Care Evolution

New Nature Medicine publication showcases a scalable, remote trial model integrating wearable, genomic, and EHR data

FOR IMMEDIATE RELEASE
Contact: [email protected]

ANN ARBOR, MI – [July 31, 2025] A new study published in Nature Medicine showcases how a digitally enabled, participant-centered research model can support high-quality AI-driven discovery—revealing hidden diabetes risk using wearable and real-world data. The PRediction Of Glycemic RESponse Study, or PROGRESS, led by the Scripps Research Digital Trials Center and powered by the CareEvolution® MyDataHelps™ app, enrolled over 1,000 participants across the U.S. using a fully digital approach, one that scaled quickly, enabled continuous participant engagement, and supported data-rich, multimodal research.

“We built our MyDataHelps infrastructure to support researchers in conducting faster, more cost-efficient studies without compromising scientific quality,” said Vik Kheterpal, Principal at CareEvolution. “This is about enabling a more scalable, participant-centered model for high-quality research.”

Participants used the PROGRESS study app built on the MyDataHelps platform. Study components such as eligibility screening, electronic consent, surveys, and connecting digital health technologies and electronic health records were completed remotely. They collected biological samples at home, including saliva, blood, and stool, and received wearable sensors (a continuous glucose monitor and an activity tracker) to measure glucose dynamics, sleep, and physical activity. All participant communications and logistics were managed digitally.

“We demonstrated that it’s possible to conduct rigorous, high-quality research at a national scale without traditional site-based infrastructure,” said Ed Ramos, PI of the study and Chief Science Officer at CareEvolution. “Participants were able to contribute from anywhere, at their convenience, and researchers were able to collect rich, multimodal data efficiently and securely.”

The study leveraged SMART on FHIR–based, patient-mediated EHR connectivity, enabling the researchers to access clinical history as well as longitudinal health records. Nearly 700 participants connected to their electronic health records from over 415 distinct healthcare organizations, including many participants with data spanning multiple provider systems. The transformation of raw EHR data into the OMOP common data model enabled scaled, real-world data analysis. 

The study’s key finding: an AI model trained on continuous glucose monitor data, microbiome profiles, diet, physical activity, and resting heart rate could detect subtle metabolic differences not captured by traditional HbA1c lab tests. This risk model revealed meaningful distinctions among prediabetic individuals with the same HbA1c levels, enabling earlier identification of those on a trajectory toward type 2 diabetes.

The PROGRESS study also serves as a milestone in improving access to research. Its digital-first design enabled participation from individuals nationwide spanning diverse geographic, socioeconomic, and age groups.

“This is about designing systems that reduce barriers and make participation easier—regardless of location, schedule, or background,” said Katie Baca-Motes, co-author on the study. “With the right incentives and digital tools, we can enable broader participation in clinical research across the country.”

Edward Ramos, PhD, and Katie Baca-Motes, MBA, also hold appointments at the Scripps Research Digital Trials Center. 

About CareEvolution

CareEvolution provides flexible, secure digital health infrastructure to support clinical research and care innovation. The company’s MyDataHelps platform enables participant-mediated data collection, remote study workflows, and integration of electronic health records, wearable sensor data, and biospecimens. Academic researchers, healthcare institutions, and industry sponsors use CareEvolution tools to design, deploy, and scale remote and hybrid studies across therapeutic areas. Learn more at www.careevolution.com.