People facing serious health threats increasingly use the Internet to obtain information and support. Internet support groups (ISGs), online communities where people come together to exchange information and/or support, are an important resource for patients grappling with serious medical conditions. Participation in health-related ISGs has been associated with significant reductions in participant-reported depression, anxiety and other indicators of psychological distress. These groups offer great potential for interventions that are highly scalable and yet easily tailored to individual needs. Unfortunately many ISG members leave too soon to benefit. In addition, there is limited scientific understanding of the active ingredients underlying the benefits that ISGs confer among their members. The proposed research uses state-of-the art machine learning and automated language analysis techniques to assess the types of interactions in online health support groups that lead to improved psychosocial well-being and health-related quality of life and how these interactions develop. It also uses these technologies and insight from our empirical research to build, deploy, and evaluate interventions that improve the interactions in health-related Internet support groups. Our work has three major aims.
In Aims 1 and 2, we will analyze data from two complementary Internet support groups to identify (a) the communication processes that sustain participant engagement in the group and (b) the communication processes through which peer-to-peer interactions improve participants' psychosocial outcomes. These datasets come from an ongoing naturalistic study of 4,400 cancer survivors and caregivers in ISGs hosted by the American Cancer Society (ACS) and from a soon-to-be-completed NIMH-funded randomized trial (MH093501) assessing the effectiveness of an ISG for depressed and anxious patients referred to the group by their primary care physicians. Both datasets include participant-level data on depression, anxiety, stress and other psychosocial outcomes and detailed server logs containing information on page views, time on site, and the content of communications between participants.
In Aim 3, we will develop and pilot test interventions to encourage the communication processes identified as effective in Aims 1 and 2. We will test whether mood, satisfaction with interactions and engagement in the group increase following interventions that (a) increase participants' receipt of individualized support from others; (b) provide participants with opportunities to offer support to others; (c) facilitate participants' expression of emotions; and (d) help participants form relationships with similar peers. In a series of randomized experiments, we will examine how these interventions affect participants' communication behaviors as well as short-term engagement and satisfaction with their online interactions. Results from this project will facilitate the development of an R01 application for a randomized clinical trial testing the impact of personalized ISG interventions on the length of time participants continue to participate in ISGs and participants' depression, anxiety and other psychosocial outcomes as well as healthcare utilization.
The proposed research will examine how social exchanges among members of online health support groups lead to improvements in participants' health-related quality of life. Our findings will clarify the mechanisms through which social support leads to improved health-related quality of life and the process by which support resources are produced. The automated techniques we use to measure the exchange of support resources will also be the basis for interventions designed to make these exchanges more efficient and effective by matching people seeking support with those who can offer the type of support they need.
