Training Youth With SCI to Improve Efficiency and Biomechanics of Wheelchair Propulsion: A Pilot Study

Participants were identified and recruited from one pediatric specialty hospital. Inclusion criteria included youth 4 to 21 years of age who were manual wheelchair users and who had sustained an SCI at least 3 months prior to participation in this pilot study. Subjects with concomitant or preexisting traumatic brain injuries were excluded. The study protocol, consent, assent, and Health Insurance Portability and Accountability Act forms were all reviewed and approved by the Institutional Review Board. Informed consent and assent when appropriate were obtained from all participants and their caregivers as appropriate, and the ethical treatment of human subjects was followed throughout the research process.

Demographic information was collected from the medical records for all participants including gender, height, weight, age, level of injury, age at time of injury, amount of time in months the participant had been using a manual wheelchair, and handedness. The subjects then completed the Wheelchair User's Shoulder Pain Index (WUSPI) questionnaire to determine preexisting UE pain.

The subjects were studied in a 3-D motion analysis laboratory where reflective markers were placed on their UEs on the acromion process, humerus, medial and lateral epicondyles of the elbow, the styloid process of the radius and ulna, and the third metacarpophalyngeal joint (MCP) and on the axle of each wheel (Figure 1). Each participant's wheelchair was fit with the appropriate size SmartWheel (22 in. to 26 in.), which was placed on the subject's chair on the dominant hand side. Additionally, an exact size matched wheel, provided by the manufacturer, was placed on the opposite side to equal the weight and inertia of the SmartWheel. After placement of both wheels, we collected four trials of baseline data by having the subjects propel their chairs across an approximately 50-foot tile floor. The only instructions given were to go down to the other side of the room at your normal speed. Feedback was provided to the subjects utilizing observational information by a physical therapist and a biomedical engineer on push mechanics and the self-selected shape of propulsion stroke along with SmartWheel data.

Following this, the subjects watched a 2-minute demonstration video describing how the training would be conducted and education on the preferred propulsion pattern. Participants were then securely fastened to a roller system (Figure 1) and aligned to view a 42-in. video monitor in our 3-D Vicon motion capture lab. Hand marker trajectory was recorded using the motion capture software by tracing the shape of the propulsive stroke, and the reflective markers on their UEs allowed them to see their own propulsion patterns in real time on the monitor (Figure 2). Subjects were instructed to create a smooth, semicircular pattern by allowing their hand to naturally drift downward in the recovery phase. After approximately 10 minutes of training, we collected another four trials of SmartWheel data by repeating the testing across the same 50-foot tile floor.

The Statistical Package for the Social Sciences (IBM SPSS, version 22.0; Armonk, NY) was used to analyze the data. Descriptive statistics were used to summarize demographic data points listed above and paired samples t tests were used to compare measures of propulsion including peak force, peak backwards force, speed, push length, push frequency, peak/average force ratio, average push force, and push mechanical effectiveness before and after training.22 Cohen's d was used as a measure of effect size (ES) to assess the magnitude of the difference between each pair of means, where 0.2 = small effect, 0.5 = medium effect, and 0.8 = large effect.

Of the 23 participants in the study (13 males, 10 females), 69% had paraplegia and the mean age at time of participation was 11.9 years (range, 7–19). The average duration of manual wheelchair use was 43.8 months with a range of 3 to 132 months, and injury levels of participants ranged from C6 to L2. Six subjects (26.0%) exhibited shoulder pain at the time of the study, while only two (8.6%) had complaints of elbow or wrist pain.

Demographic and injury-related characteristics as listed in the data analysis were unrelated to study outcomes of interest. Analyses of the biomechanical measures of propulsion were conducted using paired samples t tests. Significant findings were identified for both peak backwards force and push mechanical effectiveness. Post training, the mean peak backwards force was significantly less than the mean peak backwards force pre training, indicating improvement (−2.37 ± 1.9 vs −3.08 ± 2.1; p = .041). Additionally, the mean push mechanical effectiveness was significantly improved post training (.631 ± .17 vs .575 ± .14; p = .033), indicating that a greater percentage of the force applied by the subject contributed to acceleration of the wheelchair following training. It is also important to examine the magnitude of effect in a way that is not influenced by sample size, particularly due to the pilot nature of the current study. Consistent with the significance testing, peak backwards force and push mechanical effectiveness both demonstrated small effects according to Cohen's d (d = −0.35 and −0.36, respectively). Further, three other pairs of means came close to demonstrating a small effect, including peak force (d = 0.15), push frequency (d = 0.16), and average push force (d = 0.18) (Table 2).

Overuse injuries at the shoulder, elbow, and wrist that may contribute to destructive arthropathies and carpal tunnel syndromes are well documented.5 It has been shown that inefficient force applications and poor biomechanics may be important factors in contributing to these injuries, although most of this research has been focused on adults.23 Previous studies have shown that lower forces, slower cadence, and circular strokes during wheelchair propulsion may assist in preventing injury.23 Because of the relatively long lifespan of youth with SCI, they are most likely at high risk of developing UE pain and overuse injuries and would have to endure them for a longer time compared to those with adult-onset SCI. Therefore, intervention with children with SCI is imperative.

Employing an objective wheelchair assessment and proper training on an ongoing basis as the child is growing as well as when the child receives new wheelchairs may help reduce the development of UE injuries over time. In the current study, we evaluated such a program and chose to use real-time visual feedback because it has been shown to be particularly effective in improving accuracy and manual consistency in children.24 In our current study and with a limited number of participants, we saw significant gains in terms of peak backwards force and push mechanical effectiveness after propulsion training using the semicircular pattern. The average peak backwards force was significantly less post training, implying that overall the subjects improved in unintentionally slowing the wheelchair down with each push, which would lead to a smoother stroke. Additionally, the improvements in push mechanical effectiveness (a parameter intended to provide a red flag for inefficient push mechanics) show that post training, a greater percentage of the applied force to the hand-rim contributed to wheelchair acceleration. Push mechanical effectiveness in the current study is an equivalent parameter as FEF in previous studies.18,20 Although we were able to improve push mechanical effectiveness in this study, we did not measure oxygen consumption, so it is unknown whether this improvement was at the expense of mechanical efficiency as has been found in a previous study.19 

No significant statistical changes were found post training for peak force, push frequency, and average push force. Although these metrics were mildly improved, they had very small effect sizes and we cannot conclude that any differences were related to the training intervention. Speed and peak/average force ratio remained relatively consistent post training. Future research should continue to evaluate the potential impact of similar wheelchair assessment and training programs incorporating feedback for children with SCI including serial programs as they age.

In addition to a small number of participants, there are several other limitations to this study. Wheelchair fit is an important component of proper wheelchair propulsion, and we did not evaluate or adjust the wheelchair prior to or during the training session. By evaluating subjects only in forward propulsion at a steady speed, parts of wheelchair propulsion including turning and maneuvering around obstacles were negated. Another limitation of this pilot project was the short distance of propulsion due to the constraints of the space in the motion lab. Additionally, as the training was performed on the roller system we had available to us, the resistance during propulsion on the tile floor was not the same as on the roller system. Because of the different inertia, the hand has to accelerate the wheel at the beginning of the push on the rollers, while on the ground the hand has to catch up with an already rolling wheel and attempt not to brake it. Finally, as this was a pilot study with limited resources, there was no control group, it was only a single training session, and we were unable to conduct long-term follow-up to assess carryover of training or possible effects on UE pain.

Wheelchair propulsion is a skill that may be quickly improved with proper education, minimal training, and practice. As current rehabilitation trends allow little time for this in the traditional setting, a comprehensive assessment of wheelchair propulsion and education on techniques to prevent future UE injury must be repeated over time and throughout the child's lifespan with the goal of improving wheelchair propulsion and reducing overuse injuries to the upper extremities.