Wearable-Enhanced mHealth Intervention to Promote Physical Activity in Manual Wheelchair Users: Single-Group Pre-Post Feasibility Study

Abstract

Background: With the rapid advancement of technology, using wearable devices and mobile health (mHealth) apps to monitor and promote physical activity (PA) has become increasingly popular among individuals with various chronic conditions. However, such work remains limited among individuals with spinal cord injury (SCI), especially those who use a manual wheelchair for mobility.

objectives: The study aims to describe the development of the WheelFit mHealth app for monitoring and promoting PA in manual wheelchair users (MWUs) with SCI and evaluate its feasibility and usability in free-living conditions.

Methods: The WheelFit app, based on the Fogg Behavioral Model with inputs from stakeholders, including MWUs, physical therapists, and personal trainers, was developed to promote PA in MWUs. It works with two commercial wearable devices, that is, an Android smartwatch and a wheel sensor, which stream users' upper extremity and wheelchair movement to the app to calculate PA variables using custom algorithms. Users can set personal goals, review daily progress and PA history, and access an adaptive workout library within the app. A 4-week single-group pre-post study, consisting of a 1-week baseline and 3-week intervention phase, was conducted to evaluate WheelFit's feasibility and usability. Feasibility was evaluated using the session attendance rate, device and app usage, and implementation of action plans. Usability was assessed using the system usability scale. The preliminary effectiveness was assessed by comparing preintervention and postintervention PA variables and scores from the SCI exercise self-efficacy scale.

Results: A total of 16 participants completed the study protocol with 100% session attendance and maintained 14.2 hours of daily device and app connection. Participants demonstrated varying levels of adherence to their action plans. The excellent usability of WheelFit was indicated by an average system usability scale score of 81.8 (SD 19.2) points. Statistically significant increases between pre-post daily exercise times (preintervention: mean 26.4, SD 16.9 minutes; postintervention: mean 33.3, SD 24.9 minutes; P=.049) and exercise self-efficacy scale scores (preintervention: mean 33.9, SD 4.5 points; postintervention: mean 35.9, SD 3.2 points; P=.043) were observed.

Conclusions: The WheelFit app demonstrated promising feasibility, usability, and a positive impact on promoting PA in MWUs with SCI. Future investigation exploring the potential integration of the WheelFit app into clinical practice is warranted.

Keywords: community-based research; exercise; mHealth; spinal cord injury; telemedicine; telerehabilitation; workout.

Figure 1.. WheelFit devices. (A) Google pixel 3 smartphone, (B) Ticwatch Pro Android smartwatch, and (C) SensorTag wheel sensor.

Figure 2.. The WheelFit app.

Figure 3.. Study procedure. PT: physical therapist.

Figure 4.. Study participant diagram.