Back

Evaluating linkage approaches for address-level socioenvironmental exposure assessment

Hartlage, C. S.; Manning, E. R.; Brokamp, C.

2025-12-04 health informatics
10.64898/2025.12.02.25341378 medRxiv
Show abstract

Accurate linkage of addresses to parcel-level data is essential for hyperlocal environmental exposure assessment, yet the performance of methods, including their impact on exposure misclassification and bias, remains poorly characterized. Using a gold standard match of 853,255 National Address Database records to authoritative datasets from Hamilton and Franklin Counties, Ohio, we evaluated address tag fuzzy matching and geocoding-based (geomatching) approaches on accuracy of linked parcel identifier and parcel market total value and usage type. Address tag fuzzy matching achieved 100% agreement; address point geomatching performed moderately well (65.1% - 76.1%), and street range geomatching performed poorly (7.2% - 59.2%). Poorer agreement was more common in neighborhoods with higher address densities and more community material deprivation, highlighting potential for differential misclassification of exposure assessment. These findings emphasize a need for precise, scalable, and standardized linkage approaches to support valid address- and parcel-level exposure assessment in clinical and population health research.

Matching journals

The top 7 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.