Comparative Content Analysis and Expert Physician Risk Assessment of Aesthetic Procedures Promoted on TikTok in Taiwan: Quantitative Study.

BACKGROUND

Algorithm-driven social media platforms such as TikTok are increasingly creating stratified layers in the cosmetic medical market, thereby influencing patient decisions and safety. In Taiwan, TikTok has 2 parallel markets: one is a formal tier promoting "cosmetic surgery tourism" to the public and the other is an underground tier targeting Southeast Asian migrant workers, providing informal, high-risk services.

OBJECTIVE

In this study, we quantified the clinical risks embedded within these hierarchical markets and demonstrated how digital platforms exacerbate health inequalities through algorithm-driven social media content delivery.

METHODS

We conducted a dual-track content analysis of 60 TikTok videos (n=30 per group). A panel of 6 specialist physicians independently evaluated the videos using the newly developed Clinical Legitimacy and Risk Scoring (CLRS) scale. This scale assesses videos across 4 dimensions: depicted environment, operator identity, risk communication, and communication channels. Interrater reliability was evaluated using Fleiss κ.

RESULTS

The CLRS was used by 6 specialist physicians and showed a significant safety difference between the two groups. The medical tourism group (model A) had an average video score of 3.2 (SD 0.7), with the primary content type of short-form videos being "surgery experience vlogs" (9/30, 30%). Concomitantly, although professionalism was apparent in these videos, underlying risks were regularly obscured. In contrast, the migrant worker group (model B) had an extremely low average score of 1.1 (SD 0.3), indicating a complete deviation from medical standards (P<.001) and thus posing a higher risk to patient safety. The videos in model B were mainly "home surgery demonstrations" (15/30, 50%). The interrater reliability among physicians was high (Fleiss κ=0.85; P<.001). The primary coding team compiled the baseline descriptive data, whereas the board-certified physician panel independently evaluated the specific CLRS metrics.

CONCLUSIONS

The algorithmic architecture of TikTok creates a stratified marketplace, reinforcing socioeconomic stratification and health inequities. Specifically, regarding the informal aesthetic medical tier for migrant workers, the risk of infection has become a serious and urgent public health threat. Results of this study indicate that responsible public health interventions are urgently needed for the informal aesthetic medical tier targeting migrant workers and that there is a need to reassess TikTok's social responsibility in reviewing high-risk medical content in short-form videos.