Bayesian Superpopulation Models for Population-Based Diabetes Surveillance Using Multi-Site Electronic Health Record Data
Abstract
Electronic health records (EHRs) are increasingly used for disease surveillance. However, data collected in EHRs may not be a representative sample of the target population for inference. Methods for EHR-based surveillance remain in their nascence. In addition to simple weighting approaches, researchers have used multilevel regression and poststratification (MRP), a popular technique for finite population quantity estimation that remains understudied in EHRs. We are motivated by a US-based collaborative to modernize diabetes surveillance using EHR data from different health system centers across the country. To estimate type 1 diabetes (T1D) prevalence in children and adolescents, we propose and evaluate three Bayesian superpopulation models that are distinguished by how they account for two important features of EHR data: unobserved heterogeneity and data sparsity. The three models include an analysis of variance-inspired approach; a model with geographic effects varying by demographic subgroup; and MRP, which we consider as the reference model. We apply the models to estimate prevalence overall and in demographic- and geographic-based subgroups. Using plasmode simulations, we evaluate model estimation performance while preserving the complexity of the EHR data. Our results show that the superpopulation-based prevalence estimates were higher than those observed in the EHR sample alone, with the magnitude of the differences varying by health system center. The bias and coverage probabilities of MRP markedly worsened when the EHR sample size was smaller or the selection mechanism was more complex. Overall, the superpopulation model based on analysis of variance demonstrated the best estimation performance.
Related
- Bayesian Spatio-temporal Modeling of Multi-Site Wastewater Surveillance Data
- Counting the Uncounted: Population-Level Surveillance of Documented Pregnancy and Fetal Harm in Police Crash Narratives with a System One Model (Jev)
- Frequentist and Bayesian population pharmacometrics: similarities, differences, and practical implications
- Ensemble-based support vector machine classifiers as an efficient tool for quality assessment of beef fillets from electronic nose data
- Score-Based Generative Data Assimilation for Integrating Aggregated Surveillance Data into Agent-Based Models in Epidemic Tracking
BibTeX
@misc{anthopolos2026bayesian,
title = {{Bayesian Superpopulation Models for Population-Based Diabetes Surveillance Using Multi-Site Electronic Health Record Data}},
author = {Rebecca Anthopolos and Bo Cai and Lorna E. Thorpe and Hui Xie and Xiaochun Li and Sarah Conderino and David C. Lee and Hui Shao and Ibrahim Zaganjor and Marc Rosenman and H. Lester Kirchner and Muchiri E. Wandai and Levon Utidjian and Tessa Crume and Kristi Reynolds and Jasmin Divers and DiCAYA Study Group},
howpublished = {arXiv},
year = {2026},
month = nov,
doi = {10.48550/arxiv.2611.05661},
eprint = {2611.05661},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2611.05661},
}
RIS
TY - UNPB
TI - Bayesian Superpopulation Models for Population-Based Diabetes Surveillance Using Multi-Site Electronic Health Record Data
AU - Rebecca Anthopolos
AU - Bo Cai
AU - Lorna E. Thorpe
AU - Hui Xie
AU - Xiaochun Li
AU - Sarah Conderino
AU - David C. Lee
AU - Hui Shao
AU - Ibrahim Zaganjor
AU - Marc Rosenman
AU - H. Lester Kirchner
AU - Muchiri E. Wandai
AU - Levon Utidjian
AU - Tessa Crume
AU - Kristi Reynolds
AU - Jasmin Divers
AU - DiCAYA Study Group
PY - 2026
DA - 2026/11/03
PB - arXiv
DO - 10.48550/arxiv.2611.05661
UR - https://arxiv.org/abs/2611.05661
AB - Electronic health records (EHRs) are increasingly used for disease surveillance. However, data collected in EHRs may not be a representative sample of the target population for inference. Methods for EHR-based surveillance remain in their nascence. In addition to simple weighting approaches, researchers have used multilevel regression and poststratification (MRP), a popular technique for finite population quantity estimation that remains understudied in EHRs. We are motivated by a US-based collaborative to modernize diabetes surveillance using EHR data from different health system centers across the country. To estimate type 1 diabetes (T1D) prevalence in children and adolescents, we propose and evaluate three Bayesian superpopulation models that are distinguished by how they account for two important features of EHR data: unobserved heterogeneity and data sparsity. The three models include an analysis of variance-inspired approach; a model with geographic effects varying by demographic subgroup; and MRP, which we consider as the reference model. We apply the models to estimate prevalence overall and in demographic- and geographic-based subgroups. Using plasmode simulations, we evaluate model estimation performance while preserving the complexity of the EHR data. Our results show that the superpopulation-based prevalence estimates were higher than those observed in the EHR sample alone, with the magnitude of the differences varying by health system center. The bias and coverage probabilities of MRP markedly worsened when the EHR sample size was smaller or the selection mechanism was more complex. Overall, the superpopulation model based on analysis of variance demonstrated the best estimation performance.
ER -