Approach to the Postmarket Evaluation of Consumer Wearable Technologies.

Importance Consumer wearable technologies have wide applications, including some that have US Food and Drug Administration clearance for health-related notifications. While wearable technologies may have premarket testing, validation, and safety evaluation as part of a regulatory authorization process, information on their postmarket use remains limited. The Stanford Center for Digital Health organized 2 pan-stakeholder think tank meetings to develop an organizing concept for empirical research on the postmarket evaluation of consumer-facing wearables. Observations The postmarket evaluation of consumer wearables involves broad consideration of an individual consumer's journey from acquisition, intended and unintended use of the wearable, and access to health care resources on receipt of a notification. For individuals who do access the health care system, a wearable's downstream effects can be studied through appropriate clinical evaluation, delivery of guideline-directed treatments, shared decision-making in areas of clinical equipoise, and analysis of clinical end points and patient harms. Effective postmarket research draws from denominators appropriate to the clinical question, with clearly defined parameters for success and failure. Generalizability related to data completeness and reliability should also be considered. As patients increasingly integrate wearables into their health monitoring, cross-platform data sharing with a focus on privacy and data quality can drive patient-centered innovation and identify opportunities to bridge gaps in medical care. Relevance The think tank identified priorities in postmarket research, comprising the journey from consumer to patient and accounting for patient, clinician, health care delivery system, and societal impacts of consumer wearables. Overall, this approach serves not only to organize the study of consumer wearables but also to act as a guidepost for using real-world data in postmarket research.