Feasibility and scalability of a fitness tracker study: Results from a longitudinal analysis of persons with multiple sclerosis

BACKGROUND: Consumer-grade fitness trackers offer exciting opportunities to study persons with chronic diseases in greater detail and in their daily-life environment. However, attempts to bring fitness tracker measurement campaigns from tightly controlled clinical environments to home settings are often challenged by deteriorating study compliance or by organizational and resource limitations. OBJECTIVES: By revisiting the study design and patient-reported experiences of a partly remote study with fitness trackers (BarKA-MS study), we aimed to qualitatively explore the relationship between overall study compliance and scalability. On that account, we aimed to derive lessons learned on strengths, weaknesses, and technical challenges for the conduct of future studies. METHODS: The two-phased BarKA-MS study employed Fitbit Inspire HR and electronic surveys to monitor physical activity in 45 people with multiple sclerosis in a rehabilitation setting and in their natural surroundings at home for up to 8 weeks. We examined and quantified the recruitment and compliance in terms of questionnaire completion and device wear time. Furthermore, we qualitatively evaluated experiences with devices according to participants' survey-collected reports. Finally, we reviewed the BarKA-MS study conduct characteristics for its scalability according to the Intervention Scalability Assessment Tool checklist. RESULTS: Weekly electronic surveys completion reached 96%. On average, the Fitbit data revealed 99% and 97% valid wear days at the rehabilitation clinic and in the home setting, respectively. Positive experiences with the device were predominant: only 17% of the feedbacks had a negative connotation, mostly pertaining to perceived measurement inaccuracies. Twenty-five major topics and study characteristics relating to compliance were identified. They broadly fell into the three categories: “effectiveness of support measures”, “recruitment and compliance barriers”, and “technical challenges”. The scalability assessment revealed that the highly individualized support measures, which contributed greatly to the high study compliance, may face substantial scalability challenges due to the strong human involvement and limited potential for standardization. CONCLUSION: The personal interactions and highly individualized participant support positively influenced study compliance and retention. But the major human involvement in these support actions will pose scalability challenges due to resource limitations. Study conductors should anticipate this potential compliance-scalability trade-off already in the design phase.