A new analytical system created by University of Kentucky researchers is helping to predict and prevent opioid overdoses in Kentucky.
The surveillance system, called Rapid Actionable Data for Opioid Response in Kentucky (RADOR-KY), is used to inform overdose prevention and response efforts throughout the state by giving health departments, community organizations and policymakers access to near-real-time opioid overdose data for every county in Kentucky.
Developed by a UK research team led by Svetla Slavova, Ph.D., and Jeffery Talbert, Ph.D., work on the RADOR-KY dashboard began in 2023 with a three-year, $3.1 million grant from the National Institute on Drug Abuse.
The dashboard, now available at rador-ky.uky.edu, was presented this week at the Rx and Illicit Drug Summit by RADOR-KY team members Sharon Walsh, Ph.D., Daniel Harris, Ph.D., Lindsey Hammerslag, Ph.D., and Trish Freeman, Ph.D.
Overdose deaths in Kentucky peaked in 2021 at 2,257, coinciding with the COVID-19 pandemic and a surge in illicit drugs including fentanyl and methamphetamine. Opioids were involved in 90% of those deaths. Deaths have fallen sharply since, dropping to 1,410 in 2024, but Kentucky's drug overdose death rate remains above the national average.
Responding to the opioid crisis effectively requires timely, accurate information from many different sources. RADOR-KY brings that data together in one place, helping communities implement evidence-based practices like distributing naloxone, a medication that reverses opioid overdoses, and expanding access to medications for opioid use disorder such as buprenorphine.
"RADOR-KY is an analytical system that serves as a proof of concept for statewide health intelligence," said Slavova, professor and associate dean for research in the UK College of Public Health. "It bridges fragmented state health data into a shared analytics environment, making it a scalable framework that can be adopted in other health domains or by other states and jurisdictions."
The dashboard covers everything from overdose deaths and prescribing patterns to treatment access and naloxone distribution, drawing on data from more than a dozen sources including emergency medical services, the Kentucky All Schedule Prescription Electronic Reporting (KASPER) system, Medicaid claims and death certificate records.
RADOR-KY draws on the work of more than 20 UK researchers from UK's College of Public Health, College of Medicine and College of Pharmacy. The project also builds on UK's work on the HEALing Communities Study (HCS).
"This new tool, developed in partnership with our state and community partners, is publicly available and free to support public health planning for targeted interventions," said HCS principal investigator Sharon Walsh, Ph.D., a professor in UK's College of Medicine and College of Pharmacy and director for the Center on Drug and Alcohol Research. "Additionally, many community organizations can leverage this new resource as they prepare grants to support their own work."
The dashboard has a simplified default view that lets users select a measure, such as overdose deaths or naloxone distribution, and see results mapped by county alongside trend lines over time. An advanced version allows users to customize the data, download graphs and datasets, and generate predictions using machine learning models or identify emerging overdose "hotspots."
The team's Vision Session at the Rx Summit in Nashville on April 7 included a live dashboard demonstration and a real-world scenario using the Kentucky River District, a seven-county region in Eastern Kentucky, showing how a local health department could use RADOR-KY to identify overdose hotspots, assess treatment access gaps, flag emerging drug trends and build a targeted action plan.
The team sees RADOR-KY as a model for other states looking to build a similar public health data infrastructure. Built using open-source tools and a common data model, the system is designed to be easily updated, adapted and replicated.
"Many machine learning models in public health are built for a single study and rarely see continued use, but RADOR-KY is different," said Harris, associate professor in the College of Medicine and College of Pharmacy. "Our system learns from new data every month, continuously refining its forecasts as real-world patterns evolve. As we showed at the RX Summit, innovations in machine learning are not just a research endeavor, but practical tools with real public health impact."
The team continues to refine and expand the platform with ongoing feedback from an end-user advisory board made up of state and local partners. Planned additions include a public-facing version of the hotspot dashboard, integration of area development district data and expanded demographic comparisons.
Research reported in this publication was supported by the National Institute on Drug Abuse of the National Institutes of Health under Award Number R01DA057605. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
