Evaluating the reliability and quality of asthma educational content on TikTok and Bilibili: A cross-sectional content analysis

Abstract

Background

Asthma is a widespread chronic inflammatory airway disease posing a global health challenge. The digital era has made platforms like TikTok and Bilibili key sources for asthma education. However, the quality of asthma information on these platforms is unclear, raising concerns about content accuracy and reliability. This study evaluates the reliability and quality of asthma educational videos on TikTok and Bilibili.

Methods

Using “asthma” as a keyword, the top 100 videos from each platform were analyzed after excluding duplicates and irrelevant content, totaling 187 videos. All content was retrieved from the mainland-China versions of TikTok and Bilibili, and the videos analyzed were therefore limited to those uploaded and accessible within China. Videos were assessed by two senior physicians using modified DISCERN and GQS tools for reliability and quality, analyzing associations with uploader types and user engagement. Nonparametric statistical methods were applied for data analysis.

Results

Bilibili videos accounted for 51.34% of the sample, with higher median likes (101.50 vs 42.00, P = .011), shares (30.00 vs 12.00), and comments (19.00 vs 5.00) than TikTok. TikTok videos were longer (121.00 vs 83.50, P = .028) but showed weaker correlations with engagement and quality metrics. Professionals created 54.01% of videos, yet non-professionals had higher engagement. Asthma symptoms were covered in 17.11% of videos, causes in 15.51%, and diagnosis in 12.30%, with epidemiology and prevention underrepresented.

Conclusion

Our data indicate that Bilibili videos were rated as higher quality and more reliable than those on TikTok, yet both sites reward popularity over accuracy. Users seeking evidence-based asthma information should select content uploaded by verified hospitals or certified specialists and cross-check any health claims made by non-verified influencers.

Keywords

Asthma health information quality DISCERN GQS TikTok Bilibili

Introduction

Asthma represents a significant global health challenge, affecting approximately 7.4% of the world's population and imposing considerable burdens on healthcare systems and patients’ quality of life.¹ The disease is characterized by chronic airway inflammation and variable respiratory symptoms, often requiring long-term management to prevent exacerbations and hospitalizations.² Despite advances in therapeutic strategies, asthma remains a leading cause of emergency department visits, particularly among vulnerable populations.³ Recent evidence highlights the complex interplay between genetic, environmental, and socioeconomic factors in asthma pathogenesis and progression.⁴ Social determinants of health significantly contribute to asthma prevalence and disparities in health outcomes, with studies demonstrating that individuals experiencing multiple social risk factors have substantially higher odds of asthma diagnosis.⁵ Furthermore, environmental exposures such as air pollution and wildfire smoke have been strongly associated with increased asthma exacerbations and healthcare utilization.⁶ Asthma frequently coexists with other chronic conditions, with patients demonstrating higher rates of hypertension, hyperlipidemia, diabetes, cardiovascular diseases, and obesity compared to individuals without asthma.⁷ This multimorbidity pattern significantly complicates clinical management and increases the demand for healthcare services, particularly out-of-hospital care.⁸

In recent years, digital health interventions have emerged as promising tools for improving asthma control through online health communities and remote monitoring technologies.⁹ The digital transformation of healthcare information has fundamentally changed how patients access medical knowledge.¹⁰ In the contemporary era, individuals increasingly turn to social media platforms for health-related information, seeking easily accessible and relatable content about their conditions.¹¹ Bilibili, a leading video-sharing platform originating and operating primarily in China, has gained particular prominence among Chinese-speaking users for sharing health-related content. TikTok and Bilibili have emerged as particularly popular sources of health information, especially among younger demographics who prefer video-based content over traditional text-based resources.¹² These platforms offer vast amounts of user-generated content covering various aspects of asthma, including symptom management, treatment options, lifestyle recommendations, and personal experiences.¹³

This shift toward digital health information seeking presents both opportunities and challenges.¹⁴ On one hand, social media platforms can facilitate peer support and information sharing among patients, potentially enhancing self-management capabilities and treatment adherence.⁹ On the other hand, the quality and reliability of health information on these platforms remain largely unregulated, raising concerns about the potential dissemination of inaccurate or misleading content.¹⁵ The absence of standardized quality control mechanisms means that asthma-related content on TikTok and Bilibili may vary significantly in terms of medical accuracy, completeness, and evidence base.¹⁶

The reliability of health information is particularly crucial for chronic conditions like asthma, where inappropriate management can lead to serious complications.^17,18 Additionally, content promoting unproven remedies or discouraging conventional treatments may lead patients to make decisions that compromise their asthma control.¹⁹ This study aims to address these research gaps by conducting a systematic evaluation of asthma education videos on TikTok and Bilibili using validated quality assessment tools. we hypothesize that (1) the quality of asthma-related content varies between platforms due to differences in user demographics and platform algorithms; (2) videos uploaded by professional institutions will receive higher quality scores compared to those from non-professional sources; and (3) user engagement metrics, such as likes and shares, will not correlate with objective quality scores of asthma content.

Methods

Search and data collection

In this study, “asthma” was used as the search keyword. We retrieved the top 100 videos from TikTok and Bilibili respectively, and accessed each platform as a visitor. Only Chinese-language videos were eligible for inclusion. All content was retrieved from the mainland-China versions of TikTok and Bilibili, and the videos analyzed were therefore limited to those uploaded and accessible within China; no videos in other languages or from other regional versions of the platforms were evaluated. The evaluated clips provided general asthma education aimed at adult patients or caregivers of adult patients

These videos were selected based on their rankings to reduce the bias caused by personalized recommendation algorithms. We excluded duplicate videos and videos unrelated to the research topic (Figure 1). Our analysis was limited to the top 100 videos, and this limitation was supported by previous studies.^20,21 These studies indicated that videos ranked below 100 had negligible impact on the analysis. We recorded the detailed information of the selected videos. This information included video titles, uploader names, video content, video duration, likes, favorites, and shares. All data were collected between 1 June 2020 and 30 June 2025. For each video we recorded the number of likes, shares, comments, bookmarks and views present on the day of extraction; these metrics were captured only once at the time of sampling to provide a contemporaneous snapshot and were not re-checked later, because the platforms continuously update their counts. All extracted data were recorded in a Microsoft Excel spreadsheet.

Figure 1.

Search strategy and video filtering program.

Classification of videos

Based on the type of uploader, the videos are divided into four groups, as defined below: (1) Professionals: Individuals with formal medical or healthcare qualifications (such as licensed physicians, physical therapists, etc.) or those with relevant academic backgrounds (such as medical researchers specializing in orthopedics, etc.). (2) Non-professionals: Individuals without formal medical/hospitality training (such as arthritis patients sharing personal experiences, fitness enthusiasts without clinical qualifications, etc.). (3) Professional institutions: Organizations with a clear medical or healthcare mission (hospitals, academic medical centers, national health institutions such as the Chinese Center for Disease Control and Prevention, etc.). (4) Non-professional institutions: Organizations that mainly do not focus on healthcare (general media organizations, fitness brands, community groups unrelated to medical practice, etc.). The vast majority of “professional” creators were hospital-based physicians; “non-professional” creators included both social-media science popularizers and individuals with asthma (or their caregivers); and “professional institutions” were predominantly hospitals, with a small number of medical-society accounts.

Evaluating methodologies

We use the Global Quality Scale (GQS) (Supplemental Table 1), which is used to measure the overall quality of video content.^22,23 This metric is a widely recognized reliable and quality assessment tool for evaluating healthy information disseminated on digital video platforms. Using a 5-point rating scale, a GQS score of 1 indicates low content quality, while a score of 5 corresponds to excellent content quality. In this study, the overall reliability of video content was evaluated using the modified DISCERN tool (Supplemental Table 2).^24,25 DISCERN is an acronym for delivering information support for consumer health needs, a validated instrument designed to assess the quality of written health information; in this study, we used a 5-item modified version yielding a reliability score from 0 to 5.

This tool is widely recognized in the academic community and is often used in research to help consumers and healthcare providers assess the quality of health-related information. The DISCERN tool assesses the quality of video content based on five specific criteria. Clarity refers to the clarity, simplicity, and comprehensibility of the video. Relevance concerns the connection between the content and the topic. Traceability refers to the use of valid sources. Robustness involves the comprehensiveness and balance of the presented information. Fairness ensures that the content presentation is unbiased. Each criterion is scored on a binary scale: “Yes” earns 1 point, and “No” earns 0 points. Therefore, the cumulative score of a video can range from 0 to 5 points, providing a quantitative indicator for evaluating the reliability of the video.

This study evaluated every video for completeness in six core areas—epidemiology (e.g., prevalence in urban children or allergy-prone adults), etiology (e.g., dust-mite, pollen, viral upper-respiratory-infection or second-hand smoke triggers), symptoms (e.g., nocturnal cough, post-exercise wheeze or seasonal fluctuation), diagnosis (e.g., forced expiratory volume in one second/forced vital capacity ratio < 0.70 on spirometry, ≥20% peak-flow variability or positive bronchoprovocation), treatment (e.g., step-wise inhaled corticosteroid escalation, correct spacer technique or reliever-vs-controller distinctions), and prevention (e.g., smoking cessation, influenza vaccination, allergen-proof bedding or written action plans)—rating each topic as not involved, partially explained or fully explained.

Evaluation process

The evaluation of the videos was performed independently by two experienced respiratory physicians (Reviewer A: RY; Reviewer B: ZH), each with more than three years of clinical expertise in asthma management and based in the respiratory division of a tertiary teaching hospital. Prior to the assessment, both reviewers underwent standardized training on the adapted DISCERN instrument and GQS. This preparation included a calibration exercise involving 10 sample videos to align their understanding and application of the scoring criteria. Both reviewers (A and B) were blinded to both the platform and the uploader type while scoring.

The 187 videos under review were each examined separately by the two evaluators, who then provided scores across all DISCERN categories—covering aspects such as clarity, pertinence, source identification, credibility, and balance—as well as an overall GQS score. The score levels are divided into three categories: the low score range is from 1 to 2, the medium score is set at 3, and the high score range is from 4 to 5. Any differences in scoring that amounted to at least one point in either the aggregate DISCERN or GQS evaluations were flagged as discrepancies. In these cases, the evaluators would engage in a dialogue to identify the root of their disagreement, which might stem from differing interpretations of criteria such as “credibility” or “clinical accuracy.” If consensus was unattainable, a third evaluator—a highly experienced specialist in respiratory health with over ten years of expertise in asthma treatment (JL)—would be consulted. This senior professional would scrutinize the video material, review the initial assessments, facilitate a discussion to reach a consensus, and ultimately determine the final score, adhering to recognized clinical guidelines and evidence-based practices.

The primary outcome was the difference in median DISCERN and GQS scores between platforms. Secondary outcomes included engagement metrics (likes, shares, comments and collections) across quality strata, the correlation between these metrics and the DISCERN/GQS scores, and the distribution of uploader types within each quality category.

Statistical analysis

Normality will be assessed with the Shapiro–Wilk test and Q–Q plots; normally distributed variables will be summarized as means ± SD and compared with parametric methods (independent-samples t-test for two groups, one-way ANOVA with Tukey's post-hoc for three or more), whereas non-normally distributed data will be presented as medians (IQRs) and analyzed with non-parametric equivalents.

When comparing the GQS scores and DISCERN scores, we used the Fisher exact test. Due to the non-normal distribution of the data, Spearman correlation analysis will be employed to explore the correlations between various video-related variables and their relationship with the scores. Statistical significance is defined as a P value < 0.05. All statistical assessments were performed using R software version 4.4.2.

Result

Video characteristics and uploader types

We analyzed a total of 200 eligible videos; after excluding 13 that did not meet the selection criteria, 187 remained, consisting of 96 (51.34%) from Bilibili and 91 (48.66%) from TikTok. The median number of likes in the Bilibili group was higher than that in the TikTok group (101.50 vs. 42.00, P = .011). Similarly, the median number of shares (30.00 vs. 12.00) and the median number of comments (19.00 vs. 5.00) in the Bilibili group were greater, indicating a significant difference between the two groups (P < .05). In terms of video duration, the TikTok group was longer than the Bilibili group (121.00 vs. 83.50, P = .028). There was no statistically significant difference in the median number of collections (29.50 vs. 18.00) between the two groups. The general characteristics of the videos are summarized in Table 1.

Table 1.

Characteristics of the videos on TikTok and Bilibili.

Variables	Total (n = 187)	Bilibili (n = 96)	TikTok (n = 91)	P
Video duration(s), M (Q₁, Q₃)	100.00 (59.00, 187.50)	83.50 (52.75, 168.75)	121.00 (62.00, 235.50)	.028
Number of likes, M (Q₁, Q₃)	60.00 (11.00, 244.00)	101.50 (16.00, 304.75)	42.00 (8.50, 156.50)	.011
Number of collections, M (Q₁, Q₃)	22.00 (8.00, 119.00)	29.50 (9.00, 274.25)	18.00 (6.00, 75.00)	.066
Number of comments, M (Q₁, Q₃)	9.00 (2.00, 56.50)	19.00 (2.75, 86.25)	5.00 (1.50, 15.00)	.002
Number of shares, M (Q₁, Q₃)	18.00 (3.00, 68.00)	30.50 (8.00, 192.75)	12.00 (1.50, 29.00)	<.001

Out of the 187 videos, 38 were from non-professionals (accounting for 20.32%), 101 were from professionals (accounting for 54.01%), and 15 were from professional institutions (accounting for 8.02%). 33 videos were posted by non-professional institutions (accounting for 17.65%). This indicates that professionals play a leading role in creating and disseminating videos related to asthma. Additionally, specialists dominate both sites, contributing 53% of TikTok and 55% of Bilibili videos, whereas professional institutions supply <10% on each platform. Non-specialists and Non-professional institutions show modest, opposite shifts: Bilibili carries a higher Non-specialist share (26% vs 12% on TikTok), while TikTok hosts relatively more Non-professional content (23% vs 10% on Bilibili) (Figure 2).

Figure 2.

General information on asthma-related videos from TikTok and Bilibili. (A) Circular Pie chart showing the percentage of uploader types on all platforms (B), percentage stacked bar chart showing the percentage of uploader types on different platforms (TikTok and Bilibili).

There are differences in the median values of video-related indicators among different groups. The indicators of professional institutions are higher than those of other sources of videos. The number of likes (30.00 vs. 12.00), comments (30.00 vs. 12.00), and collections (30.00 vs. 12.00) of non-professional users are higher than those of professional educational videos (P < .05) (Table 2).

Table 2.

Characteristics of uploader types of videos on TikTok and Bilibili.

Variables	Total (n = 187)	Non-specialists (n = 38)	Nonprofessional institutions (n = 33)	Professional institutions (n = 15)	Specialists (n = 101)	P
Video duration(s), M (Q₁, Q₃)	100.00 (58.00, 187.50)	93.50 (59.00,263.75)	127.00 (49.00,211.00)	300.00 (153.50,431.50)	85.00 (55.00,140.00)	.003
Number of likes, M (Q₁, Q₃)	59.00 (11.00, 228.00)	156.50 (17.75,757.00)	28.00 (5.00,194.00)	574.00 (156.50,1203.50)	48.00 (11.00,113.00)	<.001
Number of collections, M (Q₁, Q₃)	22.00 (8.00, 111.50)	53.00 (11.00,506.75)	12.00 (6.00,42.00)	211.00 (40.00,890.00)	18.00 (5.00,72.00)	<.001
Number of comments, M (Q₁, Q₃)	9.00 (2.00, 56.50)	13.50 (3.25,148.25)	7.00 (3.00,23.00)	125.00 (38.50,344.50)	5.00 (1.00,32.00)	<.001
Number of shares, M (Q₁, Q₃)	18.00 (3.00, 68.00)	28.50 (8.25,272.25)	7.00 (2.00,37.00)	186.00 (53.50,451.00)	13.00 (3.00,37.00)	<.001

Video content analysis

The symptoms of asthma are the most frequently discussed topic, with 17.11% of the videos providing detailed explanations. The causes and diagnosis were covered by 15.51% and 12.30% of the videos respectively. Detailed information on epidemiology and prevention was rarely included, and only a few videos briefly mentioned these aspects (Table 3). When stratified by platform (Table S3), Bilibili videos were significantly more likely to provide detailed coverage of epidemiology and prevention (P < .001). When stratifying by uploader type (Table S4), professional institutions consistently outperformed non-specialists and nonprofessional institutions across all content domains. Most notably for Diagnosis (P = .040), 15.5% of professional-institution videos achieved the highest detail level (score 2) and 56.9% reached at least moderate detail (score ≥1), compared with only 7.0% and 49.3%, respectively, in the non-specialist/nonprofessional group. Similar gradients were observed for Epidemiology and Prevention, with professionals providing markedly more comprehensive information, while non-specialists/nonprofessional accounts frequently lacked detailed coverage. These differences indicate that users seeking accurate, in-depth asthma education—especially regarding diagnosis—should prioritize content from verified professional institutions.

Table 3.

Completeness of video content.

Video contents	Not involve (0 points)	Partial explanation (1 point)	Full explanation (2 points)
Epidemiology, n (%)	141 (75.4%)	35 (18.72%)	11 (5.88%)
Etiology, n (%)	103 (55.08%)	55 (29.41%)	29 (15.51%)
Symptoms, n (%)	48 (25.67%)	107 (57.22%)	32 (17.11%)
Diagnosis, n (%)	63 (33.69%)	101 (54.01%)	23 (12.30%)
Treatment, n (%)	82 (43.85%)	88 (47.06%)	17 (9.09%)
Prevention, n (%)	116 (62.03%)	62 (33.16%)	9 (4.81%)

Eighteen (9.6%) of the 187 videos—16 uploaded by non-professional individuals and two by non-professional institutions—contained statements that directly contradict current GINA guidelines (e.g., advocating routine use of short-acting β₂-agonists as the sole long-term control or discouraging inhaled corticosteroids in adults with persistent asthma); none of the professional or professional-institution videos contained such contradictions.

A comparative study of GQS and DISCERN score distribution of Bilibili and TikTok platforms

We classified the GQS and DISCERN scores. For reviewer A (GQS) score, there is a significant difference in distribution. Bilibili videos have a higher proportion of receiving 3 points (34.74% vs. 20.65%) and 4 points (23.16% vs. 11.96%), while TikTok participants have a higher proportion of receiving 1 point (39.13% vs. 14.74%). Reviewer B (GQS) score also shows a similar pattern. Bilibili videos are more frequently awarded 3 points (38.95% vs. 20.65%) and 4 points (23.16% vs. 11.96%), while TikTok videos are more frequently awarded 1 point (31.52% vs. 9.47%). For reviewer A (DISCERN score), there is a significant difference in distribution. Bilibili videos are more often awarded 3 points (38.95% vs. 14.13%), and TikTok videos receive 1 point (35.87% vs. 14.74%). Reviewer B's DISCERN score also shows significant differences. Bilibili videos are more likely to receive 3 points (36.84% vs. 14.13%), and TikTok videos are more likely to receive 1 point (36.96% vs. 13.68%), as shown in Table 4. Additionally, in Figure 3, we use scatter plots to more clearly display the tendency of the scores of the two reviewers.

Figure 3.

GQS and DISCERN score of asthma-related videos from TikTok and Bilibili. (A) Scatter plot showing the centralized trend of reviewer A's GQS score; (B) Scatter plot showing the centralized trend of reviewer B's GQS score. (C) Scatter plot showing the centralized trend of DISCERN score for reviewer A. (D) Scatter plot showing the centralized trend of DISCERN score for reviewer B. GQS: Global Quality Score.

Table 4.

Ratings of TikTok and Bilibili by different reviewers.

Variables	Total (n = 187)	Bilibili (n = 95)	TikTok (n = 92)	P
A (GQS scores), n (%)				<.001
1	50 (26.74)	14 (14.74)	36 (39.13)
2	45 (24.06)	24 (25.26)	21 (22.83)
3	52 (27.81)	33 (34.74)	19 (20.65)
4	33 (17.65)	22 (23.16)	11 (11.96)
5	7 (3.74)	2 (2.11)	5 (5.43)
B (GQS scores), n (%)				<.001
1	38 (20.32)	9 (9.47)	29 (31.52)
2	53 (28.34)	25 (26.32)	28 (30.43)
3	56 (29.95)	37 (38.95)	19 (20.65)
4	33 (17.65)	22 (23.16)	11 (11.96)
5	7 (3.74)	2 (2.11)	5 (5.43)
GQS scores, n (%)				<.001
1	44 (23.53)	10 (10.53)	34 (36.96)
2	51 (27.27)	28 (29.47)	23 (25.00)
3	52 (27.81)	33 (34.74)	19 (20.65)
4	33 (17.65)	22 (23.16)	11 (11.96)
5	7 (3.74)	2 (2.11)	5 (5.43)
A (Discern scores), n (%)				<.001
0	17 (9.09)	7 (7.37)	10 (10.87)
1	47 (25.13)	14 (14.74)	33 (35.87)
2	56 (29.95)	31 (32.63)	25 (27.17)
3	50 (26.74)	37 (38.95)	13 (14.13)
4	14 (7.49)	3 (3.16)	11 (11.96)
5	3 (1.60)	3 (3.16)	0 (0.00)
B (Discern scores), n (%)				<.001
0	13 (6.95)	7 (7.37)	6 (6.52)
1	47 (25.13)	13 (13.68)	34 (36.96)
2	60 (32.09)	32 (33.68)	28 (30.43)
3	48 (25.67)	35 (36.84)	13 (14.13)
4	16 (8.56)	5 (5.26)	11 (11.96)
5	3 (1.60)	3 (3.16)	0 (0.00)
Discern scores, n (%)				<.001
0	13 (6.95)	7 (7.37)	6 (6.52)
1	50 (26.74)	13 (13.68)	37 (40.22)
2	57 (30.48)	32 (33.68)	25 (27.17)
3	47 (25.13)	34 (35.79)	13 (14.13)
4	17 (9.09)	6 (6.32)	11 (11.96)
5	3 (1.60)	3 (3.16)	0 (0.00)

Table 5 presents a comparative analysis of video content quality and reliability ratings across different uploader types on TikTok and Bilibili. While there is a general trend indicating that content uploaded by Professional institutions and Specialists tends to be of higher quality, the differences observed do not reach statistical significance based on the provided analysis (P＞.05). When the same videos are stratified by platform (Table 5), both reviewers consistently awarded higher median GQS and modified-DISCERN scores to Bilibili clips than to those from TikTok (e.g., Reviewer-A GQS 3.0 vs 2.0, P = .001; Reviewer-B GQS 3.0 vs 2.0, P < .001). Although reviewers still judged Specialist- or Professional-institution content more favorably within each site, the platform effect remains significant after accounting for uploader type, indicating that BiliBili's asthma-related videos are rated as more reliable and of better overall quality than their TikTok counterparts.

Table 5.

Ratings by different commenters for different types of uploaders on TikTok and Bilibili.

Variables	Total (n = 187)	Non-specialists (n = 38)	Nonprofessional institutions (n = 33)	Professional institutions (n = 15)	Specialists (n = 101)	P
A (Discern scores), M (Q₁, Q₃)	2.00 (1.00, 3.00)	2.00 (1.00,3.00)	2.00 (1.00,2.00)	3.00 (1.50,3.00)	2.00 (1.00,3.00)	.235
B (Discern scores), M (Q₁, Q₃)	2.00 (1.00, 3.00)	2.00 (1.00,3.00)	2.00 (1.00,2.00)	3.00 (1.50,3.00)	2.00 (1.00,3.00)	.183
A (GQS scores), M (Q₁, Q₃)	2.00 (1.00, 3.00)	2.50 (1.00,3.00)	2.00 (1.00,3.00)	3.00 (2.00,3.00)	3.00 (2.00,3.00)	.164
B (GQS scores), M (Q₁, Q₃)	3.00 (2.00, 3.00)	3.00 (1.25,3.00)	2.00 (1.00,3.00)	3.00 (2.00,3.00)	3.00 (2.00,3.00)	.096

Spearman correlation analysis

Spearman correlation analysis was employed to explore the correlations among different video variables, GQS scores and DISCERN scores in asthma-related videos. There was a strong positive correlation among the number of likes, shares, comments and collections. However, the correlation between video duration and other variables was relatively weak, with most correlation coefficients ranging from 0.15 to 0.18. It is notable that the GQS score and the DISCERN score given by reviewers A and B were weakly correlated with the user interaction variables (such as the number of likes, shares and collections). The correlation coefficients between reviewers A and B for GQS and DISCERN scores were relatively high, indicating good consistency between the two reviewers. The specific correlation coefficients are shown in Figure 4.

Figure 4.

Spearman correlation analysis among different video variables, GQS, and DISCERN score concerning asthma videos. GQS: Global Quality Score.

Discussion

With asthma becoming more common and social media being the primary source for health information, this study looked at short asthma videos. Our analysis of 187 videos indicates that content on Bilibili generally achieved higher median user engagement metrics—likes, shares, and comments—and received superior quality and reliability scores from both reviewers using the GQS and modified DISCERN instruments, compared to TikTok.

This divergence may be attributed to inherent platform characteristics and audience expectations. Bilibili, originally centered around longer-form, community-oriented content, may attract users seeking more in-depth information, whereas TikTok's algorithm favors brevity and rapid consumption, potentially at the expense of informational depth and accuracy.^26,27 The observation that video duration was longer on TikTok yet correlated weakly with other engagement and quality metrics suggests that mere length does not equate to educational value, echoing concerns raised in assessments of health content on other short-video platforms.²⁸

A critical finding is the predominant role of healthcare professionals, who created over half of the analyzed videos, underscoring their active participation in digital health communication. This aligns with a broader trend where medical experts are increasingly leveraging social media to disseminate knowledge, driven by intrinsic motivations such as professional mission and a response to public health needs.²⁹ However, the fact that videos from non-professionals on Bilibili garnered significantly higher engagement metrics than those from professionals highlights a complex dynamic: while professional sources are crucial for accuracy, content from individual creators can resonate more strongly with certain audiences, possibly due to relatable presentation or perceived authenticity.²⁷ This presents a double-edged sword; high engagement with non-professional content risks amplifying misinformation if not properly vetted, a concern particularly relevant for managing chronic conditions like asthma.³⁰

The thematic analysis revealed a concerning skew in content coverage. Asthma symptoms, etiology, and diagnosis were relatively well-covered, but critical aspects like epidemiology and prevention were notably underrepresented. This gap is significant, as comprehensive asthma education should empower patients with knowledge not only about recognition and treatment but also about preventive strategies and understanding the condition's prevalence and risk factors. The omission of these topics may leave patients with an incomplete understanding of how to manage their condition holistically, potentially impacting long-term health outcomes. This finding is consistent with analyses of other disease-specific content on social media, where popular topics often overshadow equally important but less engaging information.³¹

The weak correlation observed between objective quality scores (GQS and DISCERN) and user engagement metrics (likes, shares, favorites) is perhaps one of the most disquieting results. It indicates that the factors driving user interaction—likely influenced by platform algorithms prioritizing watch time and engagement—are not aligned with those that characterize medically sound, reliable information. This creates an environment where visually appealing or emotionally charged content, regardless of its accuracy, can achieve widespread dissemination, while high-quality educational material may remain less visible.³⁰ This phenomenon, often termed the “reliability-engagement paradox,” has been documented across various health topics on social media, from mental health to nutrition and surgical procedures.³² It underscores a fundamental challenge: platform business models built on maximizing user engagement can inadvertently hinder effective health communication by not prioritizing quality and accuracy in their recommendation algorithms.³³

Our findings are consistent with recent audits of asthma content on specific platforms. McNally et al. analyzed the 100 most-liked English-language TikTok videos tagged asthma and found that only a third were created by doctors.³⁴ The rest were made by non-professionals, with herbal remedies and inhaler technique dominating the content. Despite the high proportion of lay content, median DISCERN scores were low. Mirroring our results, no correlation was found between DISCERN and likes or shares. In a follow-up study introducing the Video Health Information Credibility Score (VHICS), the same group again documented poor reliability across the top 100 videos, thus reinforcing the ‘reliability-engagement paradox’ that we had previously observed on the Chinese version of TikTok.³⁵ When comparing platforms, we found that Bilibili in our sample achieved significantly higher median DISCERN and GQS scores than the English TikTok samples described by McNally (median DISCERN score of approximately 3.3 versus approximately 1.2),³⁵ suggesting that language, cultural context, or platform governance may moderate, but not eliminate, the quality gap.

The high inter-rater reliability between the two independent reviewers for both GQS and DISCERN scores strengthens the validity of our quality assessment and suggests these tools can be consistently applied to evaluate asthma content on short-video platforms.³⁶ This methodological consistency is crucial for establishing standardized benchmarks in a field where content evaluation remains highly variable. However, the generally low to moderate absolute scores across many videos highlight an overarching problem: the overall quality of asthma information on these platforms is suboptimal.

To mitigate the risks of misinformation and maximize the potential of these platforms for public health education, a multi-faceted approach is necessary. This includes encouraging healthcare professionals and institutions to create more engaging, high-quality content; advocating for platform design changes that incorporate quality metrics into recommendation algorithms; and developing digital literacy interventions that empower patients to critically evaluate online health information.³⁷ By addressing these challenges, social media can evolve from a mere source of information into a powerful, reliable tool for supporting asthma management and improving patient outcomes.

Several limitations preclude more definite conclusions. The cross-sectional design provides a snapshot in time, but the dynamic nature of social media means content and engagement metrics are constantly evolving. Our search was conducted using specific keywords, and we only analyzed the top results based on platform-specific sorting algorithms (likely engagement-based), meaning our sample may not be fully representative of all asthma-related content available. Furthermore, we focused on two platforms prevalent in the Chinese digital ecosystem, and all videos were in Chinese. Future research should employ longitudinal designs to track content evolution, incorporate a wider array of platforms and languages, and directly investigate the impact of exposure to this content on patient knowledge, health behaviors, and clinical outcomes. We did not request official documentation, so some creators who claimed to be healthcare professionals may have been misrepresented.

In short, content on Bilibili was rated significantly higher in quality and reliability than that on TikTok, but both platforms showed a marked quality–engagement disconnect. Across sites, only videos uploaded by verified professional institutions or certified specialists consistently achieved top GQS/DISCERN scores and contained no guideline-contradictory statements, whereas high-like influencers and patient-experience vloggers delivered the lowest-quality information and occasional misinformation. Overall, the study demonstrates that while short-form platforms are now major sources of asthma information, quality is highly variable—key facts are frequently omitted, professional content underperforms in engagement, and algorithms favor popularity over accuracy—highlighting the need for viewers to prioritize verified medical accounts and for platforms to embed quality metrics into their recommendation systems.

Conclusion

Supplemental Material

sj-xlsx-1-dhj-10.1177_20552076261428193 - Supplemental material for Evaluating the reliability and quality of asthma educational content on TikTok and Bilibili: A cross-sectional content analysis

Supplemental material, sj-xlsx-1-dhj-10.1177_20552076261428193 for Evaluating the reliability and quality of asthma educational content on TikTok and Bilibili: A cross-sectional content analysis by Zhehao Xu, Jia Ding, Junlong Wang, Ruiyun Liang, Shuangfeng Xie, Xiwei Huang and Yuanjie Le in DIGITAL HEALTH

Supplemental Material

sj-xlsx-2-dhj-10.1177_20552076261428193 - Supplemental material for Evaluating the reliability and quality of asthma educational content on TikTok and Bilibili: A cross-sectional content analysis

Supplemental material, sj-xlsx-2-dhj-10.1177_20552076261428193 for Evaluating the reliability and quality of asthma educational content on TikTok and Bilibili: A cross-sectional content analysis by Zhehao Xu, Jia Ding, Junlong Wang, Ruiyun Liang, Shuangfeng Xie, Xiwei Huang and Yuanjie Le in DIGITAL HEALTH

Supplemental Material

sj-docx-3-dhj-10.1177_20552076261428193 - Supplemental material for Evaluating the reliability and quality of asthma educational content on TikTok and Bilibili: A cross-sectional content analysis

Supplemental material, sj-docx-3-dhj-10.1177_20552076261428193 for Evaluating the reliability and quality of asthma educational content on TikTok and Bilibili: A cross-sectional content analysis by Zhehao Xu, Jia Ding, Junlong Wang, Ruiyun Liang, Shuangfeng Xie, Xiwei Huang and Yuanjie Le in DIGITAL HEALTH

Footnotes

Acknowledgements

The authors would like to express their gratitude to the video uploaders for their contributions to public health. No artificial-intelligence tools were employed in any stage of manuscript preparation, including writing, data analysis and image generation.

ORCID iD

Jia Ding

Ethical approval

No clinical data, human specimens, or laboratory animals were used in this study. All information was obtained from publicly released TikTok, and BiliBili videos, and none of the data involved personal privacy concerns. In addition, this study did not involve any interaction with users; therefore, no ethics review was needed.

Contributorship

ZH, DJ, and JL planned and designed the study, analyzed and interpreted the data, and wrote the manuscript. RY and SF analyzed the data. XW created the tables presented in the manuscript. YJ, as corresponding authors, provided important advice in the study design, supervised and coordinated the study conduct process, and revised the manuscript and tables. All authors have contributed to the manuscript and approved the submitted version.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the Key Clinical Specialty of Zhejiang Province (2025004), the Medical Health Science and Technology Project of Zhejiang Province (No. 2023ky1092), the Project of Ningbo Leading Medical & Health Discipline (2026-A32), the Science and Technology Project of Jieyang City (ylws2024069), Guangdong Medical Association Clinical Research Fund (2024QC-A1006), and Scientific research project of Guangdong Provincial Bureau of Traditional Chinese Medicine (20262097).

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Availability of data and materials

All data that support the findings of this study is provided within supplementary information files.

Supplemental material

Supplemental material for this article is available online.

References

Pastore

Stadler

Sperling

, et al. T regulatory mechanisms in airway and interstitial lung diseases. Semin Respir Crit Care Med 2025; 46: 378–392.

Mims

. Asthma: definitions and pathophysiology. Int Forum Allergy Rhinol 2015; 5: S2–S6.

Rodriguez

Yuan

Bakare

, et al. Factors associated with acute unscheduled care visits for asthma in the all of US dataset. Am J Emerg Med 2025; 98: 276–282.

Laprise

Sladek

Ponton

, et al. Functional classes of bronchial mucosa genes that are differentially expressed in asthma. BMC Genomics 2004; 5: 21.

Mendy

Rowell-Cunsolo

Enkhmaa

. Asthma and summary measure of social determinants of health/health equity among adults in Mississippi, 2022. Prev Chronic Dis 2025; 22: E59.

Wilgus

M-L

Merchant

. Clearing the air: understanding the impact of wildfire smoke on asthma and COPD. Healthcare (Basel) 2024; 12: 307.

Hacikamiloglu

Dundar

Eren

, et al. Sociodemographic characteristics of asthma patients and accompanying comorbidities. Allergol Immunopathol (Madr) 2025; 53: 117–125.

Yang

Xiao

Zhang

, et al. Association of multimorbidity patterns with potential out-of-hospital clinical service needs: results from a nationally representative sample of older Chinese. Front Public Health 2025; 13: 1586215.

Karampatakis

Wood

Griffiths

, et al. Randomised controlled trial to measure effectiveness and cost-effectiveness of a digital social intervention promoted by primary care clinicians to adults with asthma to improve asthma control: protocol. BMJ Open 2025; 15: e104367.

10.

Cai

Coyte

. The impacts on population health by China's regional health data centers and the potential mechanism of influence. Digit Health 2025; 11: 20552076251314102.

11.

Valentine

Marques

Chen

, et al. Dissemination of decision aids about hip and knee osteoarthritis to Spanish-speaking adults through social Media. MDM Policy Pract 2025; 10: 23814683241309652.

12.

Sivaratnam

Hwang

Chee-A-Tow

, et al. Using social Media to engage knowledge users in health research priority setting: scoping review. J Med Internet Res 2022; 24: e29821.

13.

Zheng

Tong

Wan

, et al. Quality and reliability of liver cancer-related short Chinese videos on TikTok and Bilibili: cross-sectional content analysis study. J Med Internet Res 2023; 25: e47210.

14.

Han

Wang

, et al. Digital pathways to healthcare: a systematic review for unveiling the trends and insights in online health information-seeking behavior. Front Public Health 2025; 13: 1497025.

15.

Chiolero

Tancredi

Ioannidis

JPA

. Slow data public health. Eur J Epidemiol 2023; 38: 1219–1225.

16.

Cerqueira

Leite

França

, et al. Evaluation of the accuracy of ChatGPT in answering asthma-related questions. J Bras Pneumol 2025; 51: e20240388.

17.

Lycett

Raebel

Wildman

, et al. Theory-Based digital interventions to improve asthma self-management outcomes: systematic review. J Med Internet Res 2018; 20: e293.

18.

Peek

Richards

Muir

, et al. Blogging and social Media for mental health education and advocacy: a review for psychiatrists. Curr Psychiatry Rep 2015; 17: 88.

19.

Zhao

Zhang

. Consumer health information seeking in social media: a literature review. Health Info Libr J 2017; 34: 268–283.

20.

Zhu

Wang

, et al. Information quality of videos related to esophageal cancer on tiktok, kwai, and bilibili: a cross-sectional study. BMC Public Health 2025; 25: 2245.

21.

Mohammadzadeh Azarabadi

Gökkuş

Shalalfa

. Evaluating the quality of osteonecrosis and avascular necrosis content on YouTube: a multidimensional analysis using DISCERN, JAMA, and GQS tools. J Orthop Surg Res 2025; 20: 695.

22.

Wang

Yan

, et al. YouTube online videos as a source for patient education of cervical spondylosis-a reliability and quality analysis. BMC Public Health 2023; 23: 1831.

23.

Liu

Chen

Lin

, et al. YouTube/ Bilibili/ TikTok videos as sources of medical information on laryngeal carcinoma: cross-sectional content analysis study. BMC Public Health 2024; 24: 1594.

24.

Charnock

Shepperd

Needham

, et al. DISCERN: an instrument for judging the quality of written consumer health information on treatment choices. J Epidemiol Community Health 1999; 53: 105–111.

25.

Rees

Ford

Sheard

. Evaluating the reliability of DISCERN: a tool for assessing the quality of written patient information on treatment choices. Patient Educ Couns 2002; 47: 273–275.

26.

Yao

, et al. Short video platforms as sources of health information about HPV vaccine: a content and quality analysis. Digit Health 2025; 11: 20552076251379340.

27.

Liu

Yang

Tan

, et al. Web-Based video platforms as sources of information on body image dissatisfaction in adolescents: content and quality analysis of a cross-sectional study. JMIR Form Res 2025; 9: e71652.

28.

Gao

, et al. Blood or buzz? Decoding the quality of CBC-related short videos on Chinese platforms. Digit Health 2025; 11: 20552076251375721.

29.

Kong

. From clinics to clicks: investigating the motivations and competency framework of healthcare professionals on social media. Digit Health 2025; 11: 20552076251377845.

30.

Probst

Saffioti

Manche

, et al. Examination of social media nutrition information related to multiple sclerosis: a cross-sectional social network analysis. Public Health Nutr 2025; 28: e166.

31.

Kulu

Uyar

SBŞ

Kulu

, et al. Evaluating the quality and reliability of YouTube videos about phenylketonuria. Medicine (Baltimore) 2025; 104: e44309.

32.

Rohl

Jones

Nattis

, et al. Posttraumatic stress disorder content on TikTok: cross-sectional analysis of popular #PTSD posts. Online J Public Health Inform 2025; 17: e71209.

33.

Slaughter

Peytavin

Ugander

, et al. Community notes reduce engagement with and diffusion of false information online. Proc Natl Acad Sci U S A 2025; 122: e2503413122.

34.

McNally

Murray

Kent

. Clinical accuracy and risk of harm in asthma-related content on TikTok. Chest 2024; 167: 1297–1299.

35.

Irfan

Yasin

Yaqoob

. Breath of change: evaluating asthma information on TikTok and introducing the video health information credibility score. Cureus 2024; 16: e54247.

36.

Hübner

Thalmann

Henseler

. Disrupting the browsing experience: impact of sponsored social media content on affective flow without driving engagement. Front Neurosci 2025; 19: 1636848.

37.

Stifjell

Sandanger

Wien

. Exploring online health information-seeking behavior among young adults: scoping review. J Med Internet Res 2025; 27: e70379.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.01 MB

0.04 MB