A recent study in the Journal of Aging and Health showed that Omada’s digital diabetes management program helped those on Medicare to reduce their risk of diabetes with the mHealth digital coaching.

Omada is a popular, San Francisco-based company that uses an online and mobile platform to help the approximately 85 million users to prevent prediabetes. For reference, there are 86 million people in the U.S. who have prediabetes. And Omada was one of the first such digital entities to receive reimbursement from the federal government. Fast Company also named it as one of its Top 10 Most Innovative Companies.

Their technology uses personal behavioral and scientific data to help users avoid diabetes with small steps. Utilizing diet “gamification,” a wireless scale, community support, and a dedicated health coach, Omada focuses on giving its users accountability and insight. This study focused on a cohort of Humana Medicare patients who were given access to the Omada platform.

Humana has been in the news a lot lately thanks to a misfired merger with Aetna, blocked by a federal court, and subsequent layoffs that the healthcare company planned as a life raft of sorts if the merger didn’t pan out. In the digital health space, though, Humana has done some interesting things like using Goal Guru to help manage the health of their employees.

In this study, which included 501 patients, there was high participation (92%) and high engagement, with patients using the app as frequently as 19 times per week. And more importantly, they saw some decent results. Participants lost 7.5% of body weight, showed small improvements in A1c (0.14%), and reduced total cholesterol by 7 mg/dL.

Taking a step back though, there are some important points to note in the study. First, they reached out to nearly 10,000 patients but ultimately only about 600 were included here. So we have a highly selected group of participants in a purely observational study. Finally, the endpoints measured, including weight, were only captured during the 12-month program – the durability of these effects remains unknown.

While the results are encouraging, they shouldn’t be taken as proof of effectiveness. While it may be effective for motivated patients, we don’t know that it would help the other 9,500 potentially eligible patients who didn’t participate especially, if they are in some way different than the studied cohort.

Hopefully these findings will help motivate more rigorous evaluation in the future. That’s not to say proof of effectiveness in a large randomized controlled trial should be required for, say, an individual patient to sign up and use this platform. Rather, it would help us better understand how best to use this platform when working with different patients and could help guide future design changes to improve effectiveness.