Quality analysis of semaglutide-related content on Chinese short-video platforms Bilibili, TikTok, and Rednote, focusing on how side effects are presented to the public. Evaluates accuracy, reliability, and framing differences between platforms and uploader types (healthcare providers, patients, celebrities). Identifies systemic misinformation patterns about semaglutide adverse effects in Chinese social media—relevant given China's rapidly growing GLP-1 RA market and the public health implications of inaccurate drug information reaching hundreds of millions of users.
Abstract
OBJECTIVE: This study aims to evaluate the quality, reliability, and content characteristics of semaglutide-related short videos on three major Chinese short-video platforms (Bilibili, TikTok, and Rednote), with a focus on exploring differences in the presentation of side effects across different platforms and uploader types. Rather than examining specific clinical indications or dosing regimens, this study focuses on how semaglutide-related information is presented to the general public in real-world short-video contexts. It seeks to clarify the impact of platform ecology and creator identity on drug information dissemination, thereby providing a basis for optimizing the quality of online health information and conducting targeted public health communication.
METHODS: A cross-sectional content analysis was adopted. Semaglutide-related videos were retrieved from three major Chinese short-video platforms using the Chinese keyword "" (the standard Mandarin term for "semaglutide"). This exact search term was used to simulate real-world user behavior and ensure reproducibility across platforms. Because most videos did not specify approved indications (type 2 diabetes vs. obesity) or dosage levels, analyses were conducted without stratification by clinical dose or indication. After rigorous screening, eligible videos were assessed for quality and reliability using the modified DISCERN scale and Global Quality Score (GQS). Video content was categorized into five thematic dimensions, and mentions of nine predefined side effects were systematically recorded. A dual-perspective analytical approach-comparing the full sample and a side effect-focused subsample-was employed to examine platform-based differences in adverse event portrayal.
RESULTS: A total of 607 semaglutide-related videos were included (Bilibili: 309, TikTok: 153, Rednote: 145). The results showed significant differences in content quality (M-DISCERN and GQS scores), video duration, and user engagement metrics across platforms ( < 0.05). There were statistically significant differences in content distribution among platforms ( < 0.05): Bilibili focused on Policy News and Events, Personal Experience Sharing, and Commercial Promotion; Rednote centered on Medication Education and Medical Advice/Medication Guidance; TikTok mainly focused on Policy News and Events. Regarding side effects, the distribution of gastrointestinal and hepatobiliary side effects among the three platforms was statistically significant ( < 0.05), with TikTok having the highest mention rate. These side effect narratives were typically presented without reference to specific dosing levels or approved indications. The dual-perspective analysis further revealed systematic differences in side effect portrayal, with TikTok conferring greater visibility for adverse-event narratives. In terms of content quality, the three platforms showed a gradient difference of "Rednote > TikTok > Bilibili" ( < 0.001).
CONCLUSION: The quality and reliability of semaglutide-related short videos are significantly influenced by platform ecology. Publicly accessible drug information on short-video platforms often lacks clear differentiation between approved indications and dosage regimens, contributing to generalized interpretations of drug risks. The drug information accessible to the public exhibits structural biases, with side effect narratives demonstrating a clear "platformization" pattern. The dual-perspective analysis revealed that inter-platform differences in the portrayal of specific adverse events (e.g., gastrointestinal and hepatobiliary effects) were more discernible in the side effect-focused subsample. Future interventions should be tailored to platform-specific features to enhance the completeness and scientific rigor of online pharmaceutical information.