Scrolling for Surgery: Artificial Intelligence-Powered Evaluation of Labiaplasty Discourse on TikTok

Area of Science:

• Medical Aesthetics
• Digital Health Communication
• Social Media Analysis

Background:

• Labiaplasty is increasingly popular, with over 10,800 annual procedures in the USA.
• TikTok is a growing platform for health information, including labiaplasty discussions.
• Limited regulation on social media raises concerns about the accuracy and quality of health content.

Purpose of the Study:

• To analyze the characteristics, quality, and engagement of labiaplasty content on TikTok.
• To assess the reliability of AI tools in evaluating content quality and sentiment.
• To understand the influence of creator type on video engagement and quality.

Main Methods:

• Cross-sectional observational study of 110 "labiaplasty" TikTok videos (July-August 2025).
• Recorded video characteristics, engagement metrics (likes, shares, comments), and creator types.
• Assessed content quality using Global Quality Scale (GQS) by humans and AI (ChatGPT-4.5-turbo); performed sentiment analysis on comments.

Main Results:

• Surgeons (52%) and patients (40%) were primary creators; content focused on education (39%) and postoperative aspects (28%).
• Videos with ≥2000 likes were more often patient-created (52% vs. 32%) and had significantly lower GQS scores (human: 2.5 vs. 4; AI: 2 vs. 3).
• Low inter-rater reliability was observed for human sentiment classification (κ=0.161) and minimal agreement between AI and humans (κ=0.077).

Conclusions:

• Labiaplasty content on TikTok, though often surgeon or patient-generated, shows a trend where lower-quality videos achieve higher engagement.
• There is a need for surgeons to create more accurate and relatable content to counter potential misinformation.
• Current AI tools require refinement for dependable assessment of social media content quality and sentiment.