JMIR Publications

ABSTRACT

Digital health technologies such as smartphones and wearable devices promise to revolutionize disease prevention, detection, and treatment. Recently, there has been a surge of digital health studies where data is collected through a Bring-Your-Own-Device (BYOD) approach, in which participants who already own a specific technology may voluntarily sign up for the study and provide their digital health data. BYOD study design accelerates the collection of data on a larger number of participants than cohort design because researchers are not limited in the study population size based on the number of devices afforded by their budget. However, the BYOD study design may not support collecting data from a representative random sample of the target population where digital health technologies are intended to be deployed. This may result in biased study results and biased downstream technology development. In this viewpoint, we describe demographic imbalances discovered in existing BYOD studies, including our own, and we propose a Demographic Improvement Guideline to offset these imbalances.