Customizing and fine-tuning opioid overdose machine learning models for community-specific dashboards
5 affiliations
- 1Institute for Biomedical Informatics, College of Medicine, University of Kentucky, Lexington, 40536 KY United States of America
- 2Department of Biostatistics, College of Public Health, University of Kentucky, Lexington, 40536 KY United States of America
- 3Dr. Bing Zhang Department of Statistics, College of Arts and Sciences, University of Kentucky, Lexington, 40536 KY United States of America
- 4Departments of Behavioral Science and Psychiatry, College of Medicine, University of Kentucky, Lexington, 40536 KY United States of America
- 5Department of AI and Health, Medical College of Georgia at Augusta University, Augusta, 30912 GA USA
Abstract
Drug overdose is a leading cause of death in the United States. Targeted community interventions and allocation of resources for reducing overdose deaths require timely and actionable data to make a meaningful public health impact. We present our work on tuning geographic-specific machine learning forecast models to predict the number of opioid-involved overdoses involving Emergency Medical Services (EMS) in Kentucky, with a focus on optimizing covariates representing measurable opioid overdose risk and protective factors. We aggregated suspected opioid-involved overdoses in EMS data into monthly totals by county between January 2017 and September 2025. We further grouped these counties into larger regions, representing Area Development Districts (ADDs), service areas for Community Mental Health Centers (CMHCs), and local health departments (LHDs). We constructed long-term time-series forecasting models (NLinear LTSF) to predict opioid-involved overdoses three months into the future. We tested the impact of using eight covariate data sources: Kentucky State Police drug seizures, Medicaid claims, community naloxone distribution, Kentucky Department of Corrections intakes and releases, social determinants of health data, location of treatment venues, and weather. We evaluated our models using both an unseen 10% test set and an expanding window strategy, where we simulated predicting historical data and retrained the model using data available for every three-month window. We embedded predictions into community-specific dashboards where end-users can select individual counties, or larger regions covered by distinct ADDs, CMHCs, or LHDs. The dashboards are integrated in Kentucky’s Rapid Actionable Data for Opioid Response system designed to deliver ongoing near-real time analytics and forecasting at state and local levels.
Supplementary Information: The online version contains supplementary material available at https://doi.org/10.1186/s40537-026-01507-4.
Related
- Knowledge, attitudes and practices among Australian community pharmacists towards naloxone and overdose prevention: a national cross-sectional study
- Application of Artificial Intelligence for Predicting Sports Injuries and Customizing Personalized Prevention Strategies: A Scoping Review
- Respiratory virology dashboards for primary care and public health insight in England
- From data to action: high-resolution spatial dashboards for equity-oriented public health in Hungary
- Using GIS Dashboards to highlight AMR data disparities in Africa for Policy, Research, and Public Health
BibTeX
@article{harris2026customizing,
title = {{Customizing and fine-tuning opioid overdose machine learning models for community-specific dashboards}},
author = {Daniel R. Harris and Aaron Mullen and Peter Rock and Carmen Canedo and Katherine Thompson and Michelle R. Lofwall and Cody Bumgardner and Jeffery Talbert and Svetla Slavova},
journal = {Journal of Big Data},
year = {2026},
month = jul,
doi = {10.1186/s40537-026-01507-4},
url = {https://doi.org/10.1186/s40537-026-01507-4},
note = {CC BY-NC-ND 4.0},
}
RIS
TY - JOUR
TI - Customizing and fine-tuning opioid overdose machine learning models for community-specific dashboards
AU - Daniel R. Harris
AU - Aaron Mullen
AU - Peter Rock
AU - Carmen Canedo
AU - Katherine Thompson
AU - Michelle R. Lofwall
AU - Cody Bumgardner
AU - Jeffery Talbert
AU - Svetla Slavova
PY - 2026
DA - 2026/07/30
JO - Journal of Big Data
DO - 10.1186/s40537-026-01507-4
UR - https://doi.org/10.1186/s40537-026-01507-4
AB - Drug overdose is a leading cause of death in the United States. Targeted community interventions and allocation of resources for reducing overdose deaths require timely and actionable data to make a meaningful public health impact. We present our work on tuning geographic-specific machine learning forecast models to predict the number of opioid-involved overdoses involving Emergency Medical Services (EMS) in Kentucky, with a focus on optimizing covariates representing measurable opioid overdose risk and protective factors. We aggregated suspected opioid-involved overdoses in EMS data into monthly totals by county between January 2017 and September 2025. We further grouped these counties into larger regions, representing Area Development Districts (ADDs), service areas for Community Mental Health Centers (CMHCs), and local health departments (LHDs). We constructed long-term time-series forecasting models (NLinear LTSF) to predict opioid-involved overdoses three months into the future. We tested the impact of using eight covariate data sources: Kentucky State Police drug seizures, Medicaid claims, community naloxone distribution, Kentucky Department of Corrections intakes and releases, social determinants of health data, location of treatment venues, and weather. We evaluated our models using both an unseen 10% test set and an expanding window strategy, where we simulated predicting historical data and retrained the model using data available for every three-month window. We embedded predictions into community-specific dashboards where end-users can select individual counties, or larger regions covered by distinct ADDs, CMHCs, or LHDs. The dashboards are integrated in Kentucky’s Rapid Actionable Data for Opioid Response system designed to deliver ongoing near-real time analytics and forecasting at state and local levels. Supplementary Information: The online version contains supplementary material available at https://doi.org/10.1186/s40537-026-01507-4.
SN - 2196-1115
ER -