Study Watch – Another step closer to powering precision health

Publication validates the use of the Verily Study Watch in real-world settings

In 2017, we introduced the Verily Study Watch as part of our efforts to create new tools for unobtrusive biosensing, including those that may reliably produce research-grade measurements of physical activity in real-world settings. This is important because physical activity can predict health outcomes for those with chronic conditions, but measurements done within a lab or clinic only tell part of the story.

Today, we are excited to announce the publication that shares data demonstrating the analytical validity of an ambulatory status algorithm from the Study Watch. This evidence demonstrates that Verily Study Watch can reliably detect walking activity in daily life across diverse research populations.

The setup. Researchers utilized passively recorded motion sensor data from study participants in their usual, free-living settings. A large portion came from participants in the Project Baseline Health Study who provided self-reported labels of time periods while walking or remaining still. In addition, a smaller study was conducted whose participants wore an additional ankle device that monitored their ambulatory status over the course of several days.

The results. The algorithm was trained using data from both studies to enable deeper characterization of its performance. Testing showed that the algorithm was accurate in classifying ambulatory status both within 10-second periods, and on daily aggregated metrics (i.e., total daily ambulatory time). Furthermore, when results were broken down by subgroups in the demographically diverse Project Baseline Health Study cohort, no significant differences in performance were seen across groups stratified by sex, age, or race. Thus, the authors concluded the algorithm can accurately classify ambulatory status using a wrist-worn device in real-world settings with generalizability across the entire study population.

The impact. With this algorithm, researchers may now use Verily Study Watch to reliably measure certain physical activity in diverse populations outside of the clinic. Not only is this work an important milestone as the industry continues to expand representation in clinical research, it also serves to progress the field of digital biomarkers. It’s all just another way Verily is working to develop technologies that power more precise clinical research now for more precise health interventions in the future.

For more Verily Study Watch insights and information, connect with us by contacting flahertyt@verily.com.