Mental Health Pathways in Internet Support Groups

Carnegie Mellon University

Status: Completed

Conditions: Depression; Anxiety; Stress

Treatments: Behavioral: Order by time and topic; Behavioral: Order by social relationship; Behavioral: Order by self-disclosure; Behavioral: Order by information relevance; Behavioral: Order by help giving

Study type: Interventional

Funder types: Other; NIH

Identifiers

NCT02396472

CarnegieMU

Details and patient eligibility

About

People facing serious health threats increasingly use Internet health support communities to obtain informational support, emotional support and other resources. This study introduces software algorithms similar to those used by social media sites to put people in touch with helpful information and social interactions. Participants from the American Cancer Society's Cancer Support Network will have access to this online support group using the default interface that orders content by broad content category and date or with a new interface that highlights communication content and people that match users' interests and needs.

Full description

Internet support groups (ISGs) are online communities where people come together to exchange information, emotional support and other resources. They are an important resource for patients grappling with serious medical conditions. Although participation in health-related ISGs has been associated with significant reductions in participant-reported depression, anxiety and other indicators of psychological distress, many ISG members leave too soon to benefit. In a parallel study, we are using state-of-the art machine learning and automated language analysis techniques to assess the types of interactions that keep people participating in these groups and that lead to improved psychosocial well-being and health quality of life and how these interactions develop. The clinical trial described here uses these technologies and insights from our empirical research to build, deploy, and evaluate interventions that improve the interactions in Internet health support groups.

We will develop and pilot-test interventions to encourage effective communication processes identified in our empirical research. Participants from the American Cancer Society's Cancer Support Network will access this support group using either the default interface that orders content by disease diagnosis and date or with a new interface that sometimes highlights communication content and people who match their interests and needs. 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 compatible peers. In a series of small, randomized experiments, we will examine how these interventions affect participants' communication behaviors as well as short-term engagement and satisfaction with their online interactions.

Enrollment: 6,500 patients

Sex: All

Ages: 18+ years old

Volunteers: Accepts Healthy Volunteers

Inclusion criteria

• Is an existing or new registrant to the American Cancer Society's Cancer Support Network.

Exclusion criteria

Trial design

Primary purpose: Supportive Care

Allocation: Randomized

Interventional model: Single Group Assignment

Masking: Double Blind

---

6,500 participants in 5 patient groups, including a placebo group

Order by time and topic

Placebo Comparator group

Description:

Volunteers from the American Cancer Society's Cancer Survivors' Network (CSN) will see some of their messages delivered using CSN's default ordering, which shows messages within a conversational thread ordered by time stamp. Conversational threads are nested within a broad topic-based forum, like breast cancer or colorectal cancer survivors. Note that this is a within-participant trial, so that all participants participate in all arms of the trial. Messages, not people, are randomly assigned to condition.

Treatment:

Behavioral: Order by time and topic

Order by information relevance

Active Comparator group

Description:

In this condition some messages will be highlighted if they match the type of content the user has previously shown interest in, by previously contributing or reading semantically similar material.

Treatment:

Behavioral: Order by information relevance

Order by social relationship

Active Comparator group

Description:

In this condition some messages will be highlighted because they come from people the user has previously shown interest in, by previously reading their posts or communicating with them.

Treatment:

Behavioral: Order by social relationship

Order by help giving

Active Comparator group

Description:

In this condition some messages will be highlighted because they seek help and therefore provide an opportunity for participants to provide social support to others.

Treatment:

Behavioral: Order by help giving

Order by self-disclosure

Active Comparator group

Description:

In this condition some messages will be highlighted because in them the writer is self-disclosing, and they provide provide an opportunity for participants to self-disclose in return.

Treatment:

Behavioral: Order by self-disclosure