Quality and reliability of leukaemia-related short video content on Bilibili and TikTok: a cross-sectional study

Background Short‑video platforms have become a primary source of health information in China, yet the quality and reliability of disease‑related content remain largely unregulated. Numerous cross‑sectional studies have revealed suboptimal quality, significant platform disparities, and the superiority of professional‑generated content. However, haematological malignancies, particularly leukaemia, have been conspicuously overlooked, and the mechanisms underlying these patterns, as well as the clinical impact of misinformation, remain underexplored. Leukaemia is the most common childhood malignancy in China, with an incidence of 42.33 per million in children aged 0–14 years, and its treatment encompasses chemotherapy, targeted therapy, immunotherapy, radiotherapy, and stem cell transplantation. This study aimed to assess the quality and credibility of leukaemia‑related short videos on TikTok and Bilibili. Methods A cross‑sectional content analysis was conducted in November 2025. Using the Chinese search term “leukaemia,” 232 videos were retrieved from TikTok ( n = 123) and Bilibili ( n = 109). Two independent reviewers evaluated video quality and reliability using the GQS and mDISCERN. Content themes were categorised across seven dimensions, and engagement metrics were recorded. Results Overall quality and reliability were suboptimal (median GQS and mDISCERN: 2.00, IQR: 2.00–3.00). “Symptoms” was the most prevalent topic (27.13%), while “epidemiology” (4.33%) and “prevention” (6.06%) were severely underrepresented. Bilibili videos had significantly higher GQS scores than TikTok ( p < 0.05), whereas TikTok demonstrated substantially higher user engagement. Specialist‑generated content received the highest GQS scores ( p < 0.05), whereas mDISCERN scores did not differ significantly across uploader groups. No significant correlation was found between engagement and quality. Conclusions Leukaemia‑related short‑video content in China shows suboptimal quality and reliability, with critical gaps in epidemiology, prevention, and diagnosis. To address these deficiencies, we propose actionable measures: professionals should systematically cover essential domains; platforms should implement creator verification, algorithmic weighting, and triage review; warning labels for misleading content; and hospital‑run official accounts for quality‑assured education. These findings underscore the need for enhanced content governance and professional engagement in digital health communication.