For individuals with spinal cord injury (SCI) who live in the community, the risk of experiencing secondary conditions such as pressure injuries or urinary tract infections is high. Even up to 20 years after the injury, rehospitalization rates for people with SCI remain over 30% [1]. To reduce the risk of secondary complications, people with SCI require knowledge about their injuries and strategies to mitigate risk. More specifically, knowledge and skills related to the management of “symptoms, treatment, physical, and psychosocial consequences and lifestyle changes inherent in living with a chronic condition” are required [2]. However, evidence suggests that knowledge and access to self-management tools are inadequate in the SCI community. For example, although some individuals with SCI report high levels of knowledge about bladder or pressure injury management on discharge from inpatient rehabilitation, studies indicate that less than half leave the hospital with clinically adequate knowledge [3]. Moreover, lack of knowledge is not the only barrier to self-care in the community; physical barriers, cost, lack of motivation, and pain also complicate the efforts to independently monitor and maintain health [4-6].

Self-management support programs are community-based tools designed to increase the knowledge and skills required to independently manage a chronic condition. As such, these programs carry the potential to fill some care gaps experienced by people with SCI in the community. Many existing programs, such as the Stanford Chronic Disease Self-Management Program [7] or the UK Expert Patient Programme [8], address a wide variety of chronic health conditions and are led by peers with the lived experience of a chronic condition. However, although peer-led programs have been associated with positive health outcomes (eg, improvements in self-efficacy, health-related quality of life [7,8], lower hospitalization rates [9], and reduced health care expenditures [10]), the typical formulation has shortcomings for community members with SCI. In one qualitative study of the CDSMP, for example, participants with SCI reported less program satisfaction than those with other conditions due to issues such as the lack of familiarity with SCI on the part of facilitators [11]. In this study and in another by Munce et al [12], individuals with SCI expressed a preference for web-based programming over in-person programming [11] and for content tailored to SCI management [11,12]. Tailored self-management programs delivered via the internet may therefore be met with higher satisfaction by individuals with SCI.

Community-based programs that deliver services to the SCI community via telephone or the internet have been both well received and are increasingly common in the literature [13-18]. Telephone-based self-management programs for people with SCI have proven to be safe and acceptable and to improve the participants’ level of activation and awareness [13-16]. Examples of internet-based health interventions include a nurse-led telehealth intervention that focused on newly discharged people with SCI [17], a tele-exercise intervention led by remote exercise coaches [18], and programs to support consultations with specialists [19] or peer educators [20] via iPads. However, internet-based self-management support for SCI, although associated with promising usability and feasibility results [19-21], are a relatively new concept and require further evaluation. The internet may indeed have special affordances for people with SCI; in randomized trials comparing internet and phone-based nursing interventions following inpatient rehabilitation, for example, internet-based programming is associated with the least number of postdischarge hospital visits [17]. Internet programming has also been identified as preferable to phone-based programming by members of the SCI community in a qualitative study by Munce et al [12].

In response to the interest in and need for web-based self-management support, a web-based program for users with SCI was developed at the University of Toronto. This program called SCI & U was modeled after telephone-based support programs (eg, SCI Action Canada [13,14] and My Care My Call [15,16]) and is part of a growing family of web-based support interventions for SCI (eg, PHOENIX [18-20] and SCI Health Storylines [21]). Like telephone-based interventions [13-16], the program consists of highly structured interactions between participants and trained peers with SCI, or health coaches. This is because, in the context of telephone-based programs, peer coaches have been perceived by participants as powerful motivators for behavior change and relevant, credible mentors for skill development [22]. Structures have been created to support participation by members of the Canadian SCI community in the program’s development [23]; these include a community advisory board and interdisciplinary design team. This kind of participatory approach reflects that of other research groups [13-16,20] and is intended to promote end user acceptance [24] and respect for guidelines for the creation of health programs serving individuals with disabilities [25].

This study aims to evaluate the SCI & U web-based self-management program in the context of a pilot study. The specific objectives were to assess implementation feasibility and the potential impact on self-efficacy and experience of secondary conditions. Evaluations of feasibility and impact were based on an analysis of quantitative data gathered before and after program participation. Qualitative data from postintervention interviews with the pilot participants were analyzed to validate and expand upon the quantitative findings.

The study followed an explanatory, sequential mixed methods design with a quantitative component followed by a qualitative component [26]. The quantitative data were the first focus of the analysis, and the qualitative approach used was postpositivist in that the qualitative data did not serve to test the hypotheses but to create an understanding of the intervention that could be shared between the research team and participants [27]. The integration of quantitative and qualitative results took place at the interpretation and reporting levels. A joint display was created to relate the quantitative data to the relevant and explanatory quotes [28].

The health coaching intervention revolved around structured, web-based partnerships between participants and trained peer health coaches. In keeping with prior work [16], the intervention consisted of 2 principal components: trained peer health coaches and support materials.

A total of 5 peer health coaches were Canadians with SCI who acted as mentors and were trained in established health promotion techniques and paid as professional contractors. Peer mentoring after an SCI, on its own, has been associated with improved confidence in self-management skills and reduced hospital admissions [28]. However, trained peer health coaches have additional strengths as they have been educated in motivational interviewing (MI) [29] and brief action planning (BAP) [30]. Both MI and BAP dictate protocols that encourage health care recipients to actively participate in health care decisions [29,30]. Specifically, BAP helps health care recipients elicit, parameterize, and track realistic goals and plans for health behavior change [30]. It has been successfully embedded in many self-management programs [30], including My Care My Call [15,16].

Web-based support materials were housed in a publicly accessible, curated information resource for SCI management [31], and a secure platform was designed to support confidential one-on-one interactions between the participants and web-based peer health coaches [32]. Both were developed using participatory processes [23]; the development of the coaching platform was additionally supported by an application program interface provided by a Toronto-based company called QoCHealth [33]. This allowed an in-house development team to create videoconferencing tools for Google Chrome browsers, tailored lists of topic-specific health information, and secure forms for goal setting and BAP (Figure 1). Of note, participants were not restricted in using alternative tools to communicate with coaches (eg, Skype) if Google Chrome was unavailable or these tools were familiar and comfortable to the participants. No conversations between the participants and coaches were recorded.

The consenting participants were recruited via community organizations and matched with the coaches based on their availability and the participants’ review of web-based coaching profiles and coach availability. The participants were matched with coaches with whom they had no relationship before the study. For the most part, the coaches worked with 2 participants at a time. Once partnerships were established, the participants engaged in up to 6 web-based self-management sessions with their coaches, using the coaching platform. The sessions were designed to last between 45 min and 1 hour, and pacing was flexible; however, the participants were asked to schedule sessions within a span of 2 weeks to complete all sessions within 12 weeks. The coaches were asked to dedicate as much as 1 hour of their time to prepare for each hour of interaction with their clients. The length of the sessions, flexible pacing, and program duration were informed by research detailing the relationship of these factors to the effectiveness of a telephone-based exercise intervention [14].

During the first SCI & U session, the participants worked with coaches to identify health priorities; these were recorded using a web-based version of the Patient Generated Index (PGI; Figure 1) [34]. The PGI allows individuals to formulate personalized goals and expectations for health care interventions; it has been used as an outcome measure in studies of chronic conditions and to articulate patient-centered treatment objectives [35,36]. At the end of each session, the participants were invited to select a single topic from a predefined list of health management topics for the subsequent session. The list of topics was informed by a prior survey of the Canadian SCI community [12]; these included management of pain, diet, exercise, mental health, bowel, bladder, skin, and sexuality.

During sessions, coaches discussed topics with participants and shared stories, listened, worked through issues, or helped formulate goals or plans according to the BAP process. Coaches were given scripts to guide conversations, but the precise content and structure of each session was directed by the participants. The scripts contained information standards for the MI and BAP processes, as well as topic-specific information drawn from a web-based educational website for the SCI community called SCI-U [37]. These educational materials were codeveloped and validated in the context of user studies by an author (JS) working alongside a team of Canadian SCI researchers and members of community organizations [38].

Table 1 provides details of the demographics of participants. A total of 11 participants took part in the pilot, including 7 women and 4 men. Of the 11, 7 were from relatively rural Ontario (ie, municipalities with populations under 100,000); others were from the Vancouver area in British Columbia. Traumatic SCI was reported by 8 participants. Of these, 3 reported nontraumatic SCI, and 1 reported spina bifida. The participants who took part in the pilot reflected a mix of gender, level of injury, injury type, and use of mobility aids. All had been living with SCI for >5 years, with an average of 20 years and a median of 22 years. Most reported that they were regular users of both the internet and had access to email; several additionally reported regular use of social apps (eg, Facebook) and YouTube or games.

Of the 11 consenting participants, 10 completed web-based baseline assessments, and 9 completed both baseline and follow-up assessments. One participant was not comfortable using the web-based coaching platform, so the research team deferred to telephone delivery of the coaching service for this person. Although she completed her initial visit via videoconferencing software, this required the involvement of a personal support worker, and the telephone was therefore seen as preferable. This participant did not complete any web-based activities or assessments, so her data were excluded from the quantitative analysis of impact and feasibility.

Baseline assessment data are shown in Table 3. It may be noted that 2 participants reported PHQ-8 scores consistent with major depression at the outset of the pilot (ie, their scores were >10), and on the SCI-SCS, more than two-thirds reported that they had moderate or significant recent problems with urinary tract infections, circulation, spasticity, or pain. Recent moderate or significant bowel, bladder, and sexual dysfunction were also reported by more than half of the participants.

The quantitative results related to changes in measures before and after the pilot are also given in Table 3. The graphs detailing baseline scores by changes in scores for each outcome measure have been provided in Multimedia Appendix 1. Positive postintervention changes were associated with all outcome measures, and effect sizes for the UW-SES and eHEALS were moderate (ie, >0.5). Effect sizes for the measures of resilience (SCI-QOL-R), depression (PHQ-8), and secondary conditions (SCI-SCS) were small (ie, >0.3). No changes were found to be significant, although changes in SCI-QOL-R scores neared significance (P=.08).

A closer inspection of the scores on the SCI-SCS subscales indicated larger effects related to bowel dysfunction and pain subscales of the SCI-SCS. The estimate of effect size for self-reports on the pain subscale based on Hedges g was moderate at 0.52 (95% CI 0.01 to 1.04), whereas that of the bowel dysfunction subscale was large at 1.04 (95% CI 0.04 to 2.11).

The coding dictionary that was agreed upon by the 4 authors (SA, JS, TT, and SM) is presented in Multimedia Appendix 2. A tree map illustrating codes as they were applied to transcripts is shown in Figure 2.

The following results focused on the theme of “Program Impact,” as the statements related to this theme were able to directly validate and expand upon the quantitative data. In addition, we focused on the “Role of Coach” as this theme was frequently colocated in the transcripts alongside the participants’ statements about program impact.

The results indicated that the delivery of a web-based self-management program tailored for the Canadian SCI community is feasible, and that its use may be associated with positive changes in terms of self-efficacy, mood, resilience, and experience of secondary conditions. The qualitative and quantitative results illustrated the wide variety of self-management topics touched upon by the intervention and the range of self-reported behavior changes adopted as a result. In addition, the qualitative results shed light on the specific mechanisms used by the coaches to promote behavior change; these included provision of encouragement, information, social support, and accountability.

The feasibility results, in terms of adherence and retention, reflected the preliminary work related to the development of other web-based self-management programs for the SCI community, such as PHOENIX [20] and SCI Health Storylines [21]. A feasibility study evaluating the use of SCI Health Storylines by 20 individuals with SCI found that 85% were retained as inpatients and 70% were retained at 3 months post-discharge [21]. Our results were similar, as 81% of the participants remained throughout the pilot. However, the adherence measures for the two studies were not directly comparable, as SCI storylines did not require attendance at the scheduled web-based sessions with a coach. Instead, our adherence measures better reflected those from a tele-exercise intervention for the SCI community, which, like ours, reported 100% adherence [18]. This same study reported that 85% of the exercise data were successfully recorded by the developed technology; in this pilot, the data were successfully collected from 9 of 10 participants, and each study measure was completed by at least 7 of the 9 participants.

In terms of technology usability and acceptance, the results reflected those of the PHOENIX study, which demonstrated iPads and iTunes U to be acceptable for peer-led education and self-management support among users with SCI [20]. A related study exploring the use of iPads to facilitate periodic web-based medical consultations between users with SCI and health professionals revealed similarly high user acceptance of tablets and an overall preference by participants for the web-based, rather than in-person or telephone-based meetings [19]. Our results echoed the same in the context of an intervention spanning multiple sessions. In addition, most participants in this study were able to satisfactorily complete web-based activities; only one deferred to use of the telephone to complete activities such as goal setting. Both this study and the PHOENIX study illustrate participants’ relative comfort with mainstream, community-based videoconferencing tools (eg, Skype, Face Time, and Zoom) [20]. Although this study provided custom videoconferencing tools, most deferred to tools with which they were familiar, even despite potential failure to comply with Canadian Health Information Privacy Policies (eg, Personal Health Information Protection Act).

The topic of the use of videoconferencing software in the context of community-based research projects like this one is an increasingly controversial topic, particularly in the light of revelations regarding potential security loopholes [54]. Given that videoconferencing data were not collected by the research team and security risks were not well understood, research partners in this study permitted deferral to communication tools to promote accessibility. The built-in tool was, as mentioned previously, limited in that it was optimized for Google Chrome and would not work on other devices such as iPads. The issue has since been discussed with the University of Toronto REB, which is now indicating that it will explicitly recommend the use of institutionally sanctioned software (eg, Skype for Business [52]) for research communication and will ask research teams for justification to use alternatives (eg, Zoom [51]) in future studies. Moreover, these policies are actively changing as new risks are exposed. Moving forward, the SCI & U team has chosen to focus on informing the participants about potential security risks related to web-based communication, and our protocol is being amended to reflect this change in practice. The practices will likely continue to change as the situation evolves and policy comes to light.

The feasibility results also demonstrated user engagement and highlighted the variety of ways in which participants chose to tailor the SCI & U program. Although diet and exercise were the most popular topics of discussion during the coaching sessions, the participants also chose to discuss issues such as pain, bowel dysfunction, and sex. Flexibility was found to be similarly valuable in the context of the Get in Motion intervention. Participants in this study were found to have very different needs in terms of program pacing and duration based on their baseline levels of activation, for example [14]. In addition, the users in this study elected to set goals 30% of the time, indicating engagement with program activities.

In addition, both this intervention and others [15,16,20] illustrate the feasibility of participatory approaches to program development. For the PHOENIX program [20], a community-based task force, including members of the SCI community, worked to identify the needs, develop digital content for, and provide feedback on the program as it was developed. In this study, a similar task force was asked to iteratively comment on program content and design [23].

Perhaps, the impact results might best be related to those from a recent randomized controlled trial (RCT) of the My Care My Call program [15]. This was a peer-led, self-management support intervention targeting the management of secondary conditions delivered to 84 participants with SCI over 6 months. As with SCI & U, peer health coaches were trained in the use of both the MI and BAP techniques. The results showed that program participation was associated with a significant increase in measures of activation, resource awareness, social participation, and quality of life. In this study, program participation was associated with similar improvements in the measures of self-efficacy and resilience and in the ability to navigate electronic health resources.

Fewer studies have measured the health outcomes resulting from self-management interventions in the SCI population, and none that we know of are built around web-based interactions with trained peers and associated activities. However, for newly discharged people with SCI, peer mentoring has been associated with a reduced number of return hospital visits after 6 months [55]. The participation by individuals with SCI in peer-led exercise programs has been associated with improved self-reported physical activity 6 months after program participation [13]. Similar exercise outcomes, in turn, have been associated with decreased pain and improved mental health [56,57]. This study focused on relatively short-term outcomes (<12 weeks) and measured the experience of secondary health complications based exclusively on self-reports on the SCI-SCS. Nevertheless, the results pointed to potential moderate effect sizes on this scale, with larger effects on specific subscales of the SCI-SCS (eg, bowel dysfunction and pain). Some effects might have been incidental to the changes in diet or exercise as such changes were widely reported by the participants during the interviews.

The qualitative results illustrated the variety of ways in which coaches supported the participants’ health-related behavior change. The specific roles taken up by the coaches echoed those taken up by the coaches of the My Care My Call program. A content analysis of more than 500 My Care My Call telephone calls found the coaches to have fostered behavior change by acting as role models, advisors, and supporters [22]. The same study found the role of advisor as reflecting the coaches’ training and expertise in the MI and BAP techniques. The coaches in this study also acted as role models or supporters and served as advisors when they walked the participants through the BAP protocol or provided information. However, the coaches of SCI & U were additionally asked to act as advisors when they provided technical support. Technical requests might indicate a somewhat expanded role of advisor in the context of programs like SCI & U.

The results were based on a limited sample. This was a small feasibility study with a convenience sample of 11 participants from two Canadian provinces. The individuals who responded to the recruitment calls might have been people with both motivation and internet access a priori and therefore biased. Moreover, the majority of the participants experienced injuries more than 10 years before commencement of the study.

The quantitative outcome measures were based on self-reports and were only collected at baseline and as participants exited the program. It remains unclear whether the changes reported by the participants were or will be robust to time.

In addition, this study lacked a control group. Confounding factors, such as the weather, were not controlled and might have influenced outcome measures. The impact of the weather on the participants’ mood was, in fact, referenced during the interviews.

Moreover, the results presented in this study omitted the perspective of the coaches in the study on issues such as program feasibility (eg, its demands on their time and their experience of the technology). However, the perspectives of coaches were collected through postintervention interviews that are currently being analyzed. We expect a future study to touch upon the impact and feasibility of the program as perceived by these individuals. To validate the results and ensure that they are robust to both time and confounding factors, the intervention is currently being tested in the context of a longer RCT.

A web-based self-management program that has individuals with SCI partnering with trained peers has been demonstrated to be feasible. Moreover, the program is a promising means to develop self-efficacy, health literacy, and resilience in a population at high risk of health complications while living in the community. These results indicate the potential of web-based tools to bolster existing community-based support for people with SCI, particularly for those in small or rural communities.

Future work seeks to validate the results of this pilot study. An RCT is currently underway for this purpose; this will compare the effects of the SCI & U intervention over the span of a year on an experimental group and relative to wait-listed controls.