Using Monte Carlo/Gaussian Based Small Area Estimates to Predict Where Medicaid Patients Reside

Area of Science:

• Health Informatics
• Population Health Management
• Geospatial Analysis

Background:

• Electronic Health Records (EHR) are increasingly used for population health initiatives.
• Medicaid expansion discussions highlight the need for accurate health data.
• Effective small area estimation techniques are crucial for leveraging EHR data.

Purpose of the Study:

• To assess the accuracy of a Monte Carlo/Gaussian technique for small area estimation of Medicaid patients.
• To develop a predictive model for Medicaid patient distribution at the US Block Group level.
• To create a tool translating EHR data potential for population health studies.

Main Methods:

• Utilized a Monte Carlo/Gaussian technique, previously successful for voter registration data.
• Assessed methodology accuracy using Albuquerque patient address data.
• Combined Albuquerque and Chicago EHR data to build a predictive regression model.

Main Results:

• The Monte Carlo/Gaussian technique demonstrated efficacy in estimating small area populations.
• A regression model was developed to predict Medicaid patients by US Block Group.
• The study successfully laid the groundwork for a tool to translate EHR data for population health.

Conclusions:

• The developed methodology shows promise for accurate small area estimation of Medicaid populations.
• EHR data, when analyzed with advanced techniques, can significantly contribute to population health planning.
• This research supports the use of EHR data in policy discussions, including Medicaid expansion.