Health claims on TikTok: what two audits found

Two peer-reviewed audits of TikTok health content found substantial misinformation, concentrated among non-professional creators. In a 2025 study of stroke videos, 31% contained misinformation; in a 2024 study of sinusitis videos, about 44% contained non-factual information. In the stroke study, how often a video was liked or shared had no relationship with how accurate it was.

What did the TikTok stroke audit find?

Published in Scientific Reports in 2025, the study analysed 100 of the most-liked TikTok videos on stroke, selected in January 2025. It scored quality with the Global Quality Score (GQS) and compared creators:

• Healthcare professionals made 34% of the videos and scored highest (GQS 3.12 on average), versus content creators (2.31) and general users (1.95).
• Misinformation appeared in 31% of videos overall: 8.8% for professionals, 42.2% for content creators and 42.9% for general users.
• Only 23% of videos addressed stroke prevention.
• Quality did not correlate with engagement (r = -0.08, p = 0.43).

The authors list limits: a single point in time, English-language videos only, selection of most-liked videos rather than all content, and creator credentials taken from self-disclosure.

Is the pattern the same for other conditions?

A University of Chicago report on a study in Otolaryngology-Head and Neck Surgery (March 2024) looked at sinusitis videos using hashtags such as #sinusitis and #sinusinfection over a single 24-hour period. About 44% of the analysed videos contained non-factual information. Nonmedical influencers with 10,000 or more followers produced nearly half of the videos and were more likely to contain misinformation and score lower on quality. Nearly 60% of their videos contained inaccurate information versus 15% for videos by medical professionals.

The direction matches the stroke audit, though the two used different methods and neither is a census of TikTok. One condition and one day, or 100 top videos on one topic, cannot describe all health content. Together, though, they support two observations for anyone reading claims online:

• Who is speaking matters: the professional versus non-professional gap appeared in both studies.
• Popularity is not a quality signal: in the stroke audit, engagement did not track accuracy.

The habit worth building is the one this blog applies to every claim: find the source, check the sample size, note the design, and ask what was actually measured. A claim about a group of people is not automatically true of you.

The bottom line

• A 2025 audit of 100 top stroke videos found misinformation in 31%: 8.8% from professionals versus about 42 to 43% from other creators.
• A 2024 audit of sinusitis videos found about 44% non-factual, with nearly 60% of non-medical influencers’ videos inaccurate versus 15% for medical professionals.
• Neither study is a full census of TikTok, and in the stroke audit engagement did not correlate with quality.