Predictive modeling has emerged as one of West Virginia’s most effective strategies for improving overdose prevention. From January 2022 through December 2024, the West Virginia Office of Drug Control Policy (ODCP), supported by the Overdose Response Strategy (ORS) and the West Virginia Public Health Analyst (PHA), helped develop and expand a data-driven model that informs timely and targeted responses.

The effort began in Berkeley County, WV, an area heavily affected by drug trafficking due to its location along the I-81 corridor and proximity to Baltimore, MD. The model integrated real-time and historical data from public health, EMS, law enforcement and surrounding Maryland counties to identify patterns and predict overdose spikes. The model revealed a time lag between spikes in nearby counties and subsequent increases in overdoses in Berkeley County. These delays gave public health and safety partners valuable lead time to act.

The WV PHA played a central role in interpreting these trends for regional teams, helping them mobilize quickly even with limited resources. During the pilot phase, 984 doses of naloxone were deployed in model-identified high-risk areas, and over 687,000 alerts were sent to community members with information on naloxone access, overdose prevention services and treatment options. More than 250 individuals in at-risk zones received overdose supplies or education, and 31 were connected to treatment services.
Building on the pilot’s success, the model was expanded to include all ONDCP-designated action counties. The WV PHA guided this expansion by helping regional coordinators tailor local response plans based on each county’s needs. The PHA also distributed educational materials, supported overdose spike response protocols and led tabletop exercises to help teams improve their readiness. The model continues to be updated regularly to ensure its predictions remain accurate and responsive to evolving drug trends.

This work helped lay the foundation for the I-81 Corridor Multi-State Collaborative, a regional partnership that includes West Virginia, Maryland, Virginia, Pennsylvania and other nearby states. The Collaborative enables these states to share data and coordinate overdose prevention efforts based on predictive modeling. Throughout the project, the WV PHA has been instrumental in ensuring that data is not only collected and analyzed but also applied effectively in the field. This includes helping partners make data-informed decisions about where to focus naloxone distribution and outreach and ensuring that community teams understand and respond to predictive alerts in real-time. The model is now recognized as a leading example of data-informed overdose prevention and has drawn interest from other states seeking to implement similar strategies.