After two years of unprecedented decline, the historic slowdown in American overdose deaths has stalled. The critical question now is whether the nation is edging towards the start of a new surge or settling into a plateau.
To answer that question, Northwestern University scientists have launched The OD Pulse, a new national, regional and state dashboard designed to track drug overdose deaths between January 1999 to April 2025, with unprecedented timeliness and precision.
The National Institutes of Health-funded project brings together multiple overdose death data sources to answer urgent questions about how the overdose crisis is changing across time, place and drug type, said principal investigator Lori A. Post, director of the Institute for Public Health and Medicine's Buehler Center for Health Policy and Economics at Northwestern University Feinberg School of Medicine.
The OD Pulse is designed to help public health agencies, clinicians, policymakers and communities respond faster and more precisely to a constantly evolving crisis. The dashboard is now live and will be updated monthly as new data become available.
"When lives are on the line, waiting a year for answers is not an option," Post said.
A clearer, faster picture of overdose deaths
Most overdose-surveillance systems report annual rates of death many months after the fact, limiting their usefulness for prevention. The new dashboard changes that.
The OD Pulse uses newly interpolated U.S. Census population estimates to calculate monthly overdose death rates, allowing direct comparisons between regions and over time. To reduce random month-to-month swings and seasonal effects, the system relies on a 12-month moving sum of overdose deaths that can produce stable signals rather than statistical noise. The data are further modeled to detect acceleration and deceleration to identify inflection points before they happen.
The platform draws heavily on the National Vital Statistics System (NVSS), including the Overdose Death Underlying Cause of Death Predict metric, which corrects for incomplete reporting in recent months. Unlike CDC WONDER, which includes only U.S. residents, NVSS captures all overdose deaths occurring in the U.S., including approximately 400 deaths each year among non-U.S. residents. These deaths still require investigation, medical response and public resources, and they remain part of the national public health burden.
Both systems (NVSS and CDC Wonder) are important. The OD Pulse integrates and models both surveillance systems to show what is happening now, not last year.
From surveillance to early warning
Beyond death rates, The OD Pulse introduces a new way to interpret and identify pending inflection points using concepts borrowed from physics.
The dashboard reports:
- Speed: the overdose death rate itself
- Acceleration: whether that rate is increasing or decreasing
- Jerk: whether the pace of acceleration or deceleration is speeding up or slowing down
The dashboard creators developed these measures originally to track infectious disease dynamics during the COVID pandemic.
"When we apply these measures to overdose deaths, they move surveillance from a rear-view mirror to something closer to a windshield," Post said. "Traditional surveillance tells us what already happened, whereas the OD Pulse shows where trends are headed."
What happens when data updates are delayed
During the federal government shutdown, access to new federal data stalled. The most recent data update from the Centers for Disease Control and Prevention was on Sept. 7, 2025, which reflected deaths only through April 2025.
"When data updates are delayed, renewed surges go unnoticed until they are well underway," Post said. "And that is exactly what we found when we analyzed data from March and April 2025."
The scientists analyzed the most recent data using the new dashboard and found that while the national overdose death rate continued to decline, the pace slowed. At the same time, the West Census Region (which includes 13 states, eight in the Mountain West and five in the Pacific division) reversed course after a prolonged deceleration and began increasing. Several states recorded two or more consecutive months of rising overdose deaths, including Alaska, Arizona, Colorado, Delaware, Hawaii, Minnesota, Nevada, New Mexico, Oklahoma and South Dakota.
"While sawtooth patterns are common, this shift marks a clear break from a prolonged period of deceleration," Post said.
Five key conclusions emerged from the team's analysis:
- 1. The U.S. national overdose death rate decelerated for at least 22 consecutive months (between August 2023 and June 2025), representing the longest and steepest sustained decline ever observed. Historically, even periods of rapid acceleration did not match the magnitude or duration of this slowdown.
- 2. The overdose crisis is not a single outbreak. It is a mosaic of overlapping drug epidemics that interact with one another over time.
- 3. The U.S. does not experience one overdose crisis. Different regions face different drugs, different supply dynamics and different trajectories.
- 4. Understanding this complexity is not academic. It is essential if prevention efforts are to save lives rather than chase last year's problem.
- 5. A rebound in overdose death rates does not automatically signal a new outbreak. The current pattern may represent a plateau, similar to those seen after periods of rapid escalation in 2018 to 2019 and again in mid 2022 to 2023.
"Importantly, the attenuation in the decline in overdose deaths began before the transition between presidential administrations," Post said. "The great deceleration reflects a convergence of policies implemented across administrations, including multilateral actions targeting fentanyl precursors and supply, expanded access to treatment and widespread naloxone distribution. Politics did not create the trend, and politics alone will not explain what comes next."
The study, "Leveraging state drug overdose data to build a comprehensive case level national dataset to inform prevention and mitigation strategies," is funded by the National Institute on Drug Abuse (grant 1R21DA058583-01) of the National Institutes of Health.
