Quality and Reliability of Liver Cancer–Related Short Chinese Videos on TikTok and Bilibili: Cross-Sectional Content Analysis Study (Preprint)

Abstract

BACKGROUND China is a large country with hepatitis B. The incidence of liver cancer is increasing and people are becoming more and more concerned about liver cancer. Short video as a convenient source of acquiring health information is gaining more and more popularity, but the quality of short videos is not well evaluated. OBJECTIVE Our study aimed to assess the quality of information in videos associated with liver cancer on Chinese short video sharing platforms (TikTok and Bilibili ). METHODS In March 2023, we retrieved and screened the top 100 videos both on TikTok and Bilibili respectively and extracted the basic information of the videos. Later，we divided the videos into 2 groups according to different sources and different content, each with subgroups, and applied two rating tools, Global Quality Score (GQS) and DISCERN, to evaluate the quality and reliability of the videos. And then we further performed correlation analysis. RESULTS According to 166 videos, doctors uploaded most videos, accounting for 53.9%(89/166) and news and reports contributed the most(42.8%,71/166). Individuals tend to have longer videos compared with news agencies and doctors (p<.0001and p<.001, respectively). The median GQS and DISCERN scores of all 166 videos were 2.5(1,4) and 4(1,6), which indicated most videos had poor quality with only 28% (47/166) of the videos considered excellent and good, and also only 29% (48/166) of the videos considered reliable or relatively reliable. The total GQS scores of doctors were significantly higher than that of individuals(p<.001), disease knowledge higher than treatment and news and reports (p=0.04 and p<.0001, respectively) and prevention higher than news and reports(p<.001). DISCERN score for doctors as well as news agencies were significantly higher than individuals (p<.0001 and p=.03, respectively), prevention higher than treatment and news and reports (p=.03 and p<.0001, respectively), so as to videos of disease knowledge (p=.03 and p<.0001, respectively). No significant differences were found between groups based on different type of doctors. Negative correlation was observed between video durations and likes (r=-.194, p=.012) Only shares had positive correlation with GQS and DISCERN scores(r=0.22, p<.01 and r=0.203, p<.01, respectively). CONCLUSIONS Short videos can be considered a poor quality source of health information related to liver cancer, most of which are posted by doctors. People's preference for videos is negatively correlated with the duration of the video, with people preferring videos that are practical and entertaining. People need to be careful when choosing videos to obtain medical content.