Promoting Adherence to Joint Exercise Using the Donation Model: Proof via a Motion-Detecting Mobile Exercise Coaching Application

Abstract

Purpose: Maintaining or increasing user adherence to digital healthcare services is of great concern to service providers. This study aims to verify whether the donation model is an effective strategy to increase adherence to physical exercise using a mobile application.

Materials and methods: A total of 5618 users of a motion-detecting mobile exercise coaching application participated in a donation or self-reward exercise challenge with the same exercise protocol. The workout consisted of 50 squats daily for 14 days. The user's exercise was monitored through a smartphone camera, providing real-time visual and audio feedback. In the donation group, 6 USD was donated to the economically disadvantaged if a participant completed their workout each day. In the self-reward group, three people who completed the program and 20 people who completed ≥12 days of exercise were randomly selected and provided with goods worth 60 USD and 4.3 USD of online currency, respectively.

Results: The average daily exercise completion rate (% of participants who completed daily exercise) in the donation group was 1.8 times higher than that of the self-reward group (donation, 41.7%; self-reward, 22.7%; p<0.0001). The donation group completed more days of the program (donation, 5.8; self-reward, 3.2; p<0.0001). The completion rate of both groups decreased with time and decreased most on day two (donation, -9.9%; self-reward, -14.5%).

Conclusion: The donation model effectively promoted adherence to mobile app-based exercise. This donation model is expected to effectively enhance user adherence to digital healthcare services.

Keywords: Machine learning; marketing; mobile application; motion; telehealth.

Fig. 1. A machine learning-based motion-detecting mobile exercise coaching application (MDMECA).

Fig. 2. Intergroup differences in the daily exercise completion rate and the average number of exercise completion days. The exercise completion rate (A) in the donation group was higher than that of the self-reward group for each exercise day. The average number of exercise completion days (B) was also 1.8 times higher in the donation group than in the self-reward group.

Fig. 3. Individual exercise completion days in two groups. The total number of participants who completed the workout on all 14 days was 286 (12.3%) in the donation group and 114 (3.5%) in the self-reward group. The proportion of those who did not exercise for even a day among the participants was 1.9 times higher in the self-reward group (41.8%) than in the donation group (22.0%). In the self-reward group, we confirmed that more than half of the participants exercised less than 3 days (64.4%).

Fig. 4. Subgroup analysis of exercise completion rates. In subgroup analyses conducted according to the challenge slogan, all subgroups participating in the donation challenge showed significantly higher daily exercise completion rates than those in the self-reward group (A). The average exercise completion rates according to the challenge slogan are visualized in ascending order (B).