Unfortunately, few behavioral exercise interventions have been tested among adults with SCI [7]. The majority of the six SCI behavioral exercise intervention studies outlined by Nery et al were delivered in person [7]. Given that transportation is an important barrier among adults with SCI who may want to participate in trials [8] and that transportation-related costs are an obstacle to exercise participation as a whole [9], the reach of in-person interventions is limited to individuals who live in proximity to the research sites or who can easily pay to commute to the sites.

Recent technological advances within the rehabilitation context [10] allow for other alternative and viable options for delivering cost-effective and equitable exercise interventions to community-dwelling adults with SCI. One such example is telerehabilitation, which consists of delivering rehabilitative and health-related services from a distance with the use of telecommunications [11]. A systematic review found that telerehabilitation was as successful as traditional, in-person rehabilitation across various settings and populations [10]. Of importance to exercise promotion, telerehabilitation has the advantage of delivering the intervention in the person’s natural environment and is an effective tool in the self-management of chronic conditions, which includes increasing exercise participation [12]. Thus, telerehabilitation has the potential to be a viable alternative for community dwelling adults with SCI.

Telerehabilitation has also been shown to be effective at managing mental health outcomes among adults with SCI [13-15]. For instance, Dorstyn et al [13] conducted a systematic review of telerehabilitation and found, across seven studies (N=272), that telerehabilitation was effective at improving short-term psychological (eg, depression) and functional outcomes and was time- and cost-effective. Furthermore, physicians, nurses, psychologists, and physical/occupational therapists have effectively applied telerehabilitation to consult on a wide range of outcomes such as pressure ulcers, depression, and functional motor skills [14-17].

In terms of exercise participation, a telerehabilitation phone-based intervention was conducted among adults with SCI, which consisted of a 6-month national telephone-based exercise counseling service (called Get in Motion) for Canadian adults with SCI [18]. This service was derived from two previous randomized controlled trials (RCT) [19,20]. Results from this telephone counseling service demonstrated that individuals’ intentions to be active remained high throughout the 6 months. The percentage of adults with SCI who participated in regular exercise (defined as ≥30 minutes of moderate-to-vigorous intensity activity on ≥3 days/week) increased from 35% (n=16) at baseline to 52% (n=24) at 6 months. This study provided initial evidence that telephone-based telerehabilitation interventions could be an effective strategy for increasing exercise participation among community-dwelling adults with SCI. Moving to video-based telerehabilitation will allow the counselor to capture nonverbal messages, demonstrate exercises, and see the person’s natural environment in order to better adapt the counseling (a need that has been previously mentioned by clients using the Get In Motion service [21]). To our knowledge, no study has conducted a video-based telerehabilitation intervention to promote exercise participation among adults with SCI. Because no such intervention exists, we propose to develop and evaluate the effectiveness of this type of telerehabilitation intervention, which will also fill the current gap in SCI-specific exercise intervention research [7].

Experts have strongly recommended that exercise interventions be grounded in psychological theory to enhance their effectiveness [22,23]. Self-determination theory (SDT [24,25]; see Figure 1) is a motivational theory based on a humanistic perspective that acknowledges every human being has an innate tendency towards growth, integration, and well-being. According to SDT, satisfaction of the three psychological needs is essential for promoting motivation and well-being [26]. The three basic psychological needs of autonomy (ie, volition in one’s actions), competence (ie, belief in one’s actions), and relatedness (ie, sense of belongingness). In SDT, motivation is conceptualized as either autonomous, where people engage in behaviors because they want to, or controlled, where they engage in behaviors because they perceive they have to [27,28].

As elaborated in SDT, the social environment is central to facilitating the satisfaction of the three basic psychological needs. Exercise counselors who adopt an interaction style that is consistent with SDT principles (ie, a need supportive style) would create such an environment by fostering and helping satisfy the basic psychological needs of autonomy, competence, and relatedness. A key element to support the need for autonomy is to elicit and acknowledge individuals’ perspectives and life aspirations. Support for competence is intended to enhance participants’ perceptions that they have the ability to attain their behavior change goals. An empathic, nonjudgmental approach aimed at understanding participants’ viewpoint would facilitate the need for relatedness.

Previous research supports the tenet that satisfaction of the psychological needs predicts greater autonomous motivation which, in turn, predicts greater exercise behavior [29,30]. SDT-based interventions have also shown success with positively influencing psychological needs fulfillment, autonomous motivation, exercise behavior, and QOL [31-33]. Despite the success of SDT-based physical activity interventions, no such intervention has been applied among adults with SCI.

To our knowledge, only a few published studies have examined SDT-based concepts in adults with a physical disability and they used primarily correlational designs [34,35]. A recently published study demonstrated that autonomous motivation had both direct and indirect relationships with exercise activity among adults with SCI [6]. Specifically, the results showed that autonomous motivation was positively related to the likelihood of meeting the physical activity guidelines for adults with SCI. Taken together, these results suggest that the principles of SDT are applicable to adults with a disability, including SCI.

Given the compelling evidence linking exercise participation and QOL, the low rates of exercise participation among adults with SCI, and a lack of exercise participation interventions, in particular SDT-based, telerehabilitation RCTs, a pilot RCT is needed. The advantage of a pilot RCT is that it allows us to determine the feasibility and acceptability of the intervention, design, and procedures prior to running a full-scale RCT. Furthermore, the effect sizes derived from this pilot RCT can help estimate an appropriate effect size for the full-scale RCT [36].

The development of effective interventions should also be grounded in sound theory given the conceptual and methodological advantages of theory-based interventions outlined earlier. Although most interventions claim to be guided by theory, the majority does not link the intervention components to the theoretical constructs being tested [37,38]. To avoid this common flaw, we have implemented the Theory Coding Scheme [38] into the design of this pilot RCT, as recommended by Gourlan et al [37]. By following this coding scheme, we are ensuring that our intervention is theory-based. The proposed intervention is also novel for the SCI field as, to date, no exercise interventions have used SDT as the guiding framework. We will therefore be the first to extend the generalizability of SDT to adults with SCI, which is an important and often disregarded step of theory testing. Figure 2 illustrates the conceptual and process model that guides this proposed intervention [39].

The overall goal of this pilot RCT is to test an 8-week innovative, SDT- and video-based telerehabilitation intervention aimed to enhance the basic psychological needs, motivation, exercise participation, and QOL-related outcomes of adults with paraplegia. The primary purpose of this pilot RCT is to determine if individuals in the intervention group have greater increases in their basic psychological needs and autonomous motivation and a decrease in controlled motivation, compared to the control group. We expect moderate effects on these SDT variables based on the results from previous SDT exercise interventions among the general population [40].

The secondary purpose of this study is to determine if the intervention group reports greater increases in exercise behavior compared to the control group. Additionally, we will examine if adults with paraplegia who received the intervention report improved scores on QOL-related variables (eg, life satisfaction, participation in daily/social activities, depressive symptoms) compared to the control group. Given that moderate effects were found for exercise behavior in the previously mentioned Get In Motion service and theory-based exercise interventions [18,37], we hypothesize, for Objective 2, that our video-based intervention will have moderate effects on exercise participation. We anticipate small-to-moderate positive effects of the intervention group compared to the control group on the QOL-related outcomes of life satisfaction, participation in daily/social activities, and depressive symptoms given the results of a leisure activity (including exercise participation) intervention [41] and cross-sectional research [3,42] in adults with SCI. We are expecting smaller effects for these variables given the relatively short duration of the proposed pilot RCT.

Our tertiary purpose is to investigate group differences in additional and common psychosocial predictors of exercise participation (eg, action planning), as well as indicators of well-being (eg, positive affect), to obtain a broader perspective of the impact of the intervention. Similarly, we expect moderate differences between the intervention and control groups on these variables.

Finally, we are conducting an implementation evaluation to determine satisfaction with the technology and extent to which the counselor delivered the intervention as intended. No hypotheses are derived for this evaluation given its description nature.

We will recruit 24 adults with paraplegia from three rehabilitation centers, one adapted fitness center, provincial SCI organizations, and through social media. We will randomize 12 participants to the intervention group and 12 to the control group. Based on previous interventions [18], we expect a 15% dropout rate resulting in 20 individuals who will complete the proposed pilot RCT.

Eligible participants will be adults with paraplegia who (1) have sustained their injury at least 1 year prior, (2) are over the age of 18 years, (3) have access to a computer, and (4) speak and understand English or French. Eligible participants will either have the intention to become physically active in the next 2 months (ie, not amotivated as per SDT) or have been minimally active (<2 times a week [5]) in the past 2 months. Participants will be excluded if they are receiving in-patient rehabilitation services, been diagnosed with memory impairments, severe communication difficulties and/or severe visual impairments, do not require a mobility device (eg, wheelchair, cane), or have answered yes to one of the questions on the Physical Activity Readiness Questionnaire (PAR-Q) and the SCI questions on the PAR-QX. Participants who have answered yes to the PAR-Q could be eligible if they provide a note from their physician stating that it is safe for them to participate in exercise activities. We have elected to focus on adults with paraplegia who use a mobility device because we want to pilot test the intervention in a more homogeneous sample than if we included both adults with tetraplegia and paraplegia.

Participants will be invited to complete each of the primary, secondary, and tertiary outcome measures at all three time points (baseline, 6, and 10 weeks). Participants in the intervention group will also complete the relevant measures under implementation outcomes.

Prior to starting the RCT, the exercise counselor will be trained in SDT-based counseling, motivational interviewing, and adapted exercise activity. The exercise counselor will also pilot the intervention with an experienced counselor and 2 inactive participants in order to troubleshoot any potential issues. The exercise counselor and research assistants will be also trained on the REACTS software system.

At the time of submission of this protocol, data collection was ongoing with 10 participants who had completed the study (5 intervention, 5 control) and 6 who were enrolled in the study (3 intervention, 3 control). At the time of publication, all 24 participants have been randomized.

Data will be screened for statistical outliers and assumptions for each statistical test will be examined [65]. Participant attrition is a common problem in trial studies and case-wise deletion can bias results [66]. As such, missing data will be examined to determine the pattern of missing data and imputed with multiple imputations methods that are appropriate for small samples [67]. Further, we will collect reasons for missing data/attrition. These approaches will help inform the expected attrition rate, reasons why people did not continue with the trial, and required sample size to conduct a subsequent larger RCT. Once data are cleaned and imputed, demographic and injury-related covariates will be examined for each outcome through a correlation matrix. Any identified covariates will be controlled for in the subsequent analyses. The same analysis will be used to determine if the intervention group improved on the variables assessed for hypothesis 1 (basic psychological needs and autonomous motivation), 2 (exercise, life satisfaction, participation in daily/social activities, depressive symptoms), and 3 (well-being, and psychosocial predictors of exercise). A regression analysis approach is recommended for analyzing RCT data as this approach can address specific questions arising from the study design, above what could be expected from analysis of variance or other similar types of analyses [68]. As such, two hierarchical multiple regressions will be used to predict changes in participants’ scores between baseline, 6 weeks, and 10 weeks. Specifically, participants’ demographic and injury-related covariates will be entered in the first block, followed by their baseline scores on all study variables in the second block, and finally their group (ie, intervention or control) in the third block. This analysis will be conducted twice: once predicting participants’ scores at 6 weeks, and again predicting their scores at 10 weeks. The differences in variance explained by the model (r²) and the weight of the regressions (standardized betas) will be used to evaluate whether the model fit improves between Blocks 2 and 3, as well as to evaluate the success of the pilot RCT. In line with the Theoretical Coding Scheme, exploratory analyses will also be conducted to investigate the strength and direction of the relationships between the SDT variables and exercise participation, looking for moderate-to-strong relationships. Correlations of .1, .3, and .5 represent small, moderate, and strong relationships between the variables [69]. If the intervention has a moderate effect on the SDT variables and if the SDT variables have moderate-to-strong relationships with exercise participation, we can then hint at a potential mediation (ie, intervention is having an effect on exercise participation through the SDT variables). Because of the sample size, we do not have sufficient power to run full mediation analyses and it is for this reason it remains exploratory in nature. Descriptive statistics will be conducted for the implementation and feasibility evaluation.

Our results will have higher credibility due to our design. By using an experimental design, specifically an RCT, cause and effect conclusions are more plausible, which are superior to cross-sectional or longitudinal designs. In addition, all our measures have been validated and found to be reliable in past research, increasing the credibility of our findings. We have also clearly outlined our primary, secondary, and tertiary purposes and associated hypotheses, giving transparency to our proposed project. Our specific outcome-based hypotheses are based on theory or past research, which allows for ease of duplicability of the results. The pilot RCT will also aid in identifying any changes that are required to our design, measures, or procedures before running a full-scale RCT. The implementation and feasibility purpose will enable us to determine if any changes to the intervention protocol should be implemented when designing the subsequent full-scale RCT.

Because we are testing a pilot RCT and focusing on the effect size rather than P values, we believe that our results will be able to be transferred to a future, larger, interprovincial exercise telerehabilitation intervention. It is only with this pilot RCT that we can reliably estimate the sample size needed for this larger trial. Our intervention components are predetermined as they are based on sound theory (ie, SDT) and from a taxonomy of behavior change and relational techniques [70,71]. We have also used the Theory Coding Scheme to develop our intervention to ensure that it is truly a theory-based intervention.

We have set procedures to control for some biases. For one, the randomization procedure will help to reduce some biases as the control and intervention group should be equivalent. The research assistant will randomize the participants by opening a blinded, prelabeled envelope. Therefore, the research assistant cannot bias group allocation. The second research assistant will also be blinded to the participants’ group allocation when collecting the follow-up data, which should further reduce biases. We are also recruiting from multiple sources, which would allow us to reduce bias that might stem if participants were recruited from only one source and subsequently enhance the generalizability of our results. With the control group having access to the exercise counselor after the study, we hope to reduce the compensatory rivalry bias. We are also avoiding focus on P values due to their reliance on sample size [72]. Because of our sample size, if we solely focus on the P value we might be making a type II error.

In line with the general limitation of telerehabilitation, our intervention is limited to individuals who are comfortable using a computer and capable of running the Web-based audio-video REACTS software. Given the small sample of this pilot RCT, we will be unable to conduct subsample analyses to investigate whether the intervention had effects for individuals with specific demographic (eg, age, education level) and SCI-related factors (eg, years since injury, adapted home). However, we do plan to control for these variables in our regression models through a correlation matrix. These correlations will enable us to be mindful of these factors when designing the full-scale RCT.

Results of this study will be translated through three main avenues. First, we plan to share our results with the academic community by presenting the results at SCI, disability, and/or exercise-related scientific meetings, and publishing our results in scientific peer-reviewed journals. Second, we will translate our findings to local adapted exercise centers to inform programs and kinesiologists of our study’s results and approach. Finally, we will share our results with various community organizations who may want to distribute to their larger membership to reach adults with SCI. Our results will therefore be disseminated to academics, the broader SCI community, and relevant professionals.

This study will be the first to test SDT in an exercise intervention among adults with paraplegia. It is also the first to examine the use of an online video-based telerehabilitation approach in this context. We anticipate that the results of this pilot RCT will support that the intervention is appropriate for adults with SCI and feasible to test in a full-scale RCT.