Quality and reliability of chest Pain–Related short-form health videos on social media: A cross-sectional content analysis

Abstract

Background & Aims

Chest pain is one of the most common reasons for emergency medical visits. Increasingly, individuals seek preliminary explanations through short-form video (SFV) platforms before consulting healthcare professionals. In China, TikTok and Bilibili are major sources of public health information; however, the quality and reliability of symptom-focused chest pain content on these platforms remain poorly characterized. Unlike prior studies that primarily examine disease-specific videos, real-world health information–seeking often begins with symptoms rather than diagnoses. This study aimed to evaluate the quality, reliability, and content characteristics of chest pain–related SFVs in a real-world digital health context.

Methods

This cross-sectional study analyzed 200 Chinese-language short-form videos on adult chest pain from TikTok and Bilibili. Video quality and reliability were assessed using the JAMA benchmark criteria, Global Quality Scale, and modified DISCERN instrument. Content characteristics, uploader type, and engagement metrics were systematically evaluated, with high inter-rater reliability.

Results

Overall informational quality was suboptimal on both platforms, with uniformly low JAMA scores reflecting limited transparency regarding authorship and sources. TikTok videos demonstrated higher mean GQS and DISCERN scores and a greater proportion of medical professional uploaders compared with Bilibili. Content predominantly focused on symptom descriptions, differential diagnosis, and general management advice, while information on etiology, treatment options, and prevention was frequently incomplete. Videos produced by non-medical professional uploaders generated higher user engagement despite lower informational quality.

Conclusion

Chest pain-related SFVs on major Chinese platforms show substantial gaps in quality and reliability. Strengthening medical professional engagement, improving platform-level content governance, and promoting evidence-based symptom education may enhance digital health literacy and reduce risks associated with delayed care-seeking.

Introduction

Chest pain is one of the most common clinical symptoms and a frequent reason for seeking medical care, representing a substantial public health burden worldwide.^1,2 It may arise from a variety of acute conditions, including acute coronary syndrome (ACS), pneumothorax (PTX) and aortic dissection. ³ These emergent conditions progress rapidly and carry high mortality, underscoring the importance of timely recognition and intervention. Non-cardiac causes such as myofascial pain, gastroesophageal reflux, and intercostal neuralgia, generally carry lower cardiovascular risk but still require careful clinical evaluation to ensure accurate diagnosis and management.^4,5 Early recognition and appropriate healthcare-seeking behaviors significantly improve prognosis, highlighting the need for accessible, reliable health information.^6–10

Online Health Information Seeking (OHIS) has become a common approach for individuals to acquire knowledge about symptoms, diagnosis, and treatment.^11,12 Short-form videos (SFVs) on social media platforms, such as TikTok and Bilibili, have emerged as increasingly popular channels for scientific communication and health education, especially among younger populations.^13–16 However, the quality and reliability of user-generated content vary widely, and misinformation or incomplete information may exacerbate health anxiety and prompt inappropriate medical decisions.^17,18

Despite the rapid growth of short-form video platforms, most existing studies have centered on single diseases, such as coronary heart disease, ¹⁹ diabetes, ²⁰ or liver cancer. ²¹ However, real-world health information-seeking typically begins with symptoms rather than diagnoses. Individuals experiencing chest pain often lack disease-specific knowledge and rely on symptom-oriented content to guide their next steps.^22,23 This symptom-centered information is therefore critical for timely recognition and appropriate help-seeking but remains largely unexamined in digital health research.

To address this gap, this study aimed to evaluate the quality and reliability of chest pain–related short-form videos, compare content quality between TikTok and Bilibili, and assess whether such symptom-oriented videos adequately cover high-risk etiologies, management advice, and prevention strategies for chest pain.

Methods

Search strategy and data collection

This cross-sectional study examined two major Chinese short-video platforms—Bilibili and TikTok—for content related to chest pain. We used “胸痛” (chest pain) as the keyword to search videos in the Chinese versions of TikTok and Bilibili on March 1, 2025. Using a standardized search strategy, we collected publicly available video data up to February 10, 2025, and evaluated their content characteristics. To minimize algorithm-driven biases, we logged out of existing accounts, cleared all browser cache and search history, and then created new user accounts on both TikTok and Bilibili. This search was conducted using the default comprehensive sorting method (as defined by TikTok’s and Bilibili’s official documentation), which ranks videos based on a combination of metrics: video completion rate (defined by TikTok as the percentage of viewers who watched beyond 5 seconds; Bilibili similarly measures completion rate for short-form content), like rate, comment rate, follow rate, and upload time. The selection of videos strictly adhered to the platforms’ default display order without manual filtering or metric-based prioritization, ensuring that the sampled content reflected real-world user exposure. We selected this approach to simulate the content most likely to be encountered by typical users during routine browsing, rather than applying chronological or random sampling methods that may not reflect actual viewing patterns. As this is the default setting used by most viewers, we adopted it to reflect typical user behavior.^24,25 We retrieved the top 100 most relevant videos for the keyword from each platform (TikTok: n = 100; Bilibili: n = 100), which were systematically retrieved and numbered according to their default display order (TikTok videos: T001-T100; Bilibili videos: B001-B100), resulting in an initial pool of 200 videos. The search was limited to videos published before February 10, 2025, to avoid including overly recent publications that might skew user engagement metrics.

Video screening procedure

All included videos were in Chinese or in English with accurate Chinese subtitles. The following exclusion criteria were applied independently by four researchers (R.CH.F., QB.C., XX.Z., and LJ.H.), with discrepancies resolved through consensus meetings to minimize judgment errors and disagreements.

• Language: Non-Chinese videos (TikTok: n=1; Bilibili: n=1).

• Target audience: Pediatric-focused content (TikTok: n=2; Bilibili: n=2).

• Relevance: Irrelevant to chest pain (TikTok: n=0; Bilibili: n=1).

• Duration: Videos <10 seconds (TikTok: n=2; Bilibili: n=0) were excluded, as extremely short videos are unlikely to contain sufficient informational content for reliable assessment of quality and credibility using standardized evaluation tools.

• Publication time: Videos published <7 days before data collection (TikTok: n=0; Bilibili: n=0).

• Duplicates: Identical videos (TikTok: n=5; Bilibili: n=4).

• Availability: Removed/hidden during review (TikTok: n=2; Bilibili: n=1).

After exclusions, the final analytical sample comprised 179 videos (TikTok: n=88; Bilibili: n=91). A detailed flowchart of the selection process is provided in Figure 1.

Figure 1.

Video selection flowchart.

Sample size justification

We extracted the top 100 videos displayed in the default search results for each platform, without additional manual selection or exclusion at this stage. This approach ensured that the sample mirrored the content most likely to be encountered by typical users. The sample size of 100 videos per platform has been widely used in prior studies and is considered sufficient to achieve content representativeness in comparable analyses.^26–29

Video uploader categorization

Uploaders were categorized into four groups—medical professional individuals, non-medical professional individuals, medical professional institutions, and non-medical professional institutions—based on predefined operational criteria. Medical professional individuals were defined as verified healthcare providers with professional accreditation (eg, physicians, nurses, or other licensed healthcare professionals), whereas medical professional institutions included accredited healthcare organizations such as hospitals, medical schools, and medical centers. Detailed definitions of all uploader categories are provided in Multimedia Appendix 1.

Video evaluation

Video content was coded into multiple thematic categories (non-exclusive) to assess richness and comprehensiveness: Management recommendations, Differential diagnosis, Symptoms and pathophysiology, Traditional Chinese Medicine (TCM), Precipitating factors and etiology, Treatment and prevention. Given the potential severity of chest pain, we documented whether each video included appropriate management recommendations for viewers. Videos were analyzed for references to specific chest pain-related conditions including coronary heart disease (CHD), myocarditis, arterial dissection, PTX, pulmonary embolism (PE), pulmonary nodule/lung cancer, pneumonia, pleurisy, gastroesophageal and hepatobiliary disorders, breast conditions, blood and endocrine disorders, neuromusculoskeletal issues, herpes zoster, and psychiatric or psychological conditions.

Assessment tools

Two independent senior cardiologists (CL.H. and XD. Z), both with extensive clinical expertise in the diagnosis and management of chest pain and cardiovascular diseases, evaluated randomized video links using three validated tools ³⁰ :

The Journal of the American Medical Association (JAMA) benchmark (0-4 points; Table 1): Evaluates source reliability and content accuracy. ³¹

Table 1.

JAMA score criteria for assessing the reliability of video.

Description
1. Providing authorship.
2. Listing copyright information and references, as well as sources of content.
3. Providing initial date and subsequent updates.
4. Disclosing conflicts of interest, funding, sponsorship, advertising support, and video ownership.
1-point for each term, 4 scores in total

Abbreviations: JAMA, Journal of the American Medical Association.

Global Quality Score (GQS) (1-5 points; Table 2): Assesses overall video quality and informational value.^31,32

Table 2.

GQS score criteria for assessing the quality of video.

Score	Description
Score 1	Videos have poor flow, poor quality of the video, most information missing and are not at all useful for patients.
Score 2	Videos have generally poor quality, some information listed and are of very limited use to patients.
Score 3	Videos have moderate quality, and some important information is adequately discussed.
Score 4	Videos have good flow, good quality, and most of the relevant information is listed and useful for patients.
Score 5	Videos have excellent flow and quality and are very useful for patients.
5 grades, score scale ranges from 1 to 5

Abbreviations: GQS, Global Quality Score.

Modified DISCERN (mDISCERN) (0-5 points; Table 3): Evaluates comprehensiveness through five criteria (clarity, source reliability, balance, additional references, and uncertainty acknowledgment). ³³

Table 3.

Modified DISCERN score criteria for assessing the reliability of video.

Description
1. Is the video clear, concise, and understandable? Yes or No
2. Are valid sources cited? Yes or No
3. Is the content presented balanced and unbiased? Yes or No
4. Are additional sources of information listed? Yes or No
5. Are areas of uncertainty mentioned? Yes or No
1 point for answer ‘yes’, 0 points for answer ‘no’

Two authors independently assessed the quality of the remaining videos between March 16 and 23, 2025. In cases of scoring discrepancies, a third observer participated in a consensus meeting to determine the final scores. High inter-rater agreement was achieved across all three evaluation tools, with varying degrees of agreement per Landis and Koch’s classification: JAMA showed excellent agreement (κ=0.862, κ > 0.80), while GQS (κ=0.786) and mDISCERN (κ=0.682) reached substantial agreement (0.61-0.80). ³⁴

Statistical analysis

All statistical analyses were conducted using IBM SPSS Statistics version 25. The normality of all variables was assessed using Shapiro-Wilk and Kolmogorov-Smirnov tests. Medians and interquartile ranges (IQR) were used to describe video interaction metrics (likes, comments, saves, shares), video duration, and the JAMA, GQS, and mDISCERN scores. Non-parametric tests were primarily used for analysis: The Kruskal-Wallis test was used to compare medians across multiple groups (JAMA, GQS, and mDISCERN scores grouped by uploader classification). For Kruskal-Wallis tests, effect sizes were reported using epsilon-squared (ε²), with values of 0.01 - 0.08, 0.09 - 0.25, and > 0.25 indicating small, medium, and large effects, respectively. Wilcoxon rank-sum tests were used for pairwise comparisons (Basic information between two platforms). Spearman correlations were used to assess relationships between the continuous non-normally distributed variables (between video interaction metrics, and between these metrics and quality scres). Inter-rater reliability for categorical variables (disease classification) was assessed using Cohen’s kappa (κ), while chi-square tests were used to evaluate differences in the distribution of disease types across video sources.

Results

Baseline video characteristics

TikTok videos demonstrated significantly higher user engagement compared with Bilibili, including median likes (3469 vs. 181, P = 0.024), comments (166 vs. 22, P = 0.034), and shares (1431.5 vs. 51, P = 0.003), respectively. There were no significant differences in median saves (1422 vs. 119, P > 0.05) or days since publication (157.5 vs. 556, P > 0.05). Bilibili videos were significantly longer than TikTok videos (230 vs. 112 seconds, P < 0.001). Median JAMA scores did not differ significantly between the platforms (0.5 vs. 1, P = 0.599). Although the median Global Quality Scale (GQS) scores were identical for both platforms (3; interquartile range [IQR], 1-5), the overall distribution of scores was significantly higher for TikTok videos (P = 0.037). Similarly, TikTok videos had significantly higher median mDISCERN scores compared to Bilibili (2 vs. 1, P = 0.023).

Video sources

The distribution of video sources differed between platforms. On TikTok, the majority of videos were uploaded by medical professionals (n = 69, 78.4%), followed by medical professional institutions (n = 11, 12.5%), with a smaller proportion from non-medical professionals (n = 5, 3.4%) and non-medical professional institutions (n = 3, 3.4%). In contrast, Bilibili featured a relatively lower proportion of medical professionals (n = 51, 56.0%) and a notably higher share of non-medical professional creators (n = 32, 35.2%), alongside medical professional (n = 3, 3.3%) and non-medical professional institutions (n = 5, 5.5%) (Supplementary Materials 1).

Content themes

Most videos across both platforms addressed chest pain symptoms (TikTok: n = 72; Bilibili: n = 71), differential diagnoses (TikTok: n = 58; Bilibili: n = 51), and management suggestions (TikTok: n = 68; Bilibili: n = 64). Fewer videos explored treatment options (TikTok: n = 29; Bilibili: n = 40), etiologies and precipitating factors (TikTok: n = 39; Bilibili: n = 38), or underlying mechanisms (TikTok: n = 34; Bilibili: n = 42). Content related to prevention (TikTok: n = 14; Bilibili: n = 17) and TCM interpretations (TikTok: n = 4; Bilibili: n = 5) was relatively rare, although TCM-related videos were generally well-received by viewers. On Bilibili, professional-led videos were particularly favored.

Disease types

Table 4. summarizes disease categories associated with chest pain. Blood and endocrine disorders appeared more frequently on Bilibili, while neuromusculoskeletal conditions were more common on TikTok. Most disease categories showed no significant differences between platforms.

Table 4.

Characteristics of the disease types in TikTok and Bilibili.

Variable	TikTok (n=88)	Bilibili (n=91)	Chi-square test
			Chi-square (χ2)	P
CHD	67	62	1.424	0.233
Myocarditis	4	11	3.315	0.069
Arterial Dissection	28	30	0.027	0.870
PTX	13	25	4.315	0.038
PE	16	24	1.730	0.188
PN/LC	11	18	1.747	0.186
PNA	6	11	1.445	0.229
Pleurisy	20	17	0.447	0.504
GE	21	20	0.090	0.764
Hepatobiliary	6	8	0.242	0.623
Breast Diseases	4	3	0.186	0.667
Blood and Endocrine Diseases	0	8	8.098	0.004
Neuromusculoskeletal Diseases	40	25	6.255	0.012
Herpes	13	15	0.099	0.753
PPD	13	11	0.278	0.598

Abbreviations: CHD, Coronary Heart Disease; PTX, Pneumothorax; PE, Pulmonary Embolism; PN/LC, Pulmonary Nodule/Lung Cancer; PNA, Pneumonia; GE, Gastroesophageal; PPD, Psychiatric and Psychological Disorders.

Video quality and reliability assessment

Figure 2. Display the detailed results of the video quality evaluations using the mDISCERN and GQS instruments. For TikTok videos, the median JAMA score was 0.5 (interquartile range [IQR]: 0-2), the median GQS score was 3 (IQR: 1-5), and the median mDISCERN score was 2 (IQR: 0-3). Similarly, for Bilibili videos, the median JAMA score was 1 (IQR: 0-2), the median GQS was 3 (IQR: 1-5), and the median DISCERN was 1 (IQR: 0-3), indicating comparable levels of acceptability.

Figure 2.

The distribution of scores related to video quality and reliability assessment on Bilibili and TikTok. GQS, Global Quality Score; JAMA, Journal of the American Medical Association; mDISCERN, Modified DISCERN.

However, TikTok videos demonstrated significantly higher GQS and DISCERN scores compared to those on Bilibili, with P-values of 0.037 and 0.023, respectively. No statistically significant difference was found between the two platforms in terms of JAMA scores (P = 0.599).

Quality scores by uploader type

Kruskal–Wallis analysis demonstrated a statistically significant difference in JAMA scores among uploader types (H = 28.905, P < 0.001). Non-medical professional institutions exhibited the highest mean rank, whereas non-medical professional individuals had the lowest. In contrast, no significant differences were observed among uploader groups for GQS (H = 4.602, P = 0.203) or mDISCERN scores (H = 4.504, P = 0.212).

Correlation analysis

Given the non-normal distribution of the data, Spearman’s correlation analysis was performed to explore the relationships among various video-related variables. The results indicated strong positive correlations between the number of likes and comments (r = 0.95, P < .001), likes and shares (r = 0.94, P < .001), and likes and saves (r = 0.95, P < .001). Additionally, significant correlations were observed between shares and comments (r = 0.91, P < .001), saves and comments (r = 0.91, P < .001), and saves and shares (r = 0.94, P < .001). A weak but statistically significant negative correlation was found between video duration and the number of shares (r = -0.31, P < .001). In contrast, no significant correlations were detected between video release time and other variables.

Discussion

Principal findings

This study provides a comprehensive evaluation of chest pain-related SFVs across two major Chinese social media platforms, TikTok and Bilibili. Overall, the quality of symptom-oriented educational content was suboptimal, with both platforms demonstrating low transparency, limited source credibility, and insufficient coverage of essential medical information. ³⁵ Although symptom description and differential diagnosis were frequently included, key elements such as pathophysiology, treatment strategies, and preventive measures were substantially underrepresented. TikTok videos—more often produced by medical professionals—achieved higher reliability scores and greater user engagement, while Bilibili videos tended to be longer and covered more cognitively complex topics, likely reflecting differences in audience demographics. A significant content bias was also identified. Although CHD is an important and well-recognized cause of chest pain, the vast majority of videos disproportionately emphasized CHD while providing limited coverage of other high-risk etiologies. This overrepresentation may inadvertently reinforce a narrow disease schema among viewers and reduce public awareness of life-threatening but less familiar conditions such as aortic dissection, PE, and PTX. These findings highlight notable gaps in the accuracy, balance, and educational value of chest pain-related SFVs and underscore the need for more evidence-based, comprehensive, and symptom-oriented health communication. Importantly, this study adopts a symptom-centered rather than disease-centered perspective, which more accurately reflects real-world user behavior and fills a critical gap in digital health research.

Transparency concerns

Across both platforms, transparency was poor, with low JAMA scores indicating limited disclosure of authorship, institutional affiliation, source attribution, and evidence references. Content coverage also exhibited a consistent pattern: videos emphasized immediate symptom interpretation, self-management advice, and differential diagnosis, while less attention was given to underlying mechanisms (43.02%), evidence-based treatment options (38.55%), and preventive strategies (17.32%). This imbalance likely reflects user demand for rapid, actionable information as well as algorithmic amplification of concise, emotionally salient content.^36–38 However, focusing narrowly on symptoms without providing adequate etiological or preventive information may limit viewers’ ability to make informed health decisions and could contribute to misunderstanding of chest pain risks. ³⁹ Gao et al. reported strong echo chamber effects on short-video platforms, where algorithmic amplification favors highly engaging content and progressively narrows informational diversity. ⁴⁰ In addition, platform recommendation algorithms may amplify videos based on popularity, engagement, or personalization, potentially introducing content bias.^41,42 Recent computational studies have also suggested that data-driven and network-based modeling approaches can influence information dissemination patterns on online platforms, further contributing to selective exposure and content imbalance.^43,44 As a result, certain topics—such as CHD—may be overrepresented in top-ranked search results, whereas other high-risk but less common etiologies may receive comparatively limited visibility. This algorithm-driven amplification may also favor concise, symptom-focused content, thereby reducing the exposure of videos addressing underlying mechanisms, prevention, or comprehensive management, and ultimately constraining users’ ability to make well-informed health decisions. From a clinical and patient-centered perspective, the consistently modest quality scores observed in this study suggest that viewers may be exposed to information that lacks sufficient transparency, evidentiary grounding, and balanced presentation. In the context of chest pain—where symptoms may signal either benign conditions or life-threatening emergencies—such informational limitations may hinder individuals’ ability to accurately appraise symptom severity, assess the credibility of recommendations, and make timely care-seeking decisions. Collectively, these findings indicate that current short-form video content may not consistently provide the depth and reliability of information necessary to support informed and confident health-related decision-making.

Platform differences

Despite shared shortcomings, the two platforms differed in content quality and dissemination efficiency. TikTok had a higher proportion of videos uploaded by medical professionals (78.4% vs. 56.0%), better mDISCERN scores (P = 0.023), and generated greater user engagement, including likes, comments, and shares (P < 0.05). In contrast, Bilibili videos were longer (median 230 vs. 112 seconds) and more suitable for in-depth explanations but lagged behind in information reliability and engagement. Platform-specific patterns also reflected distinct content orientation features: Bilibili, which caters to a user base dominated by adolescents and young adults, tends to feature more cognitively challenging or abstract health topics, such as hematologic and endocrine disorders. In contrast, TikTok’s content ecosystem leans toward presenting more intuitive health issues, particularly those related to the neuromusculoskeletal system.^45,46

Content imbalance

A marked disease representation bias was observed in the chest pain-related videos. CHD dominated the content, accounting for 72.07% of all videos, with the vast majority of such videos focusing specifically on ACS—yet clinically, only approximately 10% of patients presenting with acute chest pain are ultimately diagnosed with ACS. ⁴⁷ Non-CHD etiologies, including PE and PTX, were underrepresented, potentially distorting public understanding even delaying appropriate medical responses. Because non-ACS causes constitute the majority of chest pain presentations in emergency settings, such disproportionate emphasis on CHD may distort risk perception and delay appropriate care.^47,48 Although TCM-related content was limited (<5%), these videos showed relatively higher user engagement. While engagement does not equate to preference, this pattern indicates that TCM content may capture user attention, warranting further exploration.

Public health implications

Given that individuals often search for online information at the onset of symptoms—before interacting with the healthcare system—the quality of symptom-centered digital content plays a critical role in shaping early decision-making. Inadequate or biased information may contribute to delayed presentation, inappropriate self-management, or unnecessary anxiety, all of which carry substantial public health consequence.^49,50 Notably, explicit guidance regarding when to seek emergency medical care was rarely provided. Clear recommendations on warning signs—such as persistent chest pain, syncope, severe dyspnea, or hemodynamic instability—are critical for reducing delays in time-sensitive conditions including ACS, PE, and aortic dissection. The absence of such actionable guidance may represent a particularly important gap in current short-form video content.

Limitations and future directions

This study has several limitations. First, the search strategy relied on a single keyword (“chest pain”), whereas real-world users may search using diverse symptom descriptions. Second, the final sample size—restricted by platform algorithms and exclusion criteria—may limit generalizability. Because video selection relied on the platforms’ default popularity-based sorting algorithms, the retrieved sample may have been influenced by algorithmic amplification mechanisms that preferentially promote highly engaging or commercially favored content. This may have affected the representativeness of the included videos and contributed to potential content bias. ⁵¹ Third, this analysis focused on Chinese-language platforms; thus, the findings may have potential relevance to other settings but should be interpreted with cultural and contextual caution. Cultural factors may shape symptom interpretation, trust in medical authority, and care-seeking norms, which in turn could influence how chest pain-related information is framed, disseminated, and understood in different environments. In addition, variations in platform governance and recommendation systems across countries may further affect content visibility and audience exposure patterns.^52,53 Fourth, this study did not evaluate how viewers interpret, understand, or respond to the video content. Assessing audience comprehension and behavioral impact through user-based studies may provide deeper insight into the real-world implications of symptom-oriented short-form videos.

Future research should explore multimodal evaluation frameworks incorporating narrative structure, visual communication quality, and viewer comprehension metrics. Comparative studies across platforms, countries, and content types could further illuminate how digital ecosystems shape health information exposure. Additionally, interventions designed to enhance video accuracy, creator training, and platform governance may offer pathways to improving public health communication outcomes.

Conclusion

This symptom-centered evaluation revealed significant limitations in the credibility, balance, and completeness of chest pain–related short-form videos. Both TikTok and Bilibili lacked transparent sourcing and adequate coverage of high-risk etiologies, treatment options, and preventive strategies. Strengthening the reliability of online symptom-oriented content requires coordinated action from content creators, healthcare institutions, and platform governance systems. Delivering accurate and comprehensive chest pain information may improve digital health literacy and support timely care-seeking, ultimately reducing adverse outcomes associated with delayed recognition.

Appendix.

Abbreviations

ACS

Acute coronary syndrome

CHD

Coronary Heart Disease

GQS

Global Quality Score

IQR

Interquartile ranges

JAMA

Journal of the American Medical Association

mDISCERN

Modified DISCERN

PE

Pulmonary Embolism

PN/LC

Pulmonary Nodule/Lung Cancer

PNA

Pneumonia

PTX

Pneumothorax

PPD

Psychiatric and Psychological Disorders

SFVs

Short-form videos

TCM

Traditional Chinese Medicine.

ORCID iD

Ren Cheng-han Fan https://orcid.org/0009-0000-2948-6962

Ethical considerations

This study analyzed content from publicly accessible social media platforms. No interaction with users occurred, and no private or restricted data were accessed. No identifiable personal information was collected, stored, or reported. All data were handled in accordance with the respective platforms’ terms of use and relevant ethical guidelines for internet-based research. As the study utilized publicly available data and did not involve human subjects as defined by institutional research regulations, formal ethics committee approval was not required.

Data Availability Statement

The datasets obtained and/or analyzed during the current study are available from the corresponding author on reasonable request.*