From perception to participation: how the sick role behaviour differs across treatment stages in diabetes online health communities

Purpose Online health communities (OHCs) have become essential platforms enabling patients to obtain social support while preserving anonymity. However, the influence of treatment stages on users’ behaviour within OHCs remains underexplored. This study delved into user participation at different treatment stages and the pivotal role of OHCs in providing support by integrating norms of sick role and community membership. Design/methodology/approach Sweet Homeland, a Chinese diabetes online community, was used as a case study. This research identified distinct treatment stages based on user-generated content. Leveraging machine learning and deep learning models, the authors classified posts including “remission”, “medicine” and “insulin”. Subsequently, using topic modelling, sentiment analysis, grounded theory and sequential pattern mining, the authors discerned user-generated topics, emotional expressions and information behaviours. Findings The findings illuminated the stage-specific differences in informational and emotional needs, as well as participation behaviours and interaction patterns among users. The authors also observed the phenomenon of emotional contagion within the community, which may underlie similar emotional changes. Originality/value This study provides evidence for extending the sick role concept to chronic illnesses and online environments, and the role adoption of social relationships. The authors also offer practical advice for designing and maintaining online diabetes communities through personalised service.