AI analysis of Reddit identified over 1,177 self-reported cannabis adverse experiences requiring medical attention, providing a novel surveillance approach for cannabis product safety.
Cannabis safety regulators, pharmacovigilance researchers, digital health surveillance specialists
What the researchers found
Using keyword filtering, transformer-based algorithms, and content coding, 1,177 self-reported cannabis adverse experiences requiring medical attention were identified and characterized from Reddit posts.
Why it matters
Traditional adverse event reporting systems miss most cannabis-related problems because cannabis isn't regulated like pharmaceuticals. Social media surveillance offers a real-time window into the adverse experiences that patients and consumers actually face.
The numbers in context
1,177 self-reported adverse experiences requiring medical attention identified from Reddit posts using AI-assisted analysis.
How the study worked
Social media surveillance study using keyword filtering, transformer-based topic modeling algorithm, and inductive content coding to identify and characterize cannabis adverse experiences discussed on Reddit.
What this study cannot tell us
Self-reported online data cannot be verified. Reddit users may not represent all cannabis consumers. 'Requiring medical attention' is self-defined. Cannot determine cannabis product specifics or confirm causation. Topic modeling may miss nuanced experiences.
How to read the evidence
Novel surveillance methodology provides descriptive data but relies on unverified self-reports from a non-representative online population.
When this study was published
Published 2026, applying AI topic modeling to cannabis safety surveillance.
The bigger picture
As cannabis products diversify and regulation lags, social media may become the most important early-warning system for cannabis safety problems — similar to how Twitter data has been used for drug side effect surveillance.
Questions still open
- Which cannabis products generate the most adverse experience reports? Could real-time Reddit monitoring serve as an official safety surveillance tool? How do Reddit-reported experiences compare to formal adverse event databases?
Common questions
Can social media track cannabis side effects?
What kinds of cannabis problems do people report online?
Read the original research
Transformer-Based Topic Modeling: Characterizing Cannabis Product Adverse Experiences Self-Reported as Requiring Medical Attention on Reddit.
Journal of medical Internet research, 28, e82661
Citation
Mackey, Tim Ken; Nali, Matthew C; Larsen, Meng Zhen; Li, Zhuoran; Taylor, Cassandra L; Wolpert, Beverly; Trice, Catharine. (2026). Transformer-Based Topic Modeling: Characterizing Cannabis Product Adverse Experiences Self-Reported as Requiring Medical Attention on Reddit.. Journal of medical Internet research, 28, e82661. https://doi.org/10.2196/82661
Explore the wider topic
- THC Potency Has Changed Everything: Why Today's Weed Is Different
- Cannabis as Harm Reduction: When Weed Is the Least Bad Option
- Cannabis Legalization and Your Brain: A Decade of Legal Weed
- The Complete Guide to Cannabis Tolerance Breaks
- How to Cut Back on Weed Without Quitting: A Moderation Guide
- Microdosing Cannabis: The Case for Less Is More
- 10 Lower-Risk Cannabis Use Guidelines (Based on Published Research)
- Sober October for Weed: Your Complete Guide
- How to Take a Tolerance Break from Weed: Complete T-Break Guide
- Using Weed to Quit Opioids: The Complicated Truth
- Greening Out Explained: What Happens When You Take Too Much THC
- How to Help Someone Who Is Greening Out
- How to Avoid Greening Out: Dosing, Tolerance, and Prevention