TikTok is a valuable data source for tracking the opioid crisis

Area of Science:

• Public Health
• Social Media Analytics
• Epidemiology

Background:

• Social media platforms like TikTok are increasingly used for public health surveillance.
• Opioid addiction and overdose remain significant public health crises requiring novel monitoring methods.

Purpose of the Study:

• To evaluate the utility of TikTok comments for tracking the opioid addiction and overdose crisis.
• To assess if incorporating TikTok-derived topics improves synthetic opioid mortality forecasting.

Main Methods:

• Collected over 569,000 TikTok comments from opioid-related videos (Jan 2021-June 2025).
• Extracted 200 topics using Latent Dirichlet Allocation (LDA).
• Integrated topics into ARIMA models for 6-month synthetic opioid mortality forecasting; analyzed conversational patterns using LIWC2015 and GPT models.

Main Results:

• Incorporating TikTok topics reduced forecasting Mean Absolute Error by up to 37%.
• Identified five key themes in opioid discourse: use, source, recovery, harm-reduction, and loss.
• Observed diverse first-, second-, and third-person accounts of opioid use within comments.

Conclusions:

• TikTok comments offer a valuable, large-scale data source for opioid use surveillance.
• Social media analysis can enhance the accuracy of opioid epidemic forecasting.
• The diverse discourse on TikTok provides insights into various facets of the opioid crisis.