less clear. The objective of this study was to perform a detailed spatio-temporal analysis of estimated residential TCE exposure in New Hampshire, US. We identified kidney cancer cases (n = 292) and age-, gender-matched controls (n = 448) from the Dartmouth-Hitchcock Health System and queried a commercial financial database for address histories. We used publically available data on TCE levels in groundwater measured at contaminated sites in New Hampshire and then modeled the spatial dispersion and temporal decay. We overlaid geospatial residential locations of cases and controls with yearly maps of estimated TCE levels to estimate median exposures over the 5, 10, and 15-year epochs before diagnosis. The 50th–75th percentile of estimated residential exposure over a 15-year period was associated with increased kidney cancer risk (adjusted Odds Ratio (OR) 1.78 95% CI 1.05–3.03), compared to <50th percentile. This finding supports the need for groundwater monitoring of TCE contaminated sites to identify potential public health risks.