POINT: Pipeline for Offline Conversion and Integration of Geocodes and Neighborhood Data.

Area of Science:

• Health Informatics
• Geographic Information Systems (GIS)
• Public Health Research

Background:

• Geocoding is crucial for linking patient addresses to neighborhood social determinants of health (SDOH).
• Existing geocoding methods may not be suitable for handling sensitive patient data due to privacy concerns.
• There is a need for efficient, accurate, and HIPAA-compliant geocoding solutions in healthcare.

Purpose of the Study:

• To develop and evaluate POINT, an integrated offline geocoding pipeline.
• To enable researchers and health systems to obtain neighborhood-level SDOH estimates.
• To support personalized care and interventions based on patient lingkungan.

Main Methods:

• Developed POINT, a web-based, containerized, offline geocoding application with a GUI and API.
• Utilized two datasets for evaluation: 1 million Open Addresses and 3,096 geocoded patient addresses.
• Assessed geocoding success rates, census tract concordance, and coordinate accuracy against reference data.

Main Results:

• Achieved high geocoding success rates: 99.4% for Open Addresses and 99.8% for patient addresses.
• Demonstrated high census tract assignment concordance (>90%) with reference data for both datasets.
• Reported median geocoding distances of 52.5m and 14.5m for the respective datasets, indicating good accuracy.

Conclusions:

• POINT offers a scalable, HIPAA-compliant solution for geocoding sensitive patient addresses offline.
• The application provides accuracy comparable to existing geocoders while protecting patient privacy.
• POINT facilitates the integration of neighborhood-level variables into health research and clinical practice to advance health equity.