To evaluate the quality, reliability, content characteristics, and engagement patterns of Sjögren’s disease (SjD)-related videos on major Chinese short-form video platforms, and to examine how uploader identity, presentation format, and content orientation are associated with informational quality and user engagement. Given that SjD is a chronic systemic autoimmune disease that is frequently underrecognized and often reduced in public discourse to symptoms such as dry mouth and dry eyes, this study further aimed to assess whether short-form video content provides complete and systemically framed disease information, rather than fragmented symptom-based messages alone.

A cross-sectional content analysis was conducted using videos retrieved from three major Chinese platforms: TikTok, Bilibili, and rednote. Data collection was performed between March 5 and March 6, 2026. To reduce the influence of algorithmic personalization, all searches were conducted using newly registered accounts with no prior viewing or interaction history. The standardized Chinese keyword corresponding to “Sjögren’s disease” was used across all platforms. Videos were screened in descending order according to each platform’s default ranking mechanism until 100 eligible videos were obtained from each platform, yielding a total sample of 300 videos. Eligible videos were required to be in Chinese and primarily focused on SjD-related medical knowledge, diagnosis, treatment, disease management, or patient experience. Duplicate videos, irrelevant content, purely promotional material lacking educational value, and inaccessible videos were excluded. Two trained reviewers independently extracted data using a standardized coding sheet, with disagreements resolved by a third reviewer. Extracted variables included platform, video duration, upload date, likes, collections, comments, shares, uploader follower count, uploader type, presentation format, content orientation, and content completeness. Uploader type was categorized as professional physicians, non-professional physicians, institutions, or individual users. Presentation format was classified as direct-to-camera explanation, interview/dialogue, key-point text-based presentation, animation-based presentation, scenario-based dramatization, or news/documentary-style presentation. Content orientation was categorized as local symptom-oriented, systemic disease-oriented, or mixed. Content completeness was assessed across seven core domains: epidemiology, etiology, clinical manifestations, diagnosis, treatment, prevention, and prognosis. Each mentioned domain was scored as 1 point, yielding a Content Completeness Score (CCS) ranging from 0 to 7. Video quality and reliability were evaluated using four established instruments: Global Quality Score (GQS), modified DISCERN (mDISCERN), Journal of the American Medical Association (JAMA) benchmark criteria, and Video Information Quality Index (VIQI). Inter-rater reliability was assessed using intraclass correlation coefficients (ICCs) for quality scales and Cohen’s kappa for categorical variables. Continuous variables were summarized as medians and interquartile ranges, and categorical variables as frequencies and percentages. Kruskal–Wallis tests, chi-square/Fisher exact tests, Dunn post hoc tests, and Spearman correlation analyses were performed as appropriate. Statistical significance was defined as two-sided P < 0.05.

Among the 300 included videos, significant cross-platform differences were observed in video duration, engagement metrics, uploader follower counts, and quality scores. Bilibili videos were significantly longer than those on TikTok and rednote. TikTok videos received significantly more likes, comments, and shares, whereas rednote videos had the highest collection counts. Although statistically significant differences in GQS, mDISCERN, JAMA, and VIQI were identified across platforms, the magnitude of quality differences was generally modest. Overall, content completeness was limited. The median CCS was 2.00 (IQR 1.00–3.00), indicating that most videos covered only a small subset of the seven core content domains. Clinical manifestations were the most frequently addressed domain (68.6%), followed by treatment (43.3%) and diagnosis (36.0%). In contrast, epidemiology (14.3%), etiology (15.6%), prevention (11.6%), and prognosis (8.6%) were markedly underrepresented. Regarding content orientation, local symptom-oriented videos predominated, accounting for 68.6% of all videos, whereas systemic disease-oriented videos comprised only 16.0%, and mixed videos 15.3%. This pattern suggested that SjD was more often presented as a symptom-based condition than as a systemic autoimmune disease. Uploader identity and content orientation were more consistently associated with video quality than presentation format. Videos uploaded by professional physicians generally achieved higher GQS, mDISCERN, and VIQI scores than those uploaded by individual users. Similarly, systemic disease-oriented videos consistently demonstrated higher GQS, mDISCERN, and VIQI scores than local symptom-oriented videos. By contrast, presentation format showed limited influence on quality, with no statistically significant pairwise differences across most formats. User engagement did not align closely with informational quality. Although some higher-quality videos achieved favorable engagement, the associations between quality indicators and engagement metrics were generally weak. CCS was more strongly correlated with quality indicators than with engagement outcomes, suggesting that videos with broader and more systemically framed content tended to be more informative, but not necessarily more visible or interactive. Scenario-based dramatization tended to generate relatively high engagement despite limited differences in informational quality across presentation formats. Systemic disease-oriented videos showed better quality than local symptom-oriented videos, but this advantage translated only weakly into engagement, with collections being the only metric that differed significantly across content orientations.

SjD-related videos on Chinese short-form video platforms were characterized by limited content completeness and a marked predominance of symptom-oriented communication. Although professional physician–uploaded videos and systemic disease-oriented videos generally provided higher-quality and more reliable information, these advantages did not necessarily result in stronger user engagement. The findings suggest that, in short-form video environments, SjD is more likely to be communicated as a collection of familiar dryness-related complaints than as a systemic autoimmune disease requiring broader clinical understanding. Future digital health communication on SjD should therefore move beyond isolated symptom descriptions and better integrate medical accuracy, systemic disease framing, and platform-adapted presentation in order to support more accurate disease recognition and public understanding.