Objective: To evaluate whether a mobile health application that employs elements of social game design could compliment medical care for unresolved concussion symptoms. Design: Phase I and Phase II (open-label, non-randomized, ecological momentary assessment methodology). Setting: Outpatient concussion clinic. Participants: Youth, aged 13–18 years, with concussion symptoms 3+ weeks after injury; Phase I: n = 20; Phase II: n = 19. Interventions: Participants received standard of care for concussion. The experimental group also used a mobile health application as a gamified symptoms journal. Outcome measures: Phase I: feasibility and satisfaction with intervention (7-point Likert scale, 1 high). Phase II: change in SCAT-3 concussion symptoms (primary), depression and optimism. Results: Phase 1: A plurality of participants completed the intervention (14 of 20) with high use (110 +/− 18% play) and satisfaction (median +/− interquartile range (IQR) = 2.0+/− 0.0). Phase II: Groups were equivalent on baseline symptoms, intervention duration, gender distribution, days since injury and medication prescription. Symptoms and optimism improved more for the experimental than for the active control cohort (U = 18.5, p = 0.028, effect size r = 0.50 and U = 18.5, p = 0.028, effect size r = 0.51, respectively). Conclusions: Mobile apps incorporating social game mechanics and a heroic narrative may promote health management among teenagers with unresolved concussion symptoms.

Postural control provides insight into health concerns such as fall risk but remains relatively untapped as a vital sign of health. One understudied aspect of postural control involves transient responses within center of pressure (CoP) data to events such as vision occlusion. Such responses are masked by common whole-trial analyses. We hypothesized that the transient behavior of postural control would yield unique and clinically-relevant information for quiet stance compared to traditionally calculated whole-trial CoP estimates. Three experiments were conducted to test different aspects of this central hypothesis. To test whether transient, epoch-based characteristics of CoP estimates provide different information than traditional whole-trial estimates, we investigated correlations between these estimates for a population of young adults performing three 60-second trials of quiet stance with eyes closed. Next, to test if transient behavior is a result of sensory reweighting after eye closure, we compared transient characteristics between eyes closed and eyes open conditions. Finally, to test if there was an effect of age on transient behavior, we compared transient characteristics during eyes closed stance between populations of young and older adults. Negligible correlations were found between transient characteristics and whole-trial estimates (p>0.08), demonstrating limited overlap in information between them. Additionally, transient behavior was exaggerated during eyes closed stance relative to eyes open (p<0.044). Lastly, we found that transient characteristics were able to distinguish between younger and older adults, supporting their clinical relevance (p<0.029). An epoch-based approach captured unique and potentially clinically-relevant postural control information compared to whole-trial estimates. While longer trials may improve the reliability of whole-trial estimates, including a complementary assessment of the initial transient characteristics may provide a more comprehensive characterization of postural control.