TikTok was queried using the hashtags #melanoma, #metastaticmelanoma, and #skincancer on two pre-existing accounts. In total, 32 videos were collected by two authors in September 2024 and independently evaluated by dermatology residents and attendings at Rush University (Chicago, United States) using the DISCERN tool, a validated measure for assessing consumer health information [5]. The DISCERN questionnaire is divided into 3 parts: reliability of the publication (questions 1-8), quality of information about treatment options (questions 9-15), and the overall score of the publication (question 16). Each item is rated on a scale from 1 (poor) to 5 (excellent). An adjusted DISCERN score was calculated by averaging only the reliability-related items (questions 1‐8), since 28 of the 32 videos lacked treatment content. Only one video per TikTok creator was considered and every video was reviewed by both authors to ensure no repeated videos.

The content creators of the 32 videos consisted of physicians (n=18, 56%), nonphysicians (n=11, 34%), private companies (n=2, 6%), and organizations (n=1, 3%) (Table 1). Among the 18 physician creators, the majority were dermatologists (13/18, 72%). The mean DISCERN score across all videos was 2.08, highlighting the overall need for improved quality in melanoma-related content on TikTok. When treatment-related questions were excluded, the adjusted DISCERN score rose to 2.68, indicating that guidance on treatment is frequently lacking.

The analysis of content creators revealed that among the 18 physician creators, a significant majority (n=13, 72%) were dermatologists. This was expected as dermatology is the specialty with the most training in diagnosing and treating melanoma. However, the mean DISCERN score for videos created by dermatologists was 2.19, compared to 3.69 for emergency medicine physicians, indicating a general need for improved quality in melanoma-related content on TikTok regardless of physician training. Notably, when treatment-related questions were excluded, the adjusted DISCERN score increased from 2.08 across all videos to 2.68, suggesting that there is a frequent absence of guidance regarding treatment options in the content. Enhancing content with accurate, treatment-focused information may improve health literacy among viewers.

Physician-created videos outperformed those by nonphysicians in both quality (DISCERN: 2.3 vs 1.79, respectively; adjusted: 2.98 vs 2.36, respectively) and audience engagement metrics (mean number of likes: 17,499 vs 6515, respectively). This underscores the importance of professional involvement in online health education. Similarly, videos longer than one minute scored higher than shorter ones (DISCERN: 2.15 vs 2.01; adjusted: 2.81 vs 2.57), and educational videos outperformed anecdotal content (DISCERN: 2.18 vs 1.72; adjusted: 2.84 vs 2.13). Longer and educational videos also showed greater audience engagement. Together, these findings suggest that TikTok’s short-form format does not inherently disadvantage content that prioritizes explanation and educational intent. Rather than implying viewer preference for longer videos, the observed associations indicate that informational clarity and depth can coexist with algorithm-driven visibility on the platform.

Limitations include the study’s cross-sectional design, use of only three hashtags, platform-specific sample, and inability to verify creator credentials. Additionally, TikTok’s proprietary search algorithm may affect the visibility and discoverability of videos among different users, introducing an element of unpredictability. Lastly, the DISCERN tool was originally developed to assess written health information and may not fully capture the unique elements of video-based content. Future research should explore tools tailored to assess video content and investigate strategies for optimizing physician-led education on social media platforms.