Human Communication Research ISSN 0360-3989ORIGINAL ARTICLEA Meta-Analysis of Research on FormalComputer-Mediated Support Groups:Examining Group Characteristics andHealth OutcomesStephen A. Rains & Valerie YoungDepartment of Communication, University of Arizona, Tucson, AZ 85721-0025This article reports a meta-analysis of 28 studies examining the health-related outcomesassociated with participation in a formal computer-mediated support group (CMSG)intervention. In particular, health outcomes related to social support were assessed and fourgroup-level characteristics of CMSGs were tested as potential moderators of interventioneffectiveness. The results show that participating in a CMSG intervention—comprisedof educational and group communication components—led to increased social support,decreased depression, increased quality of life, and increased self-efficacy to manage one’shealth condition. Changes in health outcomes were moderated by group size, the nature ofthe communication channels available, and the duration of the CMSG intervention.doi:10.1111/j.1468-2958.2009.01353.xThe benefits of group interaction have been fairly well recognized among scholarsand practitioners of health communication. Group-based health interventions suchas support groups, self-help groups, and group psychotherapy allow members togive and receive social support from others who are, or have been, afflicted witha particular health condition (Arntson & Droge, 1987). Group interventions alsoprovide an outlet for members to make social comparisons (Gottlieb, 1988), witnessthe modeling of behaviors such as coping (Posluszny, Hyman, & Baum, 1998), anddemystify their experiences with a health condition (Levy, 1979). The results of atleast two meta-analyses indicate that group interventions conducted face-to-facecan lead to health improvements for participants (Barlow, Burlingame, Nebeker, &Anderson, 1999; Burlingame, Fuhriman, & Mosier, 2003).In recent years, the widespread diffusion of personal computers and the Internethas been met with the development of I nternet-based group health interventions.Computer-mediated support group (CMSG) programs are delivered by medicalCorresponding author: Stephen A. Rains; e-mail: [email protected] Communication Research 35 (2009) 309–336c2009 International Communication Association309Computer-Mediated Support Groups S. A. Rains & V. Youngprofessionals using the Internet and consist of an educational component a nda group communication component. For example, the Comprehensive HealthEnhancement Support System (CHESS) is an established computer-based groupintervention that can be tailored t o specific medical conditions such as breastcancer (Gustafson et al., 1998, 2001), smoking cessation (Meis et al., 2002),and asthma (Wise et al., 2007). CHESS makes it possible to provide groupmembers with access to medical information, decision support, skills training,and an asynchronous discussion forum where they can share information,strategies for coping, and support. CMSG interventions are distinct from theinformal and loosely structured self-help groups found on Websites such asWebMD.com and Yahoo.com in that formal CMSGs have both educational andgroup communication components, closed membership enrollment, fixed duration,and may include some form of expert leadership (Gottlieb, 2000; Helgeson &Gottlieb, 2000).Group communication is integral to CMSG interventions as a resource formembers to exchange social support. Interacting with others who are suffering,or have suffered, from a specific health condition offers a multitude of potentialopportunities to g ive and receive information, emotional, and esteem support. Suchsupport may serve to buffer individuals’ experience of stressful events (Cohen &Wills, 1985), manage health-related uncertainty (Albrecth & Goldsmith, 2003), and,ultimately, foster positive health outcomes (Uchino, Cacioppo, & Kiecolt-Glaser,1996). Further, the computer-mediated nature of CMSGs offers several distinctadvantages that may facilitate supportive c ommunication (Walther & Boyd, 2002;Walther, Pingree, Hawkins, & Buller, 2005; Wright & Bell, 2003). Walther and Parks(2002) succinctly summarize the potential of computer-mediated communicationusing the Internet noting that: ‘‘The question for researchers is not whether theInternet is capable of providing social support, but rather why it should be soeffective as a support medium’’ (p. 545).Despite the gr owing number of studies examining the health outcomes associatedwith participation in CMSG programs, the findings from this body of research areinconsistent. To help interpret scholarship on C MSGs, we conducted a meta-analysisfocusing on the role of group communication and social support in CMSGs. Meta-analysis is a useful tool to summarize a body of research and reconcile mixed finding s(Hedges & Olkin, 1985; Hunter, Schmidt, & Jackson, 1982). The primary goal of theanalysis was to assess the influence of participating in a CMSG intervention—inwhich group communication is a key component—on perceptions of social support,depression, quality of life, and self-efficacy to manage one’s health condition.A second goal of this meta-analysis was to evaluate group characteristics that maymake CMSG interventions more or less effective at fostering the exchange of socialsupport and positive health outcomes. Drawing from research on characteristicsof social networks, we examined group size, communication channels, frequencyof contact, and time as moderators of the health benefits associated with CMSGparticipation.310 Human Communication Research 35 (2009) 309–336c2009 International Communication AssociationS. A. Rains & V. Young Computer-Mediated Support GroupsLiterature reviewSocial support and CMSGsCMSGs and informal Web-based self-help groups have received a fair amount ofattention by researchers in various academic fields. Yet, the theoretical mechanismsthat explain their functions and outcomes as they relate to social support are notwell understood. As Wright and Bell (2003) explain: ‘‘The majority of studies thathave examined [computer-mediated support] groups over the past several years havebeen descriptive in nature, and they have not linked findings to a broader theoreticalframework, despite a vast amount of theory in both the areas of social supportand c omputer-mediated communication’’ (p. 49). Thus, we first consider theorizingabout the health outcomes