Ohio State, Health Dept. Model Predicts Quarterly Overdose Deaths Using Real-Time Data

Ohio health officials could soon see overdose trends months before death certificates arrive, thanks to a Bayesian model that blends urine drug screens and EMS naloxone runs into quarterly county forecasts, medRxiv reported.

The approach, developed by researchers at Ohio State University and the Ohio Department of Health, cuts the typical reporting lag that has long hampered rapid response. Official mortality data often trails reality, leaving communities blind during sudden surges such as the one that accompanied the COVID-19 pandemic.

“Urine drug tests and naloxone administrations give us that velocity.”

The model uses Integrated Nested Laplace Approximation to fuse structured spatial effects, a random-walk temporal component, and two real-time covariates: fentanyl positivity rates from Millennium Health urine screens and EMS naloxone administration counts. A two-year rolling training window produced the lowest prediction error across 2020 through 2024.

Accuracy gains were most pronounced during the second quarter of 2020, when overdose deaths spiked unexpectedly. In large, urban counties, predictions tracked closely with final death tallies. Predictive capability in smaller counties was limited by low data availability.

Mortality data used in the study are proprietary to the Ohio Department of Health and accessible only through its Institutional Review Board or data portal. Urine drug test aggregates come from Millennium Health under a research sharing agreement. The work was funded by the National Institute on Drug Abuse under grant R01DA057668.

Ohio State University, founded in 1870 as the Ohio Agricultural and Mechanical College, is the flagship land-grant institution of the University System of Ohio. It opened with 24 students on September 17, 1873, and graduated its first class of six men in 1878; the first woman earned a degree a year later. Today the university enrolls more than 50,000 undergraduates and 15,000 graduate students, and in 2026 recorded $1.58 billion in research and development expenditures, the 12th highest nationally.

The Ohio Integrated Behavioral Health Dashboard already publishes aggregated EMS and mortality figures, offering a public window into the same streams that feed the model.