Evaluating Linkage Approaches for Address-Level Socioenvironmental Exposure Assessment

Area of Science:

• Environmental health
• Geospatial analysis
• Public health research

Background:

• Accurate address-to-parcel linkage is vital for hyperlocal environmental exposure assessment.
• Current linkage methods' performance and impact on exposure misclassification are poorly understood.
• This study addresses the need for robust linkage strategies in health research.

Purpose of the Study:

• To evaluate the accuracy of address tag fuzzy matching and geocoding (geomatching) for linking addresses to parcel data.
• To assess the impact of linkage accuracy on parcel characteristics like value and usage type.
• To identify factors influencing linkage performance and potential for differential exposure misclassification.

Main Methods:

• Utilized a gold standard match of 853,255 National Address Database records.
• Compared address tag fuzzy matching against address point and street range geomatching.
• Evaluated linkage accuracy based on parcel identifier, market total value, and usage type.

Main Results:

• Address tag fuzzy matching achieved 100% agreement.
• Address point geomatching showed moderate performance (65.1%–76.1%).
• Street range geomatching performed poorly (7.2%–59.2%), with higher error rates in dense, deprived neighborhoods.

Conclusions:

• Address tag fuzzy matching is a highly accurate method for address-to-parcel linkage.
• Geocoding methods exhibit variable performance, potentially leading to biased exposure assessment.
• Standardized, precise linkage approaches are essential for valid environmental exposure assessment in health studies.