Evaluating the content quality and trustworthiness of snakebite envenomation videos on social media platforms: insights from TikTok and Bilibili Authors: Jiaye Du, Xiaohan Fang, Ting Liu, Yuqing Song, Zhan Gao Published: 2026 ABSTRACT BackgroundSnakebite envenomation is a severe medical emergency and major global health threat. As short-video platforms increasingly shape public access to health information, accurate and reliable digital education is essential for snakebite prevention and first-aid decision-making. Although the quality of medical videos on short-video platforms has been examined for several diseases, little is known about the quality, reliability, and content characteristics of snakebite envenomation-related videos on major Chinese platforms. This study aimed to systematically assess the quality, reliability, and content characteristics of snakebite envenomation-related videos on TikTok (Douyin) and Bilibili, and to explore strategies for improving the quality of online health information.MethodsOn February 17, 2026, we searched TikTok (Douyin) and Bilibili using the keywords “snakebite envenomation” and “snake envenoming”. After videos with content irrelevant to the study keywords, commercial advertisements, duplicate videos, and videos uploaded within the preceding week were excluded according to predefined criteria, 220 eligible videos were included in the final analysis. For each video, we extracted duration, audience engagement metrics, uploader type, and other basic characteristics. Video quality was independently evaluated in a double-blind manner using the Global Quality Score (GQS), modified DISCERN (mDISCERN), and Journal of the American Medical Association (JAMA) benchmarks. Group differences were analyzed using Mann–Whitney U and Kruskal–Wallis tests, and correlations among variables were assessed using Spearman’s rank correlation.ResultsOf the 300 initially screened videos, 220 were included in the final analysis, comprising 114(51.82%) videos from TikTok (Douyin) and 106(48.18%) from Bilibili. Most videos were presented in Mandarin Chinese. Etiology, clinical manifestations, treatment, and diagnosis were frequently covered, whereas epidemiology and prevention were less commonly addressed. Bilibili videos were significantly longer than TikTok (Douyin) videos 311.00 (150.75, 514.25) vs 101.00 (58.50, 171.00) seconds (P < 0.001), whereas TikTok (Douyin) videos had significantly more shares 59.00 (10.25, 717.75) vs 34.00 (3.00, 282.25) (P = 0.049). The overall median GQS, mDISCERN, and JAMA scores were 3.00 (3.00, 4.00), 3.00 (3.00, 3.00), and 1.00 (1.00, 2.00), respectively. No significant platform-based differences were observed in GQS or mDISCERN scores, but JAMA scores differed significantly between platforms (P = 0.020). Videos uploaded by healthcare professionals and nonprofit organizations had significantly higher GQS scores than those uploaded by individual users 4.00 (4.00, 5.00) and 4.00 (3.00, 4.00) vs 3.00 (2.00, 3.00), respectively; (P < 0.001), as well as higher JAMA scores both 2.00 (2.00, 2.00) vs 1.00 (1.00, 1.00) (P < 0.001). Engagement metrics were strongly interrelated (all P < 0.001), but showed limited associations with quality scores. GQS and JAMA correlated only with shares (P = 0.015 and P = 0.045, respectively), whereas mDISCERN was positively correlated with likes, collections, comments, and shares (P = 0.042, P = 0.025, P = 0.011, and P = 0.004, respectively).ConclusionsSnakebite envenomation-related videos on TikTok (Douyin) and Bilibili provided moderately useful health information, but showed important limitations in preventive content coverage and information transparency. Videos uploaded by healthcare professionals and nonprofit organizations were generally of higher quality and transparency than those uploaded by individual users, while audience engagement was not consistently aligned with content quality. These findings highlight the need for greater professional involvement, clearer source disclosure, and more standardized prevention- and first-aid-oriented content to improve the reliability and public health value of snakebite-related information on short-video platforms.Supplementary InformationThe online version contains supplementary material available at 10.1186/s41182-026-01006-5. Introduction Snakebite envenomation is a severe medical emergency that poses a major threat to global public health [1]. Beyond its acute life-threatening effects, snakebite can cause severe complications such as respiratory failure, acute kidney injury, tissue necrosis, and coagulopathy, leading to substantial long-term health consequences [2–5]. Insufficient healthcare resources and unequal access to antivenom (ASV) further exacerbate this burden, particularly in rural and low-resource regions where delays or inappropriate treatments remain common [6, 7]. These characteristics make accurate, timely, and accessible public education especially important for snakebite prevention and first-aid decision-making. With the rapid advancement of digital technology, short-video platforms have become increasingly important in health information dissemination [8]. Platforms such as TikTok and Bilibili, characterized by extensive user bases and rapid information diffusion, have emerged as key sources of public health education [8, 9]. As a leading video-sharing platform among young users in China, Bilibili hosts diverse user-generated and professionally produced content, with approximately 107 million daily active users (DAUs) [10]. In contrast, the domestic version of TikTok (Douyin), reports over 750 million DAUs and is among the most influential mobile applications in China [11]. Short videos, through their visual and engaging formats, facilitate health awareness and disease literacy; however, the lack of professional review mechanisms and inconsistent content accuracy may contribute to the spread of misleading or erroneous information, potentially influencing public health decisions [12–14]. Previous studies have evaluated the quality and reliability of online videos on several medical topics, including transarterial chemoembolization, laryngeal carcinoma, atopic dermatitis, premature ovarian failure, and thyroid eye disease [8, 9, 11, 15, 16]. These studies generally suggest that medical videos on short-video platforms vary substantially in quality, reliability, and educational value. However, despite the highly specialized and time-sensitive nature of snakebite management [2, 17, 18], evidence regarding the quality and reliability of snakebite envenomation-related videos on Chinese short-video platforms remains limited. This gap is important because inaccurate or incomplete information about snakebite first aid, antivenom use, or complications may delay appropriate medical treatment. Therefore, this study aimed to systematically assess the quality, reliability, and content characteristics of snakebite envenomation-related videos on Douyin and Bilibili. By identifying differences and shortcomings in health information dissemination across these platforms, this study seeks to provide evidence-based insights to enhance the scientific quality of health content, inform regulatory strategies, and strengthen public health literacy in the digital era. Materials and methods Ethical considerations This study did not involve human participants, clinical data, experimental animals, or histological analysis. All data were sourced from publicly available videos on TikTok (Douyin) and Bilibili, with data collection conducted in full compliance with the platforms' terms of service. No private or personally identifiable information was involved, and no interactions with users occurred during the study. Therefore, ethical approval was not required. Data extraction of basic information On February 17, 2026, we conducted a video search using the keywords “snakebite envenomation” and “snake envenoming” on Douyin, the mainland Chinese version of TikTok (https://www.douyin.com), and Bilibili (https://www.bilibili.com). To minimize potential bias from search history and personalized recommendations, data collection was performed without user account authentication on both platforms. And all data were collected manually by researchers; no automated scraping/crawling tools, bots, APIs, or bulk export functions were used. The initial video search and preliminary screening were performed by a general practitioner and took approximately one day. Videos were sorted according to the default comprehensive ranking algorithms of TikTok (Douyin), and Bilibili. Videos were sorted based on the default comprehensive ranking algorithms of TikTok (Douyin) and Bilibili, with the top 150 videos selected according to the default ranking (300 videos total). Only publicly visible information on each video page was recorded; no access restrictions were bypassed and no private data were collected. Detailed data collection methods are shown in Fig. 1. For each included video, we collected video duration, upload date, audience engagement metrics, and uploader identity information. Video uploaders were categorized into three groups: healthcare professionals (HCPs), individual users, and nonprofit organizations (NPOs).Fig. 1Search strategy for videos on snakebite envenomation Search strategy for videos on snakebite envenomation Exclusion criteria (1) irrelevant content to the target keywords (“snakebite envenomation” and “snake envenoming”); (2) commercial advertisements; (3) duplicate videos; (4) videos uploaded within the past week. Video quality assessment To ensure methodological rigor, this study employed a double-blind evaluation process to prevent any mutual influence between researchers during the initial assessment. The two independent evaluators, both general practitioners, received standardized training on the scoring criteria before assessment. During the scoring stage, both evaluators independently assessed all 220 included videos. Each evaluator took approximately one day to complete the scoring process. When disagreements arose between the two evaluators, a third researcher, also a general practitioner, intervened to facilitate consensus. Three validated evaluation tools were used to assess video quality: Global Quality Score (GQS) [19], modified DISCERN score (mDISCERN) [20–22], and Journal of the American Medical Association (JAMA) [23, 24]. The GQS was used to evaluate the overall quality, coherence, and usability of the videos, with a subjective rating of information completeness and educational value using a 1–5 Likert scale. mDISCERN assessed five dimensions: clarity, relevance, traceability, robustness, and fairness, with a maximum score of 5. JAMA evaluated four key dimensions: information source, timeliness, content validity, and compliance with information disclosure, using a 0–4 composite score. Specific scoring details are provided in Tables 1, 2, and 3.Table 1The Global Quality Score (GQS) quality criteriaItem featuresPointsPoor quality; poor flow of the videos; most information missing; not at all useful for patients1Generally poor quality; some information listed, but many important topics missing; of very limited use to patients2Moderate quality; suboptimal flow; some important adequately discussed, but other information poorly discussed; somewhat useful for patients3Good quality and generally good flow; most of the relevant information listed, but some topics not covered; useful for patients4Excellent quality and flow; very useful for patients5Table 2The Modified DISCERN (mDISCERN) quality criteriaReliability score1. Is the video clear, concise, and understandable?2. Are valid sources cited?3. Is the content presented balanced and unbiased?4. Are additional sources of content listed for patient reference?5. Are areas of uncertainty mentioned?Table 3The Journal of the American Medical Association (JAMA) benchmark criteriaScoreScore component1 scoreAuthorshipAuthor and contributor credentials and their affiliations should be provided1 scoreAttributionClearly lists all copyright information and states references and sources for content1 scoreCurrencyInitial date of posted content and subsequent updates to content should be provided1 scoreDisclosureConflicts of interest, funding, sponsorship, advertising, support, and video ownershipshould be fully disclosed Table 1The Global Quality Score (GQS) quality criteriaItem featuresPointsPoor quality; poor flow of the videos; most information missing; not at all useful for patients1Generally poor quality; some information listed, but many important topics missing; of very limited use to patients2Moderate quality; suboptimal flow; some important adequately discussed, but other information poorly discussed; somewhat useful for patients3Good quality and generally good flow; most of the relevant information listed, but some topics not covered; useful for patients4Excellent quality and flow; very useful for patients5 The Global Quality Score (GQS) quality criteria Table 2The Modified DISCERN (mDISCERN) quality criteriaReliability score1. Is the video clear, concise, and understandable?2. Are valid sources cited?3. Is the content presented balanced and unbiased?4. Are additional sources of content listed for patient reference?5. Are areas of uncertainty mentioned? The Modified DISCERN (mDISCERN) quality criteria Table 3The Journal of the American Medical Association (JAMA) benchmark criteriaScoreScore component1 scoreAuthorshipAuthor and contributor credentials and their affiliations should be provided1 scoreAttributionClearly lists all copyright information and states references and sources for content1 scoreCurrencyInitial date of posted content and subsequent updates to content should be provided1 scoreDisclosureConflicts of interest, funding, sponsorship, advertising, support, and video ownershipshould be fully disclosed The Journal of the American Medical Association (JAMA) benchmark criteria Conflicts of interest, funding, sponsorship, advertising, support, and video ownership should be fully disclosed Statistical analysis All statistical analyses and visualizations in this study were performed using R software (version 4.3.3). Categorical variables were summarized as frequencies and percentages (%), and continuous variables as median and interquartile range (IQR). For two-group comparisons of non-normally distributed quantitative variables, the Mann–Whitney U test was used; for comparisons across multiple groups, the Kruskal–Wallis test was applied. Intra-observer and inter-observer agreement for the JAMA, mDISCERN, and GQS scores was assessed using the intraclass correlation coefficient (ICC). ICC values were interpreted according to established guidelines as excellent (> 0.90), good (0.75–0.90), moderate (0.50–0.75), or poor (< 0.50) [25]. Associations between quantitative variables were evaluated using Spearman’s rank correlation. A two-tailed P < 0.05 was considered statistically significant. Results Video baseline characteristics Of the 300 initially screened videos, 220 met the inclusion criteria and were included in the final analysis. As shown in Fig. 2, 114 (51.82%) videos were from TikTok (Douyin), while 106 (48.18%) videos were from Bilibili, with TikTok (Douyin) contributing a slightly larger proportion of videos. All included videos 220 (100%) were produced by Chinese content creators. Regarding video language, 211 (95.91%) videos were in Mandarin Chinese, 4 (1.82%) videos were in English, and 5 (2.27%) videos were in Chinese dialects. The videos were uploaded across all four seasons. As shown in Fig. 3, 43 (19.55%) videos were uploaded in spring, 128 (58.18%) videos in summer, 33 (15.00%) videos in autumn, and 16 (7.27%) videos in winter. The baseline characteristics of all included videos were analyzed. The median video duration was 154.00 (73.00, 375.25). In terms of audience engagement, the median number of likes was 720.00 (82.25, 6688.00), the median number of collections was 136.00 (16.75, 1368.25), the median number of comments was 67.00 (7.75, 623.50), and the median number of shares was 40.00 (8.75, 537.50). Regarding video content, most videos covered etiology 208 (94.55%), clinical manifestations 201 (91.36%), treatment 197 (89.55%), and diagnosis 187 (85.00%). In contrast, epidemiology 70 (31.82%) and prevention 40 (18.18%) were less frequently addressed (Table 4).Fig. 2Platform distribution of snakebite envenomation-related videosFig. 3Video upload seasonTable 4Video featuresVariablesTotal (n = 220)General informationVideo length(s),M (Q1,Q3)154.00 (73.00, 375.25)Likes,M (Q1,Q3)720.00 (82.25, 6688.00)Collections,M (Q1,Q3)136.00 (16.75, 1368.25)Comments,M (Q1,Q3)67.00 (7.75, 623.50)Shares,M (Q1,Q3)40.00 (8.75, 537.50)Video contentEpidemiology70(31.82%)Etiology208(94.55%)Clinical manifestation201(91.36%)Diagnosis187(85.00%)Treatment197(89.55%)Prevention40(18.18%)Video qualityGQS score,M (Q1,Q3)3.00 (3.00, 4.00)mDISCERN score,M (Q1,Q3)3.00 (3.00, 3.00)JMMA, M (Q₁, Q₃)1.00 (1.00, 2.00)Video languageChinese (Mandarin)211 (95.91%)English4 (1.82%)Chinese (dialect)5 (2.27%)Video upload seasonSpring43 (19.55%)Summer128 (58.18%)Autumn33 (15.00%)Winter16 (7.27%)Video creatorChinese220(100%) Platform distribution of snakebite envenomation-related videos Video upload season Table 4Video featuresVariablesTotal (n = 220)General informationVideo length(s),M (Q1,Q3)154.00 (73.00, 375.25)Likes,M (Q1,Q3)720.00 (82.25, 6688.00)Collections,M (Q1,Q3)136.00 (16.75, 1368.25)Comments,M (Q1,Q3)67.00 (7.75, 623.50)Shares,M (Q1,Q3)40.00 (8.75, 537.50)Video contentEpidemiology70(31.82%)Etiology208(94.55%)Clinical manifestation201(91.36%)Diagnosis187(85.00%)Treatment197(89.55%)Prevention40(18.18%)Video qualityGQS score,M (Q1,Q3)3.00 (3.00, 4.00)mDISCERN score,M (Q1,Q3)3.00 (3.00, 3.00)JMMA, M (Q₁, Q₃)1.00 (1.00, 2.00)Video languageChinese (Mandarin)211 (95.91%)English4 (1.82%)Chinese (dialect)5 (2.27%)Video upload seasonSpring43 (19.55%)Summer128 (58.18%)Autumn33 (15.00%)Winter16 (7.27%)Video creatorChinese220(100%) Video features Platform-based differences in video characteristics Compared with TikTok (Douyin), Bilibili videos were significantly longer, with a median duration of 311.00 (150.75, 514.25) seconds versus 101.00 (58.50, 171.00) seconds for TikTok (Douyin) videos (P < 0.001). Regarding audience engagement metrics, no significant differences were observed between Bilibili and TikTok (Douyin) in the number of likes, collections, or comments. However, TikTok (Douyin) videos had a significantly higher number of shares than Bilibili videos, with median values of 59.00 (10.25, 717.75) and 34.00 (3.00, 282.25), respectively (P = 0.049) (Table 5).Table 5The basic characteristics of snakebite envenomation videos on TikTok (Douyin) and BilibiliVariablesBilibili (n = 106)TikTok (Douyin) (n = 114)PGeneral informationVideo length(s),M (Q1,Q3)311.00 (150.75, 514.25)101.00 (58.50, 171.00) < .001Likes,M (Q1,Q3)802.00 (27.00, 5733.75)542.50 (116.25, 7126.75)0.437Collections,M (Q1,Q3)155.00 (14.25, 1466.75)99.00 (21.00, 1293.75)0.850Comments,M (Q1,Q3)126.50 (3.25, 653.00)47.00 (9.25, 547.25)0.792Shares,M (Q1,Q3)34.00 (3.00, 282.25)59.00 (10.25, 717.75)0.049Video contentEpidemiology48 (45.28%)22 (19.30%)–Etiology102 (96.23%)106 (92.98%)–Clinical manifestation95 (89.62%)106 (92.98%)–Diagnosis90 (84.91%)97 (85.09%)–Treatment96 (90.57%)101 (88.60%)–Prevention24 (22.64%)16 (14.04%)–Video qualityGQS score,M (Q1,Q3)3.00 (3.00, 4.00)3.00 (3.00, 4.00)0.990mDISCERN score,M (Q1,Q3)3.00 (3.00, 3.00)3.00 (3.00, 3.00)0.148JMMA, M (Q₁, Q₃)1.00 (1.00, 2.00)1.00 (1.00, 2.00)0.020M: median, Q₁: 1st quartile, Q₃: 3st quartile Table 5The basic characteristics of snakebite envenomation videos on TikTok (Douyin) and BilibiliVariablesBilibili (n = 106)TikTok (Douyin) (n = 114)PGeneral informationVideo length(s),M (Q1,Q3)311.00 (150.75, 514.25)101.00 (58.50, 171.00) < .001Likes,M (Q1,Q3)802.00 (27.00, 5733.75)542.50 (116.25, 7126.75)0.437Collections,M (Q1,Q3)155.00 (14.25, 1466.75)99.00 (21.00, 1293.75)0.850Comments,M (Q1,Q3)126.50 (3.25, 653.00)47.00 (9.25, 547.25)0.792Shares,M (Q1,Q3)34.00 (3.00, 282.25)59.00 (10.25, 717.75)0.049Video contentEpidemiology48 (45.28%)22 (19.30%)–Etiology102 (96.23%)106 (92.98%)–Clinical manifestation95 (89.62%)106 (92.98%)–Diagnosis90 (84.91%)97 (85.09%)–Treatment96 (90.57%)101 (88.60%)–Prevention24 (22.64%)16 (14.04%)–Video qualityGQS score,M (Q1,Q3)3.00 (3.00, 4.00)3.00 (3.00, 4.00)0.990mDISCERN score,M (Q1,Q3)3.00 (3.00, 3.00)3.00 (3.00, 3.00)0.148JMMA, M (Q₁, Q₃)1.00 (1.00, 2.00)1.00 (1.00, 2.00)0.020M: median, Q₁: 1st quartile, Q₃: 3st quartile The basic characteristics of snakebite envenomation videos on TikTok (Douyin) and Bilibili M: median, Q₁: 1st quartile, Q₃: 3st quartile In terms of content distribution, both platforms mainly covered etiology, clinical manifestations, diagnosis, and treatment. As shown in Fig. 4, Bilibili had a higher proportion of videos addressing epidemiology than TikTok (Douyin) 48 (45.28%) vs 22 (19.30%). The proportion of etiology-related videos was slightly higher on Bilibili than on TikTok (Douyin) 102 (96.23%) vs 106 (92.98%). TikTok (Douyin) had a slightly higher proportion of videos covering clinical manifestations than Bilibili 106 (92.98%) vs 95 (89.62%). Diagnosis-related content showed similar coverage on both platforms 90 (84.91%) vs 97 (85.09%). The proportion of treatment-related videos was slightly higher on Bilibili than on TikTok (Douyin) 96 (90.57%) vs 101 (88.60%). Prevention-related content was less frequently covered on both platforms, but the proportion remained higher on Bilibili than on TikTok (Douyin) 24 (22.64%) vs 16 (14.04%).Fig. 4Distribution of video content categories across platforms Distribution of video content categories across platforms Uploader distribution differences in video characteristics Among the 220 snakebite envenomation videos, the majority were uploaded by individual users 167 (75.91%), followed by healthcare professionals HCPs 32 (14.55%) and nonprofit organizations NPOs 21 (9.55%). TikTok (Douyin) showed a more diverse uploader profile, with 73 (64.04%) of videos from individual users, 27 (23.68%) from HCPs, and 14 (12.28%) from NPOs, whereas Bilibili was predominantly populated by individual users 94 (88.68%), with smaller contributions from HCPs 5 (4.72%) and NPOs 7 (6.60%) (Fig. 5).Fig. 5Distribution of uploader types across platforms Distribution of uploader types across platforms Video length varied significantly across uploader types (P < 0.001). Individual-user videos were the longest 226.00 (94.50, 430.00), followed by HCPs 101.00 (58.50, 136.50), and NPOs were the shortest 46.00 (31.00, 125.00). Comment counts differed significantly (P = 0.019), with individual-user videos having a median of 115.00 (11.00, 731.00) comments compared with 13.50 (7.50, 42.00) for HCPs and 20.00 (4.00, 1474.00) for NPOs. No significant differences were observed among uploader types in terms of likes, collections, or shares (Table 6).Table 6Characteristics, quality, and reliability of snakebite envenomation videos by different uploaders on TikTok and BilibiliVariablesNPO (n = 21)HCPs(n = 32)Individual users (n = 167)PVideo length(s),M (Q1,Q3)46.00 (31.00,125.00)101.00 (58.50,136.50)226.00 (94.50,430.00) < .001Likes,M (Q1,Q3)607.00 (29.00,16,001.00)149.00 (83.75,1034.00)942.00 (101.50,6785.50)0.117Collections,M (Q1,Q3)52.00 (11.00,1024.00)54.00 (15.75,638.25)151.00 (21.50,1621.00)0.302Comments,M (Q1,Q3)20.00 (4.00,1474.00)13.50 (7.50,42.00)115.00 (11.00,731.00)0.019Shares,M (Q1,Q3)66.00 (8.00,1272.00)34.00 (13.25,458.50)47.00 (8.50,508.50)0.793GQS score,M (Q1,Q3)4.00 (3.00,4.00)4.00 (4.00,5.00)3.00 (2.00,3.00) < .001mDISCERN score,M (Q1,Q3)3.00 (3.00,3.00)3.00 (3.00,3.00)3.00 (3.00,3.00)0.557JMMA,M (Q₁, Q₃)2.00 (2.00,2.00)2.00 (2.00,2.00)1.00 (1.00,1.00) < .001M: median, Q₁: 1st quartile, Q₃: 3st quartile Table 6Characteristics, quality, and reliability of snakebite envenomation videos by different uploaders on TikTok and BilibiliVariablesNPO (n = 21)HCPs(n = 32)Individual users (n = 167)PVideo length(s),M (Q1,Q3)46.00 (31.00,125.00)101.00 (58.50,136.50)226.00 (94.50,430.00) < .001Likes,M (Q1,Q3)607.00 (29.00,16,001.00)149.00 (83.75,1034.00)942.00 (101.50,6785.50)0.117Collections,M (Q1,Q3)52.00 (11.00,1024.00)54.00 (15.75,638.25)151.00 (21.50,1621.00)0.302Comments,M (Q1,Q3)20.00 (4.00,1474.00)13.50 (7.50,42.00)115.00 (11.00,731.00)0.019Shares,M (Q1,Q3)66.00 (8.00,1272.00)34.00 (13.25,458.50)47.00 (8.50,508.50)0.793GQS score,M (Q1,Q3)4.00 (3.00,4.00)4.00 (4.00,5.00)3.00 (2.00,3.00) < .001mDISCERN score,M (Q1,Q3)3.00 (3.00,3.00)3.00 (3.00,3.00)3.00 (3.00,3.00)0.557JMMA,M (Q₁, Q₃)2.00 (2.00,2.00)2.00 (2.00,2.00)1.00 (1.00,1.00) < .001M: median, Q₁: 1st quartile, Q₃: 3st quartile Characteristics, quality, and reliability of snakebite envenomation videos by different uploaders on TikTok and Bilibili Video length(s), M (Q1,Q3) Likes, M (Q1,Q3) Comments, M (Q1,Q3) Shares, M (Q1,Q3) GQS score, M (Q1,Q3) mDISCERN score, M (Q1,Q3) JMMA, M (Q₁, Q₃) M: median, Q₁: 1st quartile, Q₃: 3st quartile Video quality assessment and observer agreement Video quality and reliability were assessed using the GQS, mDISCERN, and JAMA scoring systems. Overall, the median GQS score of all included videos was 3.00 (3.00, 4.00), the median mDISCERN score was 3.00 (3.00, 3.00), and the median JAMA score was 1.00 (1.00, 2.00). When stratified by platform, the median GQS score was 3.00 (3.00, 4.00) for both Bilibili and TikTok (Douyin), with no statistically significant difference between the two platforms (P = 0.990; Fig. 6A and Table 5). Similarly, no statistically significant difference in mDISCERN scores was observed between Bilibili and TikTok (Douyin), with both platforms showing a median score of 3.00 (3.00, 3.00; P = 0.148; Fig. 6B and Table 5). However, as shown in Fig. 6C and Table 5, JAMA scores differed significantly between platforms, although both platforms had the same median score of 1.00 (1.00, 2.00; P = 0.020), indicating a statistically significant difference in transparency and authorship-related indicators between Bilibili and TikTok (Douyin) videos.Fig. 6Platform-based comparison of video quality and reliability scores Platform-based comparison of video quality and reliability scores When analyzed by uploader type, significant differences were observed in video quality and reliability. As shown in Fig. 7A and Table 6, GQS scores differed significantly by uploader type (P < 0.001). HCP-uploaded videos had the highest median GQS score 4.00 (4.00, 5.00), followed by NPO-uploaded videos 4.00 (3.00, 4.00), while individual-user videos had the lowest score 3.00 (2.00, 3.00). For mDISCERN scores, no statistically significant difference was observed among uploader types (P = 0.557; Fig. 7B and Table 6), with NPO, HCPs, and individual users all showing a median score of 3.00 (3.00, 3.00). JAMA scores also differed significantly by uploader type (P < 0.001; Fig. 7C and Table 6), with higher median scores for NPO and HCP-uploaded videos both 2.00 (2.00, 2.00) than for individual-user videos 1.00 (1.00, 1.00).Fig. 7Comparison of video quality and reliability scores among video uploaders Comparison of video quality and reliability scores among video uploaders Observer agreement was further evaluated using the intraclass correlation coefficient (ICC). For GQS, the average-measure ICC was 0.784 (95% CI: 0.768–0.864). For mDISCERN, the average-measure ICC was 0.924 (95% CI: 0.900–0.941). For JAMA, the average-measure ICC was 0.956 (95% CI: 0.943–0.967). All ICC values were statistically significant (P < 0.001), indicating acceptable to excellent agreement and supporting the reliability of the video quality assessment process. Correlation between engagement metrics and quality scores Spearman correlation analysis was performed to explore the associations among engagement metrics and quality scores. As shown in Table 7, among the engagement metrics, likes were positively correlated with collections (P < 0.001), comments (P < 0.001), and shares (P < 0.001). Collections were also positively correlated with comments (P < 0.001) and shares (P < 0.001), while comments were positively correlated with shares (P < 0.001). Regarding the associations between engagement metrics and quality scores, GQS was only positively correlated with shares (P = 0.015), whereas mDISCERN was positively correlated with likes (P = 0.042), collections (P = 0.025), comments (P = 0.011), and shares (P = 0.004). JAMA was positively correlated only with shares (P = 0.045), but not with likes, collections, or comments. Among the quality scores, GQS was positively correlated with mDISCERN (P < 0.001) and JAMA (P < 0.001), and mDISCERN was also positively correlated with JAMA (P < 0.001).Table 7Correlation between engagement metrics and quality scoresLikesCollectionsCommentsSharesGQSmDISCERNJAMALikes1(0.000*)Collections0.947(0.000*)1(0.000*)Comments0.946(0.000*)0.878(0.000*)1(0.000*)Shares0.904(0.000*)0.917(0.000*)0.854(0.000*)1(0.000*)GQS0.031(0.643)0.112(0.097)0.013(0.846)0.163(0.015*)1(0.000*)mDISCERN0.137(0.042*)0.151(0.025*)0.172(0.011*)0.192(0.004*)0.52(0.000*)1(0.000*)JAMA0.059(0.386)0.073(0.278)0.011(0.868)0.136(0.045*)0.546(0.000*)0.477(0.000*)1(0.000*) Table 7Correlation between engagement metrics and quality scoresLikesCollectionsCommentsSharesGQSmDISCERNJAMALikes1(0.000*)Collections0.947(0.000*)1(0.000*)Comments0.946(0.000*)0.878(0.000*)1(0.000*)Shares0.904(0.000*)0.917(0.000*)0.854(0.000*)1(0.000*)GQS0.031(0.643)0.112(0.097)0.013(0.846)0.163(0.015*)1(0.000*)mDISCERN0.137(0.042*)0.151(0.025*)0.172(0.011*)0.192(0.004*)0.52(0.000*)1(0.000*)JAMA0.059(0.386)0.073(0.278)0.011(0.868)0.136(0.045*)0.546(0.000*)0.477(0.000*)1(0.000*) Correlation between engagement metrics and quality scores Discussion This study systematically evaluated snakebite envenomation–related videos on TikTok (Douyin) and Bilibili, providing a comprehensive comparison of engagement patterns, content distribution, creator profiles, information quality and reliability, language characteristics, and seasonal upload trends. Overall, these videos provided moderately useful health information, but the low overall JAMA score and limited coverage of prevention indicated persistent gaps in transparency-related indicators and preventive content. Most videos were presented in Mandarin Chinese and were predominantly uploaded during the summer season. Platform-level differences were mainly reflected in video duration, sharing activity, and selected content themes, whereas overall educational quality and reliability, as measured by GQS and mDISCERN, were broadly comparable between platforms. More importantly, videos uploaded by healthcare professionals and nonprofit organizations showed higher educational quality and transparency than those uploaded by individual users, although individual users constituted the dominant source of available content. These findings suggest that, although short-form video platforms have substantial potential for disseminating snakebite-related health information, further efforts are needed to improve the standardization, transparency, and scientific rigor of such content. Video duration and audience engagement Overall, Bilibili videos were significantly longer than TikTok (Douyin) videos (median: 311.00 vs 101.00, P < 0.001). This finding aligns with a knee osteoarthritis study: the median duration was 262 s on Bilibili versus 108 s on TikTok (Douyin), with a significant difference. Similarly, other studies suggest that Bilibili videos are generally longer [26–28]. The reason may be related to differences in content ecosystems and recommendation mechanisms between the two platforms [29]. TikTok is characterized by short-duration, high-frequency content dissemination, whereas Bilibili places greater emphasis on knowledge-oriented and in-depth content presentation [29]. Regarding audience engagement, no significant differences were observed between Bilibili and TikTok (Douyin) in likes, collections, or comments (all P > 0.05). However, TikTok (Douyin) videos had a significantly higher number of shares than Bilibili videos. This finding is partially consistent with previous studies on amblyopia and thyroid eye disease treatment videos, which reported higher audience engagement on TikTok than on Bilibili, including higher sharing activity [16, 30]. However, unlike these studies, the present study did not observe significant platform-level differences in likes, collections, or comments, suggesting that engagement patterns may vary according to disease topic and user information needs. A possible explanation is that snakebite envenomation is an acute and practically oriented health topic, for which timely first aid, appropriate health-seeking behavior, and access to definitive treatment are clinically important [16, 31]. Viewers may therefore be more inclined to share videos they perceive as useful for emergency awareness or first-aid reference, as previous studies have shown that health information sharing on social media is influenced by users’ attitudes, subjective norms, perceived usefulness, and social-sharing functions. In addition, sharing behavior may be shaped by platform-specific dissemination mechanisms and user interaction habits, as social media engagement has been associated with information quality, source credibility, perceived relevance, and platform trust [32]. Therefore, video duration alone should not be interpreted as the main determinant of audience engagement. Future snakebite-related health communication should focus on improving content clarity, practical value, and shareability rather than simply extending video duration. Content structure of videos The videos on both TikTok and Bilibili predominantly focus on etiology, clinical manifestations, and treatment, with limited coverage of epidemiology and prevention. This finding aligns with Liao [33], which indicates that videos emphasizing disease knowledge and treatment exhibit higher quality, where disease knowledge typically encompasses etiology and clinical manifestations. Similarly, a study on lung nodule videos reports [34] a comparable pattern, though it highlights treatment as the most frequently covered theme, contrasting with the balanced emphasis on etiology, clinical manifestations, and treatment in this study. This discrepancy likely arises from differing disease characteristics: lung nodules, often incidentally detected via imaging, are typically chronic or asymptomatic in early stages, prioritizing treatment options such as surgical resection or monitoring due to lower urgency for immediate action [35–37]. In contrast, snakebite envenomation represents an acute, life-threatening emergency necessitating rapid identification of clinical manifestations (e.g., swelling, pain, coagulopathy), accurate snake species differentiation, and urgent antivenom administration, rendering etiology and clinical manifestations equally critical alongside treatment [38–40]. The acute nature of snakebite envenomation demands comprehensive content to guide immediate decision-making, explaining the observed broader thematic coverage. Further analysis of the temporal distribution of video uploads revealed that most videos were published in summer, with relatively fewer releases in spring, autumn, and winter. This seasonal pattern is consistent with the epidemiological characteristics of snakebite envenomation, which peaks during warm and rainy periods [41, 42]. Rising temperatures increase snake activity and human exposure through outdoor and agricultural activities, thereby elevating the risk of envenomation [43]. The increase in video uploads during summer may reflect content creators’ responsiveness to heightened risk or a rise in public demand for related health information during peak seasons. However, although snakebite exhibits seasonal characteristics, it can occur year-round in certain tropical and subtropical regions [42, 44]. A reduction in video uploads during non-peak seasons may limit the public’s continuous access to preventive knowledge [45]. Epidemiology and prevention are equally essential for snakebite envenomation due to its high prevalence in specific geographic regions and preventable nature [38, 41, 46]. With an estimated 5.4 million cases annually, snakebite envenomation predominantly affects rural, low-resource areas such as sub-Saharan Africa, South Asia, and Latin America [46–48]. Understanding epidemiological patterns, including high-risk areas (e.g., agricultural fields, forests) and seasonal trends (e.g., rainy seasons), supports targeted education to reduce exposure risks [49, 50]. Preventive strategies, including wearing protective clothing, avoiding high-risk zones, and providing first-aid training, are crucial for lowering incidence rates [51, 52]. Therefore, we recommend strengthening anticipatory communication prior to peak seasons, while maintaining stable year-round information dissemination and standardized thematic coverage to ensure comprehensive content on acute conditions such as snakebite envenomation, thereby enhancing public health education and alleviating disease burden. Video quality and reliability This study assessed the overall quality, transparency, and reliability of snakebite-related short videos using GQS, JAMA, and mDISCERN, respectively. Videos uploaded by NPOs and HCPs achieved significantly higher GQS and JAMA scores than those uploaded by individual users (both P < 0.001), whereas mDISCERN scores did not differ significantly among uploader types (P = 0.557). This finding suggests that uploader background may be associated with the overall quality and information transparency of snakebite-related videos. This result is broadly consistent with previous studies of health-related short videos. For example, studies on liver cancer and coronary heart disease videos have shown that content uploaded by healthcare professionals or nonprofit organizations tends to have higher quality and transparency scores than content uploaded by individual users [53, 54]. One possible explanation is that NPOs and HCPs may have stronger medical knowledge, clinical experience, and health communication capacity, enabling them to provide more structured, accurate, and publicly useful information [55, 56]. In contrast, individual users may rely more heavily on personal experience or fragmented information [57], which could contribute to their relatively lower GQS and JAMA scores. However, the lack of significant differences in mDISCERN scores indicates that the advantage of professional uploaders was not evident in all dimensions of information reliability. This may partly reflect the acute and time-sensitive nature of snakebite envenomation [58, 59]. Because snakebite envenomation is an acute and potentially life-threatening condition, public-facing videos may tend to emphasize practical information, such as symptom recognition, first aid, timely medical care, and access to antivenom [60, 61]. WHO emphasizes that improving antivenom availability, increasing prevention awareness among communities and health workers, and ensuring early administration of effective antivenom are key measures for reducing severe outcomes and deaths from snakebite envenomation [62]. Similarly, Afroz et al. reported that survival after snakebite envenoming is significantly improved by rapid first aid and the administration of life-saving antivenoms, further supporting the need for practical and immediately actionable public education [59]. In line with this, Guerrero Vargas et al. designed a massive open online course on ophidic accidents as an accessible, scalable, and dynamic training strategy to promote prevention and knowledge dissemination, further highlighting the value of structured educational resources for snakebite-related public health communication [63]. Accordingly, snakebite-related public education videos may prioritize immediately actionable messages—such as recognizing envenomation, applying appropriate first aid, seeking timely medical care, and accessing antivenom—over reliability-related features, including explicit source attribution, additional references, balanced presentation, and acknowledgement of uncertainty. Because these latter elements are central to mDISCERN-based assessment, their inconsistent presentation across uploader types may partly explain the lack of significant differences in mDISCERN scores. Therefore, short-video platforms should consider strengthening uploader verification mechanisms and prioritizing content produced by NPOs, HCPs, and qualified public health organizations. Meanwhile, creators should be encouraged to provide traceable references or source information where feasible to improve video reliability and public trust. Relationship between audience engagement and content quality The present study revealed distinct patterns in the relationships among audience engagement indicators and content quality. Overall, the engagement metrics, including likes, collections, comments, and shares, were significantly interrelated. The associations between engagement metrics and content quality scores were selective: shares were positively correlated with GQS, mDISCERN, and JAMA, whereas likes, collections, and comments were positively correlated only with mDISCERN. In addition, the three quality assessment scores were positively correlated with each other. Firstly, there were significant and strong positive correlations among the four audience engagement indicators (P < 0.001). This is consistent with a study on knee osteoarthritis, which found that engagement indicators were interrelated [64]. This finding indicates that when a video receives a high number of one type of interaction (such as likes), it is also likely to receive more of other forms of engagement (such as comments, shares, or collections). This suggests that user engagement behaviors are not isolated but instead exhibit a co-varying pattern. The reason may be attributed to the urgency and high perceived risk associated with snakebite, as emotionally intense or life-threatening content can simultaneously stimulate multiple forms of engagement [65]. Furthermore, videos that provide practical first-aid guidance or contain dramatic and controversial elements may enhance both perceived usefulness and algorithm-driven exposure, thereby promoting co-varying engagement behaviors [66, 67]. Secondly, engagement metrics were not uniformly associated with video quality scores. Specifically, both GQS and JAMA were positively correlated only with the number of shares, whereas mDISCERN was positively correlated with likes, collections, comments, and shares. This result suggests that general engagement metrics should not be directly equated with the educational quality or information transparency of health-related videos [68]. Likes, comments, and collections may reflect user engagement, emotional responses, perceived usefulness, or credibility cues, rather than serving as reliable indicators of the accuracy, completeness, or reliability of health information [69]. Therefore, the positive correlations between shares and both GQS and JAMA may suggest that videos with higher quality or greater transparency are more likely to be perceived by users as having dissemination value. For an acute condition such as snakebite envenomation, videos that provide clear first-aid guidance, timely medical advice, and trustworthy information may be more likely to be shared [58]. In addition, the positive correlations between mDISCERN and multiple engagement metrics may be explained by the fact that mDISCERN focuses more on information clarity, relevance, reliability, and balance, which are features that general viewers may perceive more directly [70]. Thirdly, the present study found significant positive correlations among GQS, mDISCERN, and JAMA scores, indicating that the three assessment tools showed good internal consistency in evaluating snakebite envenomation-related videos. Although these tools emphasize different aspects of video quality, they appear to capture overlapping dimensions of high-quality health information, including educational value, reliability, completeness, and transparency. This result is consistent with the findings of Tarihci [71]. Possible reasons include that although GQS, mDISCERN, and JAMA differ in their assessment focus, they all reflect the completeness, standardization, and credibility of health information from different perspectives. Specifically, GQS places greater emphasis on the overall quality, information flow, and educational value of videos; mDISCERN focuses more on content reliability, balance, and information sources; whereas JAMA mainly emphasizes authorship, source attribution, currency, and disclosure of conflicts of interest [72–74]. Therefore, when a video has a clear content structure, relatively complete medical information, and reliable information sources, it is likely to receive higher scores across all three assessment systems. Study limitations This study has several noteworthy limitations. First, a total of 220 videos from TikTok (Douyin) and Bilibili were included, and the relatively limited sample size may not fully capture the diversity of snakebite envenomation-related content available on these platforms, thereby restricting the generalizability of the findings. In addition, selecting videos based on the default ranking algorithms may have excluded high-quality content produced by less prominent creators, potentially resulting in a sample biased toward highly engaged but not necessarily representative materials. Second, although three commonly used evaluation tools (GQS, mDISCERN, and JAMA) were employed, mDISCERN and JAMA were not originally developed specifically for video-based health information or short-video platforms. Therefore, the results derived from these tools should be interpreted as supplementary indicators of information quality, reliability, and transparency rather than as definitive video-specific assessments. Moreover, the assessment process inherently involved subjective judgment. Despite the implementation of a double-blind review and third-party adjudication, minor discrepancies in scoring may still have occurred. Third, as a cross-sectional analysis conducted on February 17, 2026, this study could not capture the dynamic changes in quality, reliability, and transparency of snakebite-related videos over time. Conclusion This study found that snakebite envenomation-related videos on Douyin and Bilibili provide moderately useful health information, but their educational value remains limited by incomplete preventive coverage, insufficient information transparency, and weak alignment between audience engagement and content reliability. Videos uploaded by healthcare professionals and nonprofit organizations were generally of higher quality than those uploaded by individual users, highlighting the importance of professional participation in digital health communication. Given the acute and potentially fatal nature of snakebite envenomation, future efforts should prioritize evidence-based, platform-adapted, and seasonally responsive communication strategies. Strengthening uploader verification, improving source transparency, standardizing content domains, and promoting reliable prevention and first-aid messages may enhance the public health value of social media-based snakebite education. Supplementary Information Supplementary material 1. Supplementary material 1.