Type 1 diabetes is a chronic, incurable, and life-threatening metabolic illness. Management demands of the condition are complex and intensive, and clinical management requires a specialist multidisciplinary team.1,2
Despite the support of health providers, the daily management tasks of type 1 diabetes are all actions that people with diabetes must carry out themselves to prevent acute and chronic physical complications. This responsibility may bring with it significant psychological strain.3 Furthermore, although health outcomes are controllable to a significant extent, fluctuations in health status are inevitable, and complications can befall even the most diligent illness managers. The resulting uncertainty and knowledge of the potential for death or disability can compound the psychological strain.4 There can also be a tangible burden associated with the considerable time, energy, and focus required to establish and maintain metabolic control, which places considerable strain on patients' ability to attend to occupational, social, familial, and personal needs and desires.3,5–8 Social stigma may also be associated with carrying out the demands of diabetes management.
Together, these experiences often leave individuals with diabetes feeling isolated from members of their immediate social network who may fail to understand their daily experiences. Seeking connection with others with the same medical condition is one strategy that people in this situation pursue, and the resulting process has been termed “peer support.”9
Peer support has been defined as assistance by a person who has experiential knowledge of the specific health issue and similar characteristics as the target population.10,11 Fisher12 describes peer support's role in health behavior change as assisting people with chronic disease to translate a plan developed with or by a health professional into action.
Peer support can be delivered in many different forms. As theories for implementing any complex interventions state, it will have a “hard core” of functions that it delivers and a “soft periphery” of elements that are adapted through the implementation according to local and individual needs.13 Boothroyd and Fisher,14 in describing the work of Peers for Progress (a global initiative of the American Family Physicians Foundation), similarly propose that the way in which peer support programs are delivered will always need to be “shaped by cultural, organizational, and national factors,” but that there are four core functions that underpin all peer support programs:
Assistance in applying disease management in daily life. Peer supporters use their own experiences with nutrition, physical activity, and medication adherence in helping people figure out how to manage diabetes in their daily lives. They can also help in identifying key resources such as where to buy healthy foods or pleasant and convenient locations for exercise.
Emotional and social support. Through empathetic listening and encouragement, peer supporters are an integral part of helping patients cope with social or emotional barriers and stay motivated to reach their goals.
Linkage to clinical care. Peer supporters can help bridge the gap between patients and primary care providers and encourage individuals to seek out clinical care when appropriate.
The evidence for peer support's effectiveness in improving clinical and psychological outcomes is growing.11,16–18 A recent systematic review19 found 40 studies in eight countries, of which 15 showed modest benefits and 18 showed strong evidence of benefit. The review concluded that “peer support is effective in promoting complex health behaviors across a variety of diseases and national settings.”19
Some studies comparing the effects of volunteer-led peer support with that of support and education led by health professionals have shown similar or better outcomes for the peer-support interventions on both clinical and psychosocial outcomes. However, peer support has been predominantly studied and defined as a process that is complementary, not an alternative to, interventions led by health professionals.17,20
Health professionals' perspectives of volunteer-led or online peer support are not well understood. There is some evidence that some health professional groups value the complementary role that peer supporters may play in supporting chronic condition self-management.21,22 There is also some evidence that peer supporters value interaction with the medical team.23 However, peer support is not currently recognized as a core part of health service delivery by health professional groups in most countries.2,17,24,25
Research and evaluation of peer-support programs that are volunteer-led or online are both extremely rare.16,17 An online community group for women with breast cancer found that the group increased coping skills and improved psychological well-being.28 Lorig et al.29 conducted a randomized, controlled trial of a 6-week self-management program involving peer facilitators and peer discussions. They found significant improvements in health status at 1 year and showed that improvements were similar to those achieved in a face-to-face version of the program.
Our study aimed to identify what key stakeholder groups (health professionals, people with diabetes, and both users of the program and observers) most value about the Reality Check online community. Øvretveit30 argues that “value explication” is an important part of evaluation. Where the values relate to benefits, impacts, or outcomes of a program, they can be described as “success values.” Success values can be determined through what has been measured in other research studies on the topic if the primary user of the evaluation is the scientific/research community; otherwise, they should be specifically investigated with relevant stakeholders.31
The second aim of our study was to contribute the experiences from this long-running and self-sustaining online program to the growing field of peer-support research. Identifying what users value most could inform the selection of outcomes that are most important to measure in summative evaluations of this and other peer-support programs.
The study involved conducting online focus groups with both users and observers of an online community of people with type 1 diabetes and conducting an online survey of health professionals.
The University of Melbourne's Population Health Human Ethics Advisory Group granted approval for this project on 3 June 2010 (Ethics ID 1033676).
Reality Check online community
Reality Check is an Australian open-access online community (located online at www.realitycheck.org.au), where people with type 1 diabetes discuss topics of interest to them in real time. After a brief registration process requiring only a user name and an e-mail address, participants type a message on their computer that is then immediately published on the Web site, where others can similarly log in and post responses to the message, creating discussions threads. Discussions can be read and observed on the Web site freely without registering or requiring a password. At the time of this study, the program was 13 years old and had 1,558 registered members.
The online community's purpose has been defined by its founders as facilitating sharing of personal experiences of living with diabetes. Notably, provision of medical advice is actively discouraged and removed if it occurs.
The program is an active forum. A 2010 content analysis showed that, in a 1-week snapshot, 75 unique participants posted 197 messages covering 23 different topics. The median time for a participant to receive a response was 1 hour and 14 minutes, and there was a median of 15 responses in each discussion.
Volunteer moderators, all people with type 1 diabetes themselves and longstanding members of the community, actively monitor the Web site. There is no formal training system for the volunteers, but they enforce a published set of rules. A weekly roster system is used to ensure that a moderator is monitoring the Web site at least daily. Moderation can involve removing inappropriate messages such as advertising or the provision of medical advice, contacting members by e-mail to explain the Web site's rules, and banning inappropriate participants. However, the most frequent activity of the moderators is to make proactive contributions to the community and model and reinforce appropriate behavior and discussion, such as welcoming new members, contributing their own experiences, and providing links to high-quality information resources. Its founders describe Reality Check as an online peer-support program.27
Online focus groups
Online focus groups were conducted with users of the Reality Check online community. Conducting focus groups online had practical benefits in that the target population was geographically dispersed and known to be active online. Methodological benefits were also identified. Walston and Lissitz32 studied online focus groups' effectiveness as a methodology in direct comparison with in-person focus groups, concluding that they could be equally effective and that there may be added benefits of the online format, including that “the online environment may … discourage participants from withholding embarrassing information.” Stewart and Williams33 examined the use of online focus groups in qualitative research, identifying among the benefits that participants can have time to reflect and form a more considered response.
The focus groups were conducted within a private, password-protected section of the same online community Web site as that being evaluated. Participants used their own computers to access the focus groups and were asked to participate for a minimum of two 15-minute occasions during a 2-week period. Discussions were asynchronous, as is common in online communities.
Participants were recruited through a third-party e-mail invitation. The recruitment directly sought both contributors to the Reality Check forum and users who visit the site and read messages without directly contributing (observers). Participants were allocated to focus groups based on their patterns of use (observers, active observers, or occasional, regular, or frequent contributors) to achieve optimal group dynamics, which is crucial in focus groups, including those conducted online.34–36 Focus groups were led by an experienced, independent facilitator.
Participants responded to seven questions raised by the facilitator (Table 1). The facilitator viewed the focus groups daily, adding prompts and additional questions to encourage responses and discussion among participants.
At the conclusion of the focus groups, transcripts were copied and saved to secure, electronic computer files by the evaluator, and the transcripts were deleted from the Web site. All user accounts created for the purpose of the research were closed and deleted at the focus groups' conclusion.
Health professional survey
A survey of health professionals was designed to investigate their success values for the program. The survey explored health professionals' own use of the online community and their experiences of referring people with diabetes to use it (Table 2).
The survey was administered in an online format and was anonymous. It was piloted with six health professionals and subsequently disseminated via an e-mail sent to all health professionals subscribed to the managing organization (the Type 1 Diabetes Network) e-mail newsletter, which invited them to participate in a survey exploring health professionals' experiences of the Reality Check online community. Recruitment and data collection took place over 35 days.
Thematic analysis was used for both the online focus group transcripts and the health professional survey data. The analysis was conducted by the internal evaluator together with an independent person; for the focus groups, this was the independent focus group facilitator, and for the health professional survey, it was a diabetes nurse educator with experience in social research.
Data were analyzed using a three-stage approach endorsed for qualitative research in public health: data immersion, coding, and then creating categories and subcategories37 (Figure 1). In stage three, a deductive process was applied to determine whether coded data could be categorized into the four core functions that underpin all peer-support programs as described by Peers for Progress,14 namely, 1) assistance in applying disease management in daily life, 2) emotional and social support, 3) linkage to clinical care, and 4) ongoing support. A fifth category, acceptability of specific program delivery features, was also used, consistent with the Peers for Progress14 assertion in their definition that issues of delivery and implementation of peer-support programs should be considered separately from those describing the functions of the peer-support process.
Initially, the definition of each function of peer support was used to create subcategories, and coded responses were matched to these subcategories. New subcategories were created for remaining data that fit within a function area but did not fit the precise definition of that function. For example, “assistance in applying diabetes management in daily life” has been given a precise definition as described above, and from that, two subcategories were created: “real-life experiences of diabetes” and “identifying key resources.” Three new subcategories were created by the researchers to describe data grouped in this category that did not fit into the definition above. These were “problem solving,” “learning,” and “improve management of diabetes.”
Online focus groups
Participation and recruitment. A recruitment e-mail was sent to 4,119 subscribers to the managing organization's monthly e-mail newsletter. The recruitment information was also placed as an advertisement within the online community for 2 weeks. Ninety-nine people registered their interest in participating over a 2-week period. Ten did not meet the study's inclusion criteria; eight had not visited the Web site in the preceding 3 months; one exceeded the age limit, and one was a moderator of the Web site. Eight-nine people were invited to participate in the focus groups.
Forty-two people participated in six focus groups. The average age of participants was 37.2 years (range 22–60 years). Twenty-six percent were male. The average duration of type 1 diabetes was 17.5 years (range 1–46 years). All participants spoke English at home. Thirty-nine participants (93%) resided in Australia, two (5%) in the United Kingdom, and one (3%) in New Zealand. Participants' duration of participation in the program was most commonly reported as > 5 years (by 14 people, 33%), with 12 (29%) at 2–5 years, 8 (19%) at 1–2 years, 4 (10%) at 3–12 months, and 2 (5%) at 1–3 months. The frequency of participation in the program (by visiting the Web site) was most commonly reported as daily (36%), with 26% monthly, 24% weekly, 10% several times per day, and 5% visiting once in the past 3 months. Thirty-one percent were observers, having visited the Web site but not posted any messages in the preceding 3 months.
People with diabetes' success values. Through the first phase of analysis, 66 codes were developed to describe what participants most valued about the program. The 66 codes were examined to determine whether they could be categorized into the four core functions of peer support or the fifth category, which we called “program delivery features.” A large volume and wide diversity of data were found within the first three of the five categories, and, to accurately represent the data, a total of eight subcategories were drawn from the Peers for Progress definitions of the categories. The data in those categories were allocated to subcategories as appropriate, as shown in Table 3.
Finally, there were some data describing success values for the program held by people with diabetes that fit into the broader categories of the Peers for Progress model of peer support but did not fit the more specific definitions of each function of peer support and therefore did not fit the subcategories created directly from the definition. A total of eight new subcategories were created to describe these remaining data, as shown in Table 4.
Program delivery features valued by people with diabetes. In addition to the success values that aligned with the core functions of peer support, participants identified a range of program delivery features they valued. Consistent with the Peers for Progress14 definition of peer support, data relating to delivery and implementation issues were considered separately from those describing the function of the peer support process and were therefore considered as a fifth data category. Within this category, five subcategories were created to describe the program delivery features participants valued, as shown in Table 5.
Participants' negative experiences. Specific focus group questions purposefully sought to elicit negative experiences, and together with those recounted throughout general discussions, 14 codes were developed to describe these data.
Twenty-two participants (53%) stated that they had no bad experiences to report. One representative comment stated, “I have not had a bad experience on Reality Check. Obviously not all posts are equally important or interesting to me, but I just don't read those ones. That's the beauty of a forum.”
Discussion about how the program is managed or moderated elicited some concerns that the manner in which participants who break the rules are “reprimanded” could be taken badly. Generally, however, discussion of program management was positive, with an understanding of the complexity and challenges that moderators face.
Other negative experiences reported by participants can be summarized through the codes developed: not allowing provision of medical advice within the forum is a form of censorship; feeling judged by other participants; negativity/victim mentality of some participants; criticism of health professionals is wearying; repetition and predictability of some topics discussed; being misunderstood led to someone being offended; strong negative views of type 2 diabetes and the public's confusion with type 1 diabetes and the differing management demands between the conditions; not enough people using it at one point; scary experiences discussed (e.g., complications); conflicting opinions causing hostility; and incorrect information is sometimes posted but it is quickly corrected.
Survey of health professionals
Participants and recruitment. An e-mail invitation was sent to 588 people who had identified themselves as health professionals when subscribing to the managing organization's monthly e-mail newsletter. Eighty-one (14%) complete survey responses were received and analyzed.
The average age of respondents was 44 years, ranging from 21 to 64 years. The majority of respondents were female, with only 15 (19%) male. The majority (83%) were based in Australia, with some from New Zealand (5%). Thirty-four (42%) were in rural or regional areas, with the remaining 47 (58%) in metropolitan areas.
The majority (54, or 67%) identified their profession as diabetes nurse educator; the remainder were 7 (10%) other types of nurses, 2 (3%) pharmacists, 1 (2%) paramedic, 1 (2%) medical, and 1 (2%) community program coordinator. There were no specialist physicians or general practitioners. Respondents' work settings were diverse: 40 (49%) hospital, 12 (15%) community-based health service, 6 (7%) primary care/medical clinic, and 7 (9%) private practice. Thirty-two (40%) reported that they or someone in their immediate family had type 1 diabetes, 41 (51%) did not, and 8 (10%) did not state.
Health professionals' success values regarding use of the program by people with diabetes. The health professional survey investigated participants' success values for the Reality Check online community, that is, what they personally valued as defining success for the program. Success values were explored with regard to what health professionals valued about both people with diabetes participating in the program and their own personal use of the program. This article reports what health professionals valued with regard to people with diabetes who they refer to the program. Ninety-one percent of respondents indicated that they refer people to the program.
Through the first phase of analysis, 64 codes were developed, and, as with the analysis process for the focus groups, these codes were subsequently grouped into categories and subcategories drawn from the Peers for Progress definition of peer support,14,15 as shown in Table 6, which offers examples of success values identified.
As with the focus group data analysis, some data fit into the broader categories of the Peers for Progress14,15 model but did not fit the more specific definitions of each function of peer support. Therefore, six new subcategories were created to describe these remaining data, as shown in Table 7.
Acceptability of program delivery features. The program delivery features that health professionals most valued included three aspects. First, they described a group of success values relating to program delivery features that are intentional parts of the program delivery, including it being Australian, focused on type 1 diabetes, anonymous, not moderated by health professionals, online, technically accessible, and well monitored. Second were a group of realized program characteristics, including the positive and constructive attitudes to living with diabetes, the welcoming and supportive nature of the community, the fact that it is easy to understand, and the honesty of participants. Finally, professionals described specific client groups for whom the program is acceptable, including newly diagnosed patients, young people, clients with the necessary intellect to use it, clients who have the “capacity to filter” information, and people living in rural and remote areas or anyone experiencing geographical barriers to other forms of peer support.
Negative experiences. The survey actively sought information about respondents' negative experiences with the program. Thirty-six (44%) stated that they had not had any bad experiences. Thirty-one (38%) did not answer the question. Of the 18 (12%) respondents who described a negative experience, the following codes describe their experiences, and the number of respondents who identified each type is shown in parentheses.
Professionals' responses related both to feedback from patients referred to the program and to personal experiences of using the program. Feedback from patients referred included: the participant could not find someone who had had a similar experience to own(1); participant was attacked for having a different perspective (2); discussing anxiety heightened anxiety (1); negative feedback about my diabetes service was discussed on the Web site (1); moderation of posts was overzealous, and a participant therefore withdrew participation (1); negativity of some discussions had an adverse effect(2); and participants perceived to be “needy” (1). Health professionals personal experiences included: cynical attitudes toward health services (1); negative discussion of type 2 diabetes (1); diabetes management techniques discussed are not consistent with standard recommendations (3); gossip about health professionals (2); and “inappropriate advice”(1). One further negative experience related to other services provided by the managing organization: difficulties in supply of some printed resources.
Differences between stakeholder groups' success values. Table 8 shows what was valued, which of the values aligned with the established definitions of peer support, and how the values vary between the stakeholder groups studied: health professionals and people with diabetes.
Values that were associated with the first core function of peer support, “assistance in applying diabetes management in daily life,” were similar between health professionals and people with diabetes. In the second core function of peer support, “emotional and social support,” people with diabetes identified more values for the program than health professionals, with the additional concepts that they valued being encouraged, staying motivated, facilitating self-reflection, feeling a sense of community, and assisting others. Health professionals' main differential was that parents and caregivers could gain support from the program. Finally, although health professionals valued the program for providing linkages to clinical care, which is a part of the Peers for Progress definition of peer support,14 people with diabetes did not identify this as something that they valued.
Online community valued for peer-support function
A consumer-led online community is valued for a function that is similar to more structured peer-support programs that are more commonly delivered in person. People with diabetes said the online community provides them with a wide range of supports that are distinct from what they may receive from clinical care provided by health professionals and provides complementary processes that participants value highly as contributing to their ability to manage their diabetes.
The success values that people with diabetes identified were consistent with the Peers for Progress definition of peer support,14 with the exception of it providing a “linkage to clinical care.” The success values that health professionals identified were also consistent with the peer-support definition, covering all four core functions of peer support encompassed in the definition.
Despite the definition having been developed predominantly around more structured programs delivered in person, the study confirms that an online community can be conceptualized as providing peer support.
Health professionals' views of online communities
What health professionals reported as being valuable about the online community for their patients was similar to what people with diabetes valued for themselves. However, health professionals identified fewer forms of emotional and social support than people with diabetes themselves valued receiving from the program. Although health professionals and people with diabetes both identified “empathetic listening,” “coping with social or emotional barriers,” “feeling supported,” and “connection and interaction with peers,” people with diabetes also valued some additional types of support, including “encouragement,” “staying motivated to reach their goals,” “self-reflection,” “a sense of community,” and “assisting others.”
There was some similarity among those concepts that people with diabetes identified and health professionals did not, in that they may be conceptualized as broader and deeper forms of social and emotional support. Given that there has been limited exploration of health professionals' perceptions of peer-support programs and barriers to them referring to and supporting such programs, these differences may make a valuable contribution to understanding the opportunities for increasing health professionals' understanding of what peer support can provide for people with chronic conditions.
Expanding the definition of peer support
Although the range of concepts that both people with diabetes and health professionals consistently identified as valuable about the online community fit broadly within the established definition of peer support, some new concepts were also identified that increase our understanding of the many benefits that participants may receive from peer support. For example, within “ongoing social and emotional support,” the concepts of “feeling supported,” “self-reflection,” “connection and interaction with peers,” “a sense of community,” and “assisting others” were identified in this study but not addressed in current definitions.
The study also identified a bias in the current definition of peer support towards place-based programs. The peer-support function of helping people to “identify key resources” is explained as “where to buy healthy foods or … locations for exercise,” which assumes a common location, whereas people with diabetes valued Reality Check for assisting them to identify key resources, yet they described this differently, as involving “new products and equipment,” “hot topics,” “current issues,” and “upcoming events.” Given that there have been few previous exploratory studies of long-running peer-support programs to identify what stakeholders value about them, these concepts provide insight into what people with long-term conditions value about peer support and may make valuable additions to the current definition of peer support.
This study had some limitations. In the health professional survey, nurses were overrepresented (77% of respondents), and no specialist physicians participated. A high proportion of health professional respondents had type 1 diabetes themselves or had an immediate family member with diabetes (40%, when population prevalence in Australia is 0.7%). The focus groups were conducted within the same Web site as the program itself; although the high response rate of 42 participants suggests that that choice was acceptable, the bias toward those for whom the program is already acceptable must be noted.
An additional potential limitation is that the evaluation was designed and led by one of the founders of the program, thereby making it an internal evaluation, which has both benefits and risks. Cummings38 describes the advantages of an internal evaluator as knowing the nuances of the agency and knowing how evaluation can make a difference, as well as having credibility that fosters stakeholder commitment and the ability to promote the use of evaluation findings. Owen39 also argues that an internal evaluator can play a crucial role in promoting learning within an organization, analyzing organizational culture and surfacing and challenging assumptions.
Owen's guidelines for evaluation by “insiders”39 were used to assess the risks of an internal evaluation. Three potential issues were identified: compromise of objectivity, coercive recruitment, and resistance to negative findings. Strategies were embedded in the evaluation design to address these risks, including third-party recruitment, ongoing engagement with the program managers, an independent focus group facilitator, and independent coanalysts of the qualitative data.
Risks associated with the use of an internal evaluator appear to have been minimized, as shown by the high response and participation rates and the successful collection of negative experiences, which suggests that participants felt that their anonymity was protected. The immediate uptake of some early findings from the evaluation by the program managers has demonstrated that the risk of resistance to negative findings was also minimized.
The generalizability of this qualitative study's findings to other peer-support programs is not clear, and further study of other online communities formed around health conditions is required to understand the current and potential role of online communities in health care. However, this study has provided a rare opportunity to investigate two fast-growing phenomena—online communities and peer-support programs—and to explore in some depth the perspectives of both participants and health professional referrers.
The fact that the program has been long-running, allowing stakeholders' viewpoints and experiences the opportunity to have developed over time and through repeated interactions with the program, enhances the findings. Enhancements to the program itself, our understanding of peer support, and our ability to evaluate peer-support programs have all resulted from the study.
This study was submitted as part of a Masters of Public Health degree undertaken at the University of Melbourne, and the project was supervised by Rosemary McKenzie. Dr. Sarity Dodson facilitated the online focus groups and assisted with data analysis. Marie Gill assisted with interpretation and analysis of data from the health professional survey. The authors would like to thank the volunteer management committee of the Type 1 Diabetes Network and moderators of the Reality Check online community, without whom the subject of this evaluation would not exist, for their generous and wholehearted support of this effort. Special thanks go to Colleen Clarke, administration coordinator with the Type 1 Diabetes Network, who managed the third-party recruitment for the focus groups.
Note of disclosure: Ms. Gilbert was one of the founders of the online community that is the subject of this study and, at the time the study was undertaken, was president of the Type 1 Diabetes Network, which manages the online community.