Characterizing development patterns of health-care social networks

Abstract

The growing amount of data in health-care social media requires innovative new analysis methods, which are elementary to exploration of relationship dynamics, in a bid to understand the new roles social media plays in health care. In this work, we use network analysis to explore the temporal nature of two large diabetes social networks, with a view to enhancing our knowledge of the development of community structures and cohesion factors. We compare our findings with analysis of two other nonhealth-care social networks. Current results reveal how diabetes online communities are very dynamic, suggesting diabetes patients are usually actively engaged for periods of less than a year, typically immediately following diagnosis. Additionally, we observe shrinking of both diameter and density, as well as disassortative mixing. The presented empirical study informs future online intervention strategies for promoting health behavior and lifestyle changes among people with diabetes.