This study was conducted at Baylor Scott & White Institute for Rehabilitation in Dallas, Texas. Recruitment began in late 2016, and the study activities were concluded by May 2017. The Baylor Scott & White Research Institute human subjects committee approved the study activities (Institutional Review Board #016-091), and all participants provided informed consent prior to beginning study participation. Eligible participants were between 18 and 70 years of age and were not pregnant; lived with an SCI at the level of C5 or below for more than 6 months; used a wheelchair; spoke and read English; had a computer and internet access; and could independently navigate the study website. Former patients with SCI from our inpatient rehabilitation hospital who agreed to be contacted for future studies were invited to participate in the study via telephone call or email, and staff handed out fliers at the outpatient clinic and the monthly support group for people with SCI. Study staff screened potential participants for eligibility by telephone. For those deemed eligible to enroll, study staff obtained a medical clearance by fax to participate in an intervention targeting physical activity from each participant’s primary care provider. Participants in this pilot study were not eligible to continue to participate in the full randomized controlled trial (RCT) study, which is currently under review.

The WOWii intervention comprises three key components: (1) the WOWii exercise program provided via modules on the WOWii website, (2) a starter package of exercise equipment, and (3) weekly virtual videoconference meetings led by the study staff to facilitate support for exercise. Study staff who led the videoconference meetings held master’s degrees in public health. They were trained to deliver the intervention and supervised by the principal investigator, an applied behavioral psychologist who has delivered exercise interventions to people with disabilities for over 20 years.

For this pilot study, participants were given a username and password to access 4 weeks of the WOWii website modules, which addressed getting started (eg, importance of exercise to health, benefits for people with disabilities, progression and safety, exercise basics), goal setting, identifying and addressing barriers to activity, and enlisting support (Figure 2). There was an activity at the end of each module that was designed to reinforce the material covered. Screenshots of the website and activities are available to view in Multimedia Appendices 1-4. The WOWii website uses an application programming interface to collect data from each participant’s Polar account, allowing the study staff to download participant exercise data weekly in addition to their completed module activities.

The starter package of exercise equipment included resistance bands, an aerobic exercise DVD, and a pedal exerciser. Participants were also given Polar A300 wristwatch activity monitors [29] and Polar H7 heart rate monitors [29] (Polar Electro Oy) for recording the frequency and intensity of exercise bouts. These devices operate with proprietary Polar software, Polar Flow [30], which tracks and displays its data on the watch that can be downloaded through a compatible smartphone app or a web browser directly from the Polar Flow website.

Weekly videoconference meetings conducted over Skype were facilitated by a study coordinator and/or the principal investigator. The content of these discussions was designed to reinforce the information covered during each weekly module from the website. The format of each session provided time for the leader to briefly review the key points of the web-based module while facilitating conversation among the group regarding their experiences with each module topic. At the end of each session, leaders prompted the participants to complete the weekly module activity that reinforced the material discussed and asked participants to share their activity responses with the group. Session leaders recorded weekly attendance in a web-based spreadsheet.

Summary statistics were generated to describe the levels of individual and overall participant engagement as measured by percentage of sessions attended, percentage of weeks of exercise tracking via the Polar device, and percentage of website module activities completed. The data were summarized using SAS, version 9.4 (SAS Institute).

Although individual participation levels varied from week to week, the participants demonstrated consistent engagement during the 4-week pilot, with an overall engagement rate among the sample of 75% for attending the weekly videoconference meetings, completing the web-based activities, and wearing and syncing the arm-based Polar activity monitor. The participants showed the highest rates of engagement for completing the web-based module activities (85%) and attending the videoconference sessions (83%). These activities offer participants the space to learn about behavioral skills critical to successful behavior change and to discuss with the group their plans and personal experiences implementing those plans. The rate of syncing the Polar activity monitors was lower at 56%, although 90% of participants synced the watch in at least 1 week. It was unclear whether participants had trouble syncing their watches or were not exercising each week.

This pilot study offers preliminary evidence that a web-based platform may be feasible for delivering a PA intervention program to meet the unique needs of people with SCI. Feedback from participants in this feasibility study was vital in implementing changes to the final website used during the RCT. For example, participants recommended incorporating a more robust peer mentoring component than we originally envisioned. The participants emphasized the importance of having individuals with SCI available to discuss their successes in increasing PA as a motivational piece for those starting the program. Additionally, participants advocated for adding cost-effective resources to the website, such as equipment and fitness programs, to best accommodate those whose disability may have limited access to disposable income. After conducting this initial feasibility study, the research team initiated testing the full 16-week theory- and evidence-based program in a randomized controlled trial with 168 total participants with SCI. That trial offered a more robust test of the effectiveness of the program in promoting increased moderate-intensity PA among a large SCI sample as well as the program’s effects on the participants’ self-reported self-efficacy and barriers to engaging in health promoting behaviors.

In contrast with the study by Kosma et al [25], which included a web-based message board that received only 6 postings over the course of the study, attendance at the virtual videoconference sessions was relatively high at 83%. During these weekly sessions, we observed that the participants were actively engaged in the conversation, and several reported that their favorite part of the program was connecting with other people with SCI. We believe that the high engagement we observed was due to participants being able to see one another and connect by videoconference, which mitigated any discomfort participants may have felt with sharing information on the web. Notably, a study published in 2019 that obtained qualitative feedback regarding a web-based portal aimed at increasing self-efficacy for exercise among a population with SCI reported that participants endorsed many of the features present in our program, such as self-regulation strategies, knowledge, and action planning [31].

Our observed attrition rate of 20% falls within an expected and acceptable range based on the previous WOW trial and research on weight loss and internet-based interventions. The WOW trial saw a 33% attrition rate [27], while a 2011 systematic review examining dropout among participants of intensive lifestyle weight loss interventions (61 total studies) found that the reported attrition rates among participants ranged from 9% to 90% depending on the length and type of the intervention [32]. Other systematic reviews examining participation in internet-based interventions for anxiety and depression found attrition rates of up to 83% [33,34].

A potential study limitation was the lack of racial diversity among the participants despite good diversity in terms of education and income. Additionally, initial problems with Polar watch data exports limited our ability to accurately measure exercise in minutes. However, this pilot study yielded valuable knowledge that was implemented in the full RCT. For example, due to confusion regarding the difference between PA and exercise, more information was added to the first website module. Changes were also made to bolster the participants’ interaction by having peer mentors with SCI join virtual sessions, content was conveyed by adding more informational videos, and other content was added about important disability-related legislation.

The WOWii program incorporated the following three novel components to increase PA among people with SCI: (1) group videoconferencing to leverage social support, (2) a web-based portal for information related to increasing PA, and (3) provision of exercise equipment to facilitate PA while mitigating barriers. Our pilot study showed promise regarding the ability of WOWii to engage participants. The pilot study also led to improvements that were made before implementing the full RCT to study the effectiveness of the intervention. Given the prevalence of transportation issues and computer use among people with SCI, this intervention may be a valuable contribution to address the scarcity of lifestyle interventions which are accessible to this population.