Balance Assessment Using Gamified Digital Technology in Community-Dwelling Older Adults: Mixed Methods Validation Study and Randomized Controlled Trial.

Background: Falls cause injury and mortality among older adults, necessitating reliable, scalable, engaging balance assessment tools to use in community settings. Traditional clinician-administered assessments like the Brief Balance Evaluation Systems Test (Brief-BESTest) are limited by subjectivity and accessibility constraints. Computer vision-based digitalization combined with gamification may address these limitations; yet, validation evidence remains limited. Objective: We (1) digitalized and validated a computer vision-based Brief-BESTest against clinician scoring and (2) investigated whether a gamified interface improves older adults' user experience during balance assessment, without compromising assessment performance. Methods: This mixed methods study comprised (1) a concurrent validity substudy in a convenience subsample (n=10) and (2) a parallel-group randomized controlled trial (RCT; n=30) with 1:1 allocation. Participants were community-dwelling older adults aged ≥60 years recruited from Hongqi Community, Shanghai, through community announcements and health care worker referrals. Phase 1 (n=10; mean age 64.9, SD 2.76 years) evaluated concurrent validity of a computer vision-based digitalized Brief-BESTest using OpenPose skeletal tracking (Carnegie Mellon University Perceptual Computing Lab; Logitech Brio 4K webcam, 27-inch touchscreen) against the clinician-administered version. Phase 2 (n=30; mean age 66.7, SD 3.93 years) used a parallel-group RCT with 1:1 coin-flip allocation. Primary outcome measures include perceived exertion (Borg Rating of Perceived Exertion scale 6-20), intrinsic motivation (Intrinsic Motivation Inventory 7-point Likert, including interest and enjoyment, perceived competence, and pressure and tension subscales), and intention to continue use (7-point Likert scale). Semistructured interviews (mean 4.8 minutes) assessed engagement factors. Data collection occurred in a controlled indoor setting with safety railings. Results: Phase 1 demonstrated excellent intrasession reliability (intraclass correlation coefficient=0.89-0.92) and strong concurrent validity (Spearman ρ=0.91; 95% CI 0.68-0.98; P<.01), with no significant mean difference (MD; paired t test: MD 0.23; P=.77; d=-0.07). In phase 2, Gamified Digital Balance Assessment (GDBA) users reported significantly lower perceived exertion (Mann-Whitney U: MD -2.67; 95% CI -4.60 to -0.74; P=.01; d=-1.08), higher enjoyment (MD 1.53; P=.009; d=1.17), higher perceived competence (MD 1.14; P=.02; d=0.89), and higher intention to continue use (MD 1.66; P=.001; d=1.25). Pressure and tension (P=.09; d=0.63) showed no significant difference. Thematic analysis (Cohen κ=0.68) identified 2 themes: motivational rewards (80% cited real-time feedback) and perceived usability (87% emphasized avatar demonstrations). Conclusions: This study validated a computer vision-based digital Brief-BESTest and experimentally tested a gamified interface for balance assessment in community-dwelling older adults. Unlike prior work focused largely on single-task digital tests or nongamified interfaces, the GDBA integrates comprehensive, clinically grounded balance assessment with evidence-based gamification tailored to older users. These findings advance digital geriatric assessment by demonstrating that gamified designs can enhance motivation, perceived competence, and tolerability of testing without sacrificing measurement quality. If replicated in longitudinal and real-world settings, such systems could provide scalable, low-cost tools for routine fall-risk screening, self-monitoring, and targeted preventive interventions in community and primary-care environments.