About this app. This app helps you see state-level patterns driving CDC's national predicted overdose data. This app was created by the Opioid Data Lab at the University of North Carolina, and is not endorsed by or affiliated with the US DHHS. We are just visualizing two public CDC data sources in a new way that is revealing.
How to use this app
This app works best in Chrome or Firefox. You may need to refresh or hit βΆοΈ Run in the top right corner to wake up the app (~10 seconds). Select a state to see results. Use the download icon in the top right for a printable PDF. App last updated January 15, 2025.
Check out our post using this app and the many caveats about mortality data. And, sign up for our email newsletter so we don't have to make cringey social videos when we have updates.
What the lines mean
The π prediction line is the official predicted overdose trend from CDC NVSS provisional overdose data -- you know, the data that media and policymakers use to understand what's happening with overdose trends nationally. The orange π line is the state's monthly count of final certified overdose deaths, aka the messy "truth." But it takes years for those data to post in WONDER, and are currently only reliably available through 2022 (final) or 2023 (provisional). The blue π line is a yearly line we calculated ourselves by summing the monthly π "truth" data, a recreation of the official yearly violet π prediction line.
About the heatmap
The heatmap below the lines shows π prediction line in shades of purple, the 12-month predicted overdose count. The aqua box is the 12 month window corresponding to the highest overdose month (aqua bar). By comparing the peak month (aqua) to the orange line, you can see get a sense the OD pattern contributing to the predicted peak. Sometimes, the orange peak may not align with the aqua peak month, which we explain here.
Notes and caveat
January 2023 onwards are provisional and/or partial data from WONDER.
Recreated data using monthly finalized WONDER counts from 2018 to 2022, and unadjusted provisional in 2023.
States with less than 5 OD deaths in a given month will show up as zero for that month, like Montana.
New York State consolidates New York (minus NYC) and New York City.
Suggested Explorations
Data and Code
This app was built by Nabarun Dasgupta at the University of North Carolina at Chapel Hill in January 2025. This app is not scheduled for regular updates, but can be if y'all find it useful send a request (opioiddatalab@unc.edu). Thanks to Svetla Slavova (Univ. of Kentucky) and Adams Sibley (UNC) for input.
All data used in the app are from public open sources. The app is written in Python 3.10 for data science. The app is hosted within a distributed Deepnote environment. The GitHub repo has all the files and documentation. All data processing and methods are shown in this .ipynb rendered on GitHub. The code for the app is displayed here.
Reuse permission
All graphics, data, and code are made public using the MIT License to allow others to use and share.
Copyright 2025 UNC Opioid Data Lab
Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the βSoftwareβ), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:
The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.
THE SOFTWARE IS PROVIDED βAS ISβ, WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.