Regency EF
The Importance of Timestamping for Data Integrity in Studies Incorporating Mobile Health Data (304019)
*Vidhya Balasubramanian, Stanford University Quantitative Sciences Unit
Manisha Desai, Stanford School of Medicine
*Ariadna Garcia, Stanford University
Rebecca Gardner, Stanford University Quantitative Sciences Unit
Santosh Gummidipundi, Stanford University Quantitative Sciences Unit
*Haley Hedlin, Stanford University Quantitative Sciences Unit
Justin Lee, Stanford University Quantitative Sciences Unit
Ken W Mahaffey, Stanford University
Marco Perez, Stanford University
Mintu Turakhia, Stanford University
Keywords: timestamps, data integrity, wearable devices, mobile health study, pragmatic trial design
The collection of big data from wearable devices provides an opportunity to learn more about participants in the wild than would have been possible in traditional trials. To optimize utility of such data, investigators need to be mindful in designing studies so that dates and times are collected with minimal noise for ongoing continuous measurements of endpoints of interest. Dates and times of additional data collected outside of the wearable device (e.g., in the clinic or through the electronic health records) need to also be carefully considered to address goals. In a recent digital health study, there were multiple data streams collected on the study’s population throughout the study’s course with the intention of relating them to one another. In this study standardized timestamps were necessary for determining concordance between two devices, and in ensuring overall data integrity as well. We demonstrate the critical impact of collecting timestamp data for providing principal interpretation of findings and for maintaining integrity of the study data and illustrate the importance of timestamping in all studies that incorporate data from wearable devices.