Small area environmental health data: Challenges and solutions from a surveillance system perspective

Abstract

Timely, locally-relevant information is important for conducting environmental epidemiology studies; however, public health data are often lagging, spatially unresolved, or unavailable. The Centers for Disease Control and Prevention’s Environmental Public Health Tracking Program (Tracking Program) is building a system of sub-county data to enhance the spatial resolution. This presentation covers the main challenges of increasing spatial resolution. The validity of epidemiologic research depends on positional accuracy of geocodes. Geocoding presents several challenges (e.g., handling special cases like PO Boxes and group quarters, match score thresholds, and how to categorize geocoding quality). The Tracking Program developed geocoding guidelines to transform address-level health data to census tract while addressing these issues. Many health outcomes cannot be displayed at census tract for reasons of confidentiality and stability due to too few cases and/or people in an area. To overcome this, the Tracking Program created standardized sub-county geographies (census tract aggregations) that can be compared over time, space, and dataset to allow data to be displayed at a finer scale while ensuring minimal suppression and instability. Multiple challenges were addressed, including selection of population thresholds and handling group quarters. There are multiple choices of denominator datasets available for public health surveillance (e.g., decennial census). The choice of denominator dataset may or may not have an impact on calculated rates and public health interpretation. To address this and answer the question of which denominator dataset to use, the Tracking Program created a framework to evaluate denominator datasets. The efforts from this project will help meet increasing demand for finer resolution data. These data can inform environmental health studies, where there is often local-level heterogeneity in the environment. Having small area environmental health data can allow for better understanding of local variation within counties, prioritize needs, and advance our understanding of environmental health processes and impacts.