Thematic patterns of social media content by patient-led breast cancer key opinion leaders: A topic modeling study

To investigate the thematic focus of patient-led breast cancer key opinion leaders (PBKOLs) on Xiaohongshu by analyzing their postings and identifying underlying narrative patterns.

PBKOL accounts were identified through systematic keyword searches on Xiaohongshu. Their publicly available text-image and video posts were retrieved, transcribed, and standardized. Latent Dirichlet Allocation (LDA) was used to extract latent themes, and subthemes were inductively classified to capture narrative nuances.

A total of 65 PBKOLs contributed 6852 valid posts between December 2017 and July 2025. Four core themes were identified: recovery life and functional self-management (13.60%), medical decision-making and treatment journey (26.80%), emotional adjustment and meaning-making (31.30%), and family routine and relational support (28.30%). Emotional narratives predominated, while family-oriented posts emphasized relational support and treatment-related content reflected practical but less frequent engagement. PBKOLs varied widely in influence, ranging from grassroots users to highly followed opinion leaders.

PBKOLs shape breast cancer health communication by prioritizing emotional resonance and community support over professional instruction. While their narratives enhance trust, empowerment, and cultural relevance, concerns remain regarding scientific accuracy. Integrating patient narratives with evidence-based guidance may help develop more balanced and sustainable digital health strategies.