Can peer support be effective online?

Research and development on social media and online technology for health has exploded in the past decade. Online services have been used as a resource for information about illness and treatment for that illness. They connect consumers to real-world resources where treatments can be obtained. Online platforms may be especially useful for people with serious mental illness given the special needs that arise from their disabilities; they are often disabled by cognitive difficulties that significantly undermine personal goals. Online services may help people with cognitive disabilities (Medalia & Revheim, 1999). Programs have included rules and strategies of cognitive behaviour therapy to help users manage the depression and anxiety that emerge from irrational thoughts. Similar platforms set up decision trees that help people consider the pros and cons of specific goals; e.g., ‘what are the advantages and disadvantages of returning to college to finish my accounting degree?’ Online programs can serve as cognitive ‘prostheses,’ external processes that help to manage and organise one's memory.

Naslund and colleagues (this issue) expand the potential of social media to an equally important realm of psychiatric rehabilitation, peer support. I briefly review the key ingredients of peer support for people with serious mental illness and then review the ways these authors believe that online platforms help people accomplish these goals. Next, I contrast both benefits and limitations of online services for peer support. Finally, I expand on research directions outlined by Naslund and colleagues, suggesting questions and corresponding methods that should be added to make sense of this important area.

Peer support is often defined as combinations of emotional and instrumental support provided by individuals with lived experience who come together with the specific intent of bringing about social and personal change (Davidson, 2013). Peer support is mutually beneficial through a reciprocal process of giving and receiving based on principles of respect and shared responsibility. Feelings of rejection, discrimination, frustration and loneliness are combated through this system of sharing, supporting and assisting others. Peer support provides a venue for personal empowerment and self-determination, two processes key to recovery. Copeland & Mead (2004) distinguished peer from clinical support. Peers understand each other because they have ‘been there,’ shared similar experiences and model a willingness to learn and grow. Peers assemble in order to change unhelpful patterns, get out of ‘stuck’ places and build relationships that are respectful, mutually responsible and mutually transforming.

Naslund and colleagues list several general benefits to online platforms – overcomes geographic and time boundaries, represents user-driven environments and allows anonymity – as well as benefits that are specific to the needs of people with serious mental illness: helps them manage interpersonal skills and emotions that might typically be overwhelming in face-to-face encounters. They nicely contrast benefits to limitations of online services. Among others, they note that online dependencies may cause problems that challenge face-to-face communications. This one concern echoes the one, which I think is a central research questions going forward. How does one compare online peer support with in vivo interactions? How do online services expand the potential of support services? Where does social media fall short compared with face-to-face interaction? This likely will reflect a cohort effect. Younger people, immersed in social media, may accept online services more, thus leading to greater benefits. Still, this work needs to be framed in light of a burgeoning research on adolescents that tempers benefits of social media with such concerns as social isolation and depression (Best et al. 2014).

Two research principles need to be considered as investigators examine peer support in the online sphere, principles relevant to the personal empowerment dimension of this support. (1) Work to adapt and evaluate online platforms needs to take the form of community-based participatory research (CBPR) (Minkler & Wallerstein, 2003). According to CBPR, research has its greatest impact when it relies on partnership among the community affected by the phenomenon of concern and experts in methods and analyses. CBPR is especially important for disenfranchised groups – people of colour, those with low income, people with serious mental illness – who are traditionally left out of the kind of social decision-making that leads to health policy. People with serious mental illness participating in CBPR are full partners in the research enterprise and not relegated to being subjects of study. Research partners with mental illness are especially important to CBPR coming into and going out of a research enterprise. Consider CBPR start-up and draw down for research examining online peer support. Research partners with mental illness are essential to beginning research, when questions and hypotheses are developed. People with mental illness, especially those who have participated in peer support and/or online platforms, have significant insights on the approaches, likely to exceed what can be gleaned from written resources. Research partners are also importantly coming out of a study. Researchers typically wrap up projects in neatly summarised reports and journal articles. People with lived experience are likely to use findings to improve the status quo of peer support. (2) Research needs to strategically incorporate treatment preference; i.e., in expressing personal empowerment, people will have predilections towards, or away from, a specific intervention like peer support (Corrigan & Salzer, 2003). Randomized controlled trials (RCT) remain the gold standard in services research despite concerns expressed by methodologists and advocates. Among these concerns is the expectation recognising that research participants approach services research without significant preferences for intervention – e.g., anti-depressants v. cognitive behaviour therapy – and so are likely to engage in the intervention to which they are randomly assigned. Such preference is essential to peer support; namely, the empowered individuals recognise the importance of mutual support and commit to participating. People randomized to peer support when they do not cherish it are unlikely to benefit, whether it be online or in vivo. One alternative to RCTs gaining traction in patient-centred outcomes research is partially randomized preference trials. In this design, the research participant's preference for an intervention is assessed at entry into the study. Those who prefer one of the treatments would then be assigned to that treatment. Those who report no preference for treatment would be randomized to the research arms in the investigation. If everyone has a strong preference for treatment, the study essentially tests feasibility. If all research participants expressed no preference, a randomized clinical trial remains.

The point here is to realise the complexity of the research agenda that emerges from Naslund and colleagues. Elegant questions need to correspond with sophisticated methods and designs.