This study used a cross-sectional observational design to evaluate TikTok videos discussing 5 rare genetic conditions: EDS, Marfan syndrome, cystic fibrosis, Wilson disease, and Gaucher disease. Videos were identified using TikTok’s native search function within the mobile app. Raters created new TikTok accounts to avoid the influence of personalized algorithms. For each condition, standardized English-language hashtags were entered into the search bar and manually screened in descending order of view count as displayed by the platform. The following search terms were used:

The first 40 videos with the highest visible view counts were included for eligibility screening, yielding 200 total videos. Videos were eligible if they were posted between January 2022 and October 2024, directly discussed the target condition, and were between 15 seconds and 4 minutes in duration, consistent with the length range of short-form health content on TikTok reported in prior studies [17]. Exclusion criteria included advertisements, promotional content, duplicate uploads, and non-English videos. After exclusions, 184 unique videos remained for analysis. All searches were conducted in December 2024 to ensure that the dataset reflected the most current content available at the time of analysis. The inclusion period (January 2022 to October 2024) was selected because TikTok expanded its maximum video length from 50 seconds to 10 minutes in early 2022, allowing for more substantive educational content. The cutoff of October 2024 was chosen to include approximately 3 years of postexpansion content while maintaining a manageable and temporally consistent dataset.

Videos were categorized by creator type based on account biographies, stated credentials, and linked websites. Categories included physicians, medical professionals, patients, influencers, nonprofit organizations, and others. Creators were classified as physicians if their profile or linked page listed an MD, DO, or an international equivalent medical degree (eg, MBBS). Medical professionals included other licensed health care workers such as registered nurses, genetic counselors, pharmacists, or physical therapists who reported professional credentials but did not hold physician degrees. This distinction was made to explore whether engagement and information quality differed between physician-led and allied health–led educational content, reflecting the unique training and communication roles of each group in digital health spaces. Influencers were defined as creators with a substantial follower base (typically exceeding 10,000 followers) or self-identified content creators who regularly produced health-related or lifestyle videos for a broad public audience but did not hold formal medical credentials or represent nonprofit organizations. Any ambiguity in classification was resolved by consensus among 5 reviewers to minimize misclassification bias.

All videos were coded using a standardized data extraction form developed for this study. The form included fields for creator type, disease, category, engagement metrics, and the quality assessment instruments (GQS and mDISCERN), allowing reviewers to score each video within the same structured framework. Five reviewers independently coded all videos, with each reviewer assessing approximately 20% of the total sample and at least 2 reviewers overlapping on every video to ensure consistency. Coding discrepancies were reviewed collectively after initial scoring and resolved by consensus at the end of data collection to ensure consistency across raters. For videos that claimed to cite peer-reviewed literature or clinical guidelines, reviewers verified the presence of such sources through in-video text, captions, or linked resources. If verification was not possible, the citation claim was coded as absent.

Video quality was assessed using 2 established tools: the GQS and the mDISCERN instrument. The GQS, originally developed for patient-facing websites and adapted for short-form videos, rates content on a 5-point Likert scale (1=poor, 5=excellent) based on flow, educational value, and comprehensiveness (Table 1) [20]. The mDISCERN, adapted from the original DISCERN framework for written materials, evaluates reliability, balance, and overall quality (Table 2) [21]. The full mDISCERN tool can be found in Multimedia Appendix 1.

Both instruments have been successfully applied in prior analyses of TikTok health content [12,17]. These instruments were selected because they provide standardized, validated frameworks for assessing health information quality, which allow cross-study comparisons despite their limitations in capturing certain stylistic features of short-form videos, such as pacing and audiovisual emphasis. To address these limitations, we applied the mDISCERN instrument, which has been adapted for short-form and audiovisual health content in prior TikTok-based research [18] and used alongside GQS in several video-based quality evaluations [17,19], and conducted reviewer calibration to ensure consistent scoring. Prior to data collection, all reviewers completed calibration exercises using sample videos to ensure consistent scoring. Any discrepancies in scoring were resolved through consensus discussion among the 5 reviewers. Table 2 was developed for this study using criteria derived from the original DISCERN instrument.

Descriptive statistics summarized the distribution of videos across creator types and diseases. GQS and mDISCERN scores were compared across creator types using Kruskal-Wallis tests due to nonparametric score distributions, followed by Dunn post hoc tests for pairwise comparisons. Chi-square tests examined associations between creator type and the citation of peer-reviewed literature or clinical guidelines. All analyses were conducted using SPSS (version 28.0; IBM Corp), with significance set at P<.05.

Engagement metrics, including views, likes, shares, and comments, were recorded. To adjust for variability in posting time, a normalized engagement ratio (NER; interactions per 1000 views) was calculated using the formula:

This allowed for standardized comparisons of audience interaction, as supported by prior TikTok engagement studies [15,19]. A higher NER reflects proportionally greater viewer interaction relative to exposure; however, it does not necessarily indicate the quality or meaningfulness of that engagement.

This study analyzed publicly available TikTok videos related to rare genetic conditions. All data were deidentified and obtained from open-access sources that do not require an account or login for access. In accordance with the determination provided by the Sam Houston State University Institutional Review Board (IRB), the project was classified as not human subjects research under 45 CFR 46.102. Therefore, formal IRB review was not required. The IRB confirmed that this classification applied because no identifiable private information or interaction with human participants occurred.

This study highlights the growing importance of TikTok as a platform for rare disease awareness and education, while also emphasizing critical disparities in the quality of medical information shared. Most videos analyzed were created by patients and influencers, who collectively attracted the highest levels of viewer engagement. Despite this popularity, these videos generally contained lower-quality medical information compared with content produced by medical professionals and physicians. This pattern aligns with previous research across other health conditions, where content created by medical professionals typically receives less audience engagement despite higher-quality scores [14,16]. Although physician-generated videos represented fewer total uploads and reached smaller audiences overall, they demonstrated higher NER compared with those produced by nonphysicians, suggesting that audience interaction may be influenced by perceived credibility or informational clarity rather than entertainment value, reflecting prior TikTok research showing a mismatch between content popularity and quality [11,13].

This study found substantial differences in quality scores across creator types. Physician- and medical professional–generated videos consistently demonstrated higher quality based on the GQS and the mDISCERN. Conversely, influencer-created and patient-generated videos scored notably lower on these scales. These findings mirror results from previous TikTok evaluations of various health topics, including diabetes, fibromyalgia, and coronary artery disease, where professionally created content consistently demonstrated greater accuracy but attracted less viewer attention [11-13].

An important observation of this analysis was the pronounced mismatch between content quality and viewer engagement. Physician- and medical professional–generated videos, despite higher quality, received fewer views and interactions compared with influencer- or patient-generated content. Influencer-created videos, in particular, garnered substantial viewership despite notably lower-quality ratings. This discrepancy, previously documented in studies on sinusitis and eating disorders on TikTok, raises significant concerns regarding public health literacy and underscores the risks associated with widespread misinformation on short-form video platforms [13,16]. The NER further emphasizes this mismatch, demonstrating that high-quality physician-created videos, despite fewer total views, generated significantly higher engagement per view. This suggests that viewers might interact more meaningfully with accurate, high-quality content even when encountered less frequently.

In the context of rare diseases, this disconnect between popularity and accuracy is particularly problematic. Patients frequently experience prolonged diagnostic delays, inconsistent expertise of health care professionals, and fragmented health care management [1,5]. Therefore, many individuals rely on platforms like TikTok for disease-specific information, emotional support, and practical management strategies [4,8]. This trend aligns with broader findings within rare disease communities, where patients and caregivers commonly use digital tools to access reliable information, reduce isolation, and navigate fragmented health care systems [22]. Foster and Ellis [23] suggest that self-diagnosis behaviors driven by platforms like TikTok may reflect complex motivations, including identity exploration, community seeking, and barriers to formal health care access. However, while patient-generated videos offer considerable emotional and community building value, they often include medically inaccurate, incomplete, or misleading information, reinforcing concerns highlighted by prior research [12,15].

Another critical finding was the notably low rate of citing peer-reviewed literature or clinical guidelines across all types of creators. Although medical professionals and physicians cited evidence slightly more frequently than influencers or patients, overall citation rates remained exceptionally low. This aligns with previous analyses of TikTok content in osteoporosis, cancer, and gastrointestinal health, where minimal referencing of credible sources was also noted [17,19]. The absence of evidence-based citations severely limits the educational effectiveness of these videos, hindering viewers’ ability to evaluate content credibility and posing tangible risks to patient safety, particularly within vulnerable rare disease populations [6].

Despite these challenges, TikTok presents significant opportunities for community support, increased visibility, and patient empowerment within rare disease communities. Patient-generated content can effectively raise awareness, provide emotional support, and amplify visibility for conditions often overlooked in mainstream health care and research [2,9]. Prior research demonstrates that hashtags like #RareDisease and #ChronicIllness serve as valuable online gathering spaces, facilitating supportive communities and collective advocacy [8,24]. For individuals living with rare conditions, often marked by isolation and limited support, sharing personal experiences can foster identity formation, validation, and belonging. Patient-created videos may help reduce stigma, increase awareness among peers, and promote self-advocacy, even when medical accuracy is imperfect. These peer-to-peer interactions complement rather than replace professional education and highlight the platform’s potential as a space for empowerment and community building.

To effectively leverage TikTok’s potential while addressing content accuracy and reliability, a multifaceted approach is required. Institutional support, such as training health care professionals in digital health communication, actively involving patient advocates in content creation, and implementing platform-level enhancements, including verified creator badges and algorithms prioritizing credible, evidence-based content, could significantly improve the quality and reliability of shared information [25-27]. Additionally, enhancing digital health literacy through targeted educational initiatives can empower viewers to critically evaluate online health information, fostering safer online communication practices [3,28].

In addition to institutional and platform-level interventions, collaboration between health care professionals and patient influencers may represent a promising strategy for improving the quality and reach of health information. Training programs that equip patients and influencers with foundational health communication and media literacy skills could strengthen the accuracy and clarity of their content while preserving authenticity. Likewise, structured partnerships between physicians, allied health professionals, and experienced content creators could expand credible outreach and help translate complex medical concepts into accessible, patient-centered narratives. Within rare disease communities, such training initiatives could be developed in collaboration with established advocacy organizations such as the National Organization for Rare Disorders, which already supports educational outreach and patient empowerment programs. Such approaches may ultimately enhance trust, reduce misinformation, and promote more equitable participation in digital health education.

While these strategies may offer promising directions for improving content quality and engagement on social media platforms, several limitations of this study should be acknowledged. While EDS, Wilson’s disease, and cystic fibrosis showed significant variation in information quality across creator types, results for Marfan and Gaucher diseases were not statistically significant. This likely reflects smaller sample sizes and limited creator diversity within these subgroups, which reduced statistical power to detect true differences.

The cross-sectional design captures only a single snapshot in TikTok’s rapidly evolving content environment, potentially limiting generalizability. The reliance on hashtags and English-language videos may have excluded relevant content, highlighting the need for broader multilingual research approaches. Additionally, the inclusion of only the most viewed videos introduces potential selection bias, as this approach may overrepresent popular content while underrepresenting less visible yet potentially higher-quality educational videos. The GQS and mDISCERN tools, though validated, were originally developed for longer-form content, underscoring the need for specialized assessment tools tailored specifically to short-form videos [29]. It is possible that certain diseases were more likely to include peer-reviewed citations than others, reflecting differences in research visibility, professional engagement, or the maturity of patient advocacy communities; however, this relationship was not specifically examined in this analysis. Future research using longitudinal designs could provide deeper insights into the direct impact of social media content on health behaviors and outcomes, further elucidating the platform’s long-term educational efficacy.

TikTok has become a significant platform for health communication, especially for rare disease communities, facilitating patient empowerment, advocacy, and education. However, our findings highlight substantial gaps in the quality and reliability of TikTok content on rare genetic conditions. Videos from patients and influencers are highly popular yet frequently lack medical accuracy and proper citations. In contrast, physicians and medical professionals produce consistently higher-quality content that remains comparatively underviewed.

Although physician-generated content had significantly higher-quality scores, influencer- and patient-generated videos dominated views. This discrepancy raises concerns about misinformation and its implications for patient safety, consistent with prior findings. Institutional support, such as targeted training in digital health communication, could empower health care professionals to produce engaging, evidence-based content. Addressing this imbalance requires increased collaboration between clinicians and patient creators, institutional training in digital health communication, and targeted platform-level interventions such as verified badges and enhanced algorithmic visibility for credible content. Additionally, improving digital health literacy is essential to equip users to critically evaluate health information online. Such coordinated efforts can transform TikTok into a more reliable and effective tool for rare disease education and community support.