At the cutting edge of digital health you’ll find health apps being integrated with traditional health processes to create hybrid digital health systems. Among these digital health systems diabetes stands out as one of the first chronic conditions to demonstrate the benefits of marrying digital health with established health practice.
Early indications from randomized controlled trials of diabetes digital health systems are hopeful. They point to the qualities that make diabetes ripe for integrating digital health technology into daily management. We’re beginning to see some positive results for people who are affected by diabetes. And we’re beginning to understand some of the characteristics that make diabetes a good candidate for a digital health approach.
Diabetes dominates digital health apps
Worldwide, the top app stores contain over 318,000 digital health apps. Of these, IQVIA reports in “The Growing Value of Digital Health,” 40% are focused on managing a health condition, as opposed to being focused on wellness. Additionally, 16% of the health-condition apps (roughly 280 apps) are specific to diabetes. Only mental health and behavior disorders had a larger share of health-condition apps at 28%.
Clearly, the developers of digital health apps see potential in applying their technology to diabetes care or it wouldn’t take up such a significant share of the available catalog.
What makes diabetes well-suited for a digital health care approach?
Several key aspects put diabetes among the chronic health conditions that are uniquely suited for digital health applications.
With diabetes there are agreed-upon, specific, scientifically measurable indicators for treatment and management. Blood glucose readings, along with A1c results, embody the accepted standard for measuring the efficacy of diabetes treatment and management. Because these standards are in place, the state of a person’s blood glucose management can be measured and analyzed mathematically. Blood glucose readings and the severity of hypoglycemic episodes can be categorized across individuals. For app developers, this means that the bulk of what goes into managing diabetes on a daily basis can be expressed through mathematical algorithms.
People living with diabetes have the means to measure and monitor those indicators as part of their daily routine. Access to blood glucose data is widespread, from either a Blood Glucose Meter (BGM) or a Continuous Glucose Monitor (CGM). Despite cost and coverage barriers that limit the amount of checking a person can afford to do, people with diabetes still have personal access to a reliable, standardized, scientific measure of the state of their diabetes. Not every chronic health condition has a way to monitor its status outside the doctors office. For app developers, this means that there’s a reliable source of health data that can be collected.
The most commonly used indicators for diabetes management respond quickly to treatments and interventions. For example, blood glucose levels move within minutes or hours of a dose of insulin, or eating quick-acting sugar, or engaging in physical exercise. This immediacy makes it possible for diabetes health apps to provide meaningful feedback and guidance to people with diabetes and their caregivers. They can reliably measure the outcome of their efforts to bring their blood glucose levels up or down within a relatively short time. For app developers, this means that they can build meaningful feedback loops into their algorithms and expect that this feedback will drive positive outcomes.
Digital health care is being integrated into everyday diabetes care
Particularly illuminating against this backdrop are preliminary reports on two ongoing diabetes digital care studies presented at the IDF Congress 2019 in Busan, South Korea.
Dr. Young Min Cho, MD, PhD, from the Seoul National University College of Medicine, gave status on an ongoing study of people with type 2 diabetes using an Android app for logging. The app is part of a “smartphone-based, patient-centered care system” that he referred to as mDiabetes. It includes mechanisms for monitoring glucose readings, physical activity, and diet along with providing an online social network and clinical decision support system.
Dr. Soo Lim, MD, PhD, from the Department of Internal Medicine, Seoul National University College of Medicine, gave status on an ongoing study that is implementing what he called a Ubiquitous health care System (uhealth care) for people with type 2 diabetes. uhealth care, is much like the care system in Dr. Cho’s study, except that participants also received feedback via automated text messages. Based on the reported blood glucose readings, the uhealth care system sends participants tailored, in-real-time feedback via text message. These text messages are generated by software algorithms developed by Dr. Lim’s team.
Both doctors shared that they are seeing initial positive results from their studies. Speaking in general terms, Dr. Cho and Dr. Lim mentioned that A1c results have been reduced, fewer hypoglycemic episodes were experienced, and weight gain was avoided. All of these early indications point toward promising future results.
These are the kind of results that have been demonstrated by other diabetes health apps. Published studies of diabetes digital apps have been summed up by IQVIA in this way:
“[D]iabetes apps have consistently delivered statistically and clinically significant HbA1c improvements...in type 2 diabetics, with greater benefits to T2D than T1D, younger patients rather than older patients, and patients who received health care professional feedback via the app vs. those who did not.”
It’s remarkable that such a definitive statement can be made about the effectiveness of diabetes digital health apps, especially because there’s such a scarcity of published studies on health apps overall. IQVIA reports that only 571 efficacy studies were published between 2007 and 2017 for all of the roughly 127,000 health-condition focused apps available.
Where will diabetes digital health care go next?
We’re seeing positive indicators for the potential efficacy of diabetes digital health. But we are far from done developing the digital health systems that meet all of the demands of daily diabetes management.
Both Dr. Cho and Dr. Lim acknowledged that the systems they are working on still need improvement. They frame the results they presented as preliminary and talk about needing further development before the algorithms and systems can be more widely used for daily diabetes management.
But even at this early stage of their studies Dr. Cho and Dr. Lim point to a number of key considerations that can either help ensure or scuttle the success of future diabetes digital health systems.
Health apps do not operate in isolation. Strikingly neither Dr. Cho or Dr. Lim talked about using their diabetes digital health apps in isolation. The apps they use are integrated into a larger health care system. And that system includes traditional approaches to therapies and personal contact with health care professionals. They acknowledged that some fear that digital health will eliminate the need for medical professionals. Neither doctor expects this to happen.
AI has the potential to take on the most complex aspects of diabetes care. Through the use of AI more and more complex algorithms can be generated. Potentially this will lead to further automating some of the most complex aspects of diabetes care, like calculating real-time insulin dosing in a way that takes into account more than just the person’s current blood glucose reading. Dr. Cho or Dr. Lim acknowledged that this application of AI opens up the possibility for truly personalized care and in-real-time feedback.
AI also has the ability to analyze large groups of data quickly. For the individual this may mean a much more complete and clear picture of their current state of health. For population studies this points to better understanding of community health and the impacts of social and physical environments.
Barriers remain to the widespread adoption of diabetes digital health systems. Beyond access and affordability, the regulatory environment across the world stands in the way of widespread adoption of digital health systems. Ironically, Dr. Cho and Dr. Lim stated that the digital health systems they are working on are currently prohibited in South Korea because of concerns about the potential for a data breach and loss of privacy. This barrier exists, despite these systems receiving approval for study from South Korea’s equivalent agency to the FDA.
Diabetes serves as a proof of concept for other health conditions
Both Dr. Cho and Dr. Lim talk about the need for further development in both the human and software systems that make up the digital diabetes care systems that they are working on.
Dr. Lim looks beyond diabetes and sees the potential to apply similar digital health systems to other health conditions. He specifically mentioned asthma, cognitive health, wound care, and telemedicine as potential candidates. For all these, and maybe even more, health conditions diabetes digital health systems can serve as the proof of concept that spurs further development and implementation of digital health systems.
Bigfoot Perspective
“We reduce the burden of daily diabetes management from the individual when we think of building digital health systems instead of focusing on a single digital device or app. With such a system the work that goes into daily diabetes management is shared among the person with diabetes, their medical team, and devices.”
—Jeffrey Brewer