DreaMed Simplifies Diabetes Care via AI-Driven Recommendations
This Israeli-born startup is making significant clinical and regulatory headway in their effort to automate care recommendations for people with diabetes. Their platform endo.digital processes mountains of diabetes-related data, from new and emerging devices and sources, so physicians can focus on patient care.
Investors, learn how you can back Health Transformers like Dr. Moshe Phillip, Eran Atlas, and Dr. Revital Nimri.
Challenge
Over the last decade we’ve witnessed an explosion in the availability of personal health data. Whether it comes through wearable fitness trackers, remote patient monitoring devices or in-hospital technology, the stream of data is getting deeper and wider by the day.
This increase in data volume opens up awesome opportunities for healthcare innovators to predict illness, diagnose disease and improve care. But it also creates new challenges. One place we’re seeing the two sides of this health data coin play out is in diabetes.
According to the IDF Diabetes Atlas, there are nearly 783 million people suffering from diabetes around the world. More and more of these individuals are gaining access to continuous glucose monitors (CGMs) that track glucose levels throughout the day, without the need for a finger stick. Just recently, US Medicare has decided to expand reimbursement for CGMs for people with Type 2 that uses basal insulin. Last week, the authorities in France made a similar decision. This is life-saving, game-changing technology, but it brings with it a wave of health data that can be overwhelming for both patients and their physicians to manage. And that is just one device and data source of many in the diabetes space.
On the physician side, the tidal wave of diabetes health data causes endocrinologists and their staff to spend so much time combing through digital platforms and information that they experience information overload. They then can end up giving less attention to their patients, leading to clinician burden, frustration, and concern for patient wellbeing.
As access to diabetes devices expands, data overload also becomes a problem for primary care physicians. These doctors aren’t diabetes specialists, so the web of new diabetes data sources is even more overwhelming. As a result, these doctors end up sending individuals back to an endocrinologist — of which there are too few — who now have a backlog of patients with data they need to wade through.
In the end, an incredibly valuable well of data can turn from an opportunity for healing to a burden. Thankfully, there are dedicated health innovators working on a solution. DreaMed, a team based in Israel and MN, is using artificial intelligence to make sense of disparate health data streams and offer care pathway recommendations at the click of a button.
Origin Story
Professor Moshe Phillip, MD, is the Director of the Institute of Endocrinology and Diabetes at the National Center for Juvenile Diabetes at Schneider Children’s in Israel. Over the years he noticed how innovative technology companies were collecting and disseminating so much digital diabetes data that it was becoming overwhelming for physicians and actually taking attention away from treatment and quality patient care.
Dr. Phillip knew what was good for his young patients, and that there had to be a better way to access and focus this data. He decided to create his own tech team and build a solution.
The professor reached out to Eran Atlas, a biomedical engineering student at Tel Aviv University, for help. The two, together with Revital Nimri, MD, from Schneider Children’s Institute and other team members, focused on the development of an artificial pancreas system suitable for people with Type 1 diabetes. Together with their partners, Prof Thomas Danne and Prof Tadej Battelino, their team was the first in the world to take a group of patients into a hotel, simultaneously in three different countries, give them laptops, and monitor the use of an algorithm to automate their insulin delivery while they slept. This initial study was published in the New England Journal of Medicine — the first such findings to be shared in the publication. Eventually this solution was sold to Medtronic and is part of MiniMed 780G.
Then they considered what more they could do for people with diabetes, knowing that the vast majority of the population would not be able to use artificial pancreas systems. Their vision was focused on how they could take that large amount of digital data and automate it to reflect how physicians actually think: treatment, recommendations, even a way to incorporate billing all in one place through the use of artificial intelligence.
In 2015, the team began work on an end-to-end automation of the care pathway with this mountain of diabetic data. They didn’t just want to collect information or even interpret it but develop a platform that did both in an accessible, manageable way and was also able to digitally share physician recommendations and treatment plans between doctor and patient. If they could accomplish this, the team believed they could also incorporate automated billing to capture the appropriate codes for the different clinician services provided to make a seamless experience. They established DreaMed and hit two key milestones that prompted them to keep moving forward.
First, they published a multicenter study that proved they could achieve the same clinical results with their algorithm and digital platform as doctors in Joslin Diabetes Center and Yale New Haven Health System. This study was followed by additional clinical studies and real-world clinical data, showing the similarity between the recommendations of their platform — called endo.digital — and expert physicians. Next, they were the first in the world to receive FDA clearance to use continuous glucose sensor data and provide a recommendation for insulin therapy. This was new territory for the FDA, so DreaMed worked with the agency to create a new product code. Since that time, the company has received three more FDA clearances.
Under the Hood
The DreaMed product is called endo.digital, and on the website the first thing you read is, “We treat the data, you treat the person.” The goal of the digital experience is to “make the digitally complex simple” through the AI engine, and it works with the user in whatever way fits their needs.
The solution is elegantly simple to use. The clinician opens their EMR user account and goes into the appropriate patient chart. There is a button that reads: “Give Me a Recommendation.” The provider clicks on the button, and a recommendation comes up based on the available data the AI has collected. The healthcare provider reviews the recommendation and edits as needed, then clicks “Approve.” The visit summary is auto populated, a notification is sent to the patient via mobile or email, and the app calculates their insulin needs.
DreaMed tells doctors they can decrease the number of process steps without having to engage with more technology or increase their data management. The company can cut a process from 30 steps down to 12. What once took 25 minutes to complete could now take nine minutes less — and it would include automating all billing.
Data is pulled two different ways: cloud to cloud if the patient device is set up for that, or old-school USB cable plug-in to a provided endo.digital device uploader that then sends the data to the cloud. The AI is agnostic as to how it gets the information.
On top of the randomized controlled study published in Nature Medicine, DreaMed conducted a study with a set of patients spread across 27 physicians across the world. The company asked how they would treat a particular data set of patients. When physicians conferred with each other, the level of agreement for treatment was between 42–60%. However, AI-recommended treatment received a 67–70% agreement. In other words, when blinded, physicians trusted the DreaMed platform more than their own colleagues.
StartUp Health’s Take
DreaMed is building relationships with some exciting partners with an eye toward optimizing the care provided by their AI. Their first customer was Boston Children’s Hospital, and they’ve established a strong relationship with the Billings Clinic and Yale New Haven Health System. The company is also expanding to evaluate the solution in primary care physicians through their work with Barbara Davis Center for Diabetes in Colorado Springs, CO. DreaMed partners are willing to help them collect the necessary data to keep abreast of marketplace trends, which makes it easier to gain the trust of potential investors.
DreaMed is acutely aware of the complications of those dealing with diabetes — for example 66% of the Type 2 diabetes population also have hypertension and 23% have a chronic heart disorder. The company sees a big future where they expand their AI-driven health recommendations to new verticals, like “nephro-digital” for kidney health then moving to “cardio-digital” for heart problems and hypertension. They believe their platform can address a range of verticals and within a year or two, incorporating the care of diabetics ingesting oral medicine in addition to syringe and pump. Proving themselves in T1D opens the door to support other health needs for the population.
Join us in welcoming Eran Atlas and the DreaMed team to the StartUp Health community.
→ Connect with DreaMed via email
Call for T1D Innovation
Are you a scientist or innovator focused on T1D innovation who would benefit from education about how to navigate and build a company that will be successful in attracting mission-aligned capital, customers, and collaborators to pursue scientific discoveries in the field of Type 1 diabetes? Learn more and apply for a T1D Fellowship.
Become a Health Moonshot Champion
Health moonshots are fueled by passionate families, foundations, and industry organizations committed to achieving health moonshots. Learn how you can join other champions of the T1D Moonshot or one of our other global health moonshots.
Funders: Learn how you can become a Health Moonshot Champion and invest in Health Transformers.
Founders: Don’t make the journey alone. Learn how Health Transformer University fuels your health moonshot.
Follow us on social media for daily updates on Health Transformers: Twitter, LinkedIn, Facebook, and Instagram.
Published: Jun 23, 2023