Diabetes Digital Health, Telehealth, and Artificial Intelligence

By David C. Klonoff, David Kerr and Juan Espinoza

Diabetes Digital Health, Telehealth, and Artificial Intelligence explains how to develop and use the emerging technologies of digital health, telehealth, and artificial intelligence to address this important public health problem to deliver new hardware, software, and processes. The book explores trends in developing and deploying the three most important emerging technologies for diabetes: digital health, telehealth, and artificial intelligence. This book is essential to clinicians, scientists, engineers, industry professionals, regulators, and investors, offering the tools that will be used to create the next generation products to support a precision medicine approach to manage diabetes. According to the CDC, in the US there are 37 million people with diabetes and 96 million people with prediabetes. Diabetes triples the risk of myocardial infarction and stroke and is the leading cause of blindness, end stage renal failure, and amputations. The management of diabetes is becoming increasingly dominated by digital health tools consisting of wearable sensors, mobile applications providing decision support software, and wireless communication tools. Digital health provides new data streams that can be combined to create unique approaches for diabetes based on a precision medicine paradigm.

• Includes Artificial intelligence (AI) data for the prediction, diagnosis, treatment, and prognostication for diabetes as a model disease
• Describes the most important issues of our time that comprise the most important technologies currently being applied to diabetes
• Presented in a consistent easy to help those new to the field understand and compare/contrast various elements of digital health, telehealth, and artificial intelligence for diabetes

• Language: English
• Publisher: Academic Press
• Release date: Jun 14, 2024
• ISBN: 9780443132438

Book preview

Part I

Digital health

Outline

Chapter 1. Trends in Digital Health for Diabetes

Chapter 2. Using Digital Health Tools in Medical Practice

Chapter 3. Diabetes Digital Health in the Hospital

Chapter 4. Digital Pharmacy for Diabetes

Chapter 5. Digital Health and Pharmacoadherence

Chapter 6. Food Recognition and Nutritional Apps

Chapter 7. Accessing and Acting Upon Patient-Generated Health Data

Chapter 8. Cybersecurity of Digital Health Tools

Chapter 9. The Role of Digital Health in Tackling India’s Diabetes Epidemic

Chapter 10. Investment Opportunities in Diabetes Digital Health

Chapter 1: Trends in Digital Health for Diabetes

Sang Youl Rhee ¹ , ² , and Eun Jung Rhee ³       ¹Department of Endocrinology and Metabolism, Kyung Hee University College of Medicine, Seoul, South Korea      ²Center for Digital Health, Medical Science Research Institute, Kyung Hee University Medical Center, Seoul, South Korea      ³Division of Endocrinology and Metabolism, Department of Internal Medicine, Kangbuk Samsung Hospital, Sungkyunkwan University School of Medicine, Seoul, South Korea

Abstract

Chronic metabolic disorders, including diabetes and obesity, represent a significant and growing global health problem, characterized by high prevalence rates and a heavy socioeconomic burden. These conditions require ongoing lifestyle assessment and management, a requirement that exceeds the capacity of current healthcare services, revealing pressing unmet needs. As a result, the adoption of digital health methodologies, supported by advanced information and communication technologies, is expected to revolutionize the therapeutic landscape and management protocols of these pervasive metabolic diseases. This article explores a spectrum of emerging digital health technologies and their transformative role in the field of diabetes and metabolic disease management.

Keywords

Artificial intelligence; Blood glucose self-monitoring; Diabetes mellitus; Digital technology; Mobile application; Obesity; Telemedicine; Wearable electronic devices

Summary

• Driven by advances in Information and Communication Technology, there has been an active pursuit of digital health solutions designed to integrate these technological elements into routine clinical care.

• Currently, contemporary technological advances such as mobile applications, wearable devices, continuous glucose monitors, telemedicine, and artificial intelligence are being used to effectively treat and manage diabetes patients, with a growing body of evidence supporting their utility.

• However, there is a lack of robust research on these emerging digital health technologies. There is an urgent need for rigorous validation studies to determine the comparative effectiveness of digital health services against the traditional standard of care outlined in current clinical practice guidelines.

Abbreviations

AGP    Ambulatory Glucose Profile

AHRQ    Agency for Healthcare Research and Quality

AI    Artificial intelligence

CGM    Continuous glucose monitoring

DALYs    Disability-Adjusted Life Years

GERD    Gastroesophageal reflux disease

HbA1c    Glycated hemoglobin

ICT    Information and Communication Technology

RCTs    Randomized controlled trials

SMD    Standardized mean difference

Statistics

• In 2021, the number of individuals living with diabetes worldwide is estimated to be 536.6 million, accounting for approximately 10.5% of the total population. In particular, the prevalence of diabetes among the adult population in Korea, especially those aged 30 years and above, is estimated to be 16.7%, a figure that exceeds the global prevalence rate.

• In a recent study conducted, it was observed that ICT interventions were associated with a decrease in HbA1c levels in patients with type 1 diabetes (standardized mean difference (SMD), −0.05; 95% confidence interval (CI), −0.08 to −0.02), type 2 diabetes (SMD, −0.06; 95% CI, −0.07 to −0.05), and gestational diabetes (SMD, −0.07; 95% CI, −0.10 to −0.03).

• According to records from the Ministry of Health and Welfare of the Republic of Korea, from February 2020, when telemedicine services were first allowed, through January 2023, a remarkable 36.6 million telemedicine services were provided by 25,697 healthcare institutions, benefiting a total of 13.8 million individuals. Despite the program's initial launch under the banner of COVID-19, the majority of these services were actually extended to patients with chronic diseases, including diabetes.

Introduction

The advent of Information and Communication Technology (ICT) has ushered in an era of dramatic change, affecting the fundamental way humans conduct their lives. This paradigm shift was first triggered by the ubiquitous use of personal computers and later accelerated by the rapid proliferation of smartphones. The core elements of ICT have transcended the boundaries of our daily lives to permeate medical and healthcare services. We are now on the cusp of an unprecedented revolution in healthcare.

Current Status

Diabetes is one of the most prevalent chronic diseases in the world today. By 2021, an estimated 536.6 million adults worldwide will be living with diabetes, representing approximately 10.5% of the global population. ¹ In Korea, the prevalence of diabetes is even higher, estimated at 16.7% among adults aged 30 years and older, which is higher than the global average. ² Furthermore, diabetes is identified as the chronic disease with the highest disease burden in Korea in terms of Disability-Adjusted Life Years (DALYs). ³ Diabetes can have a disproportionate impact, especially for individuals facing health disparities.

Effective treatment and management of diabetes requires continuous assessment and management of individuals' daily lifestyle habits, including diet and exercise, in addition to medical services provided by healthcare professionals. Systematic analysis of these data can help optimize an individual's glycemic control and prevent chronic complications associated with diabetes. However, traditional healthcare services are often episodic and tied to hospital or outpatient clinic visits, limiting the ability to achieve optimal outcomes. As a result, a variety of ICT-enabled digital health technologies that can help collect and assess data about an individual's daily life are gaining significant attention as potential allies in improving the health outcomes for people with diabetes. Table 1.1 presents representative digital health-related factors that can be used in diabetes and metabolic diseases.

Barriers

The effectiveness of mobile applications in improving clinical outcomes in diabetes has yet to be robustly demonstrated. An Agency for Healthcare Research and Quality (AHRQ) report found that only a minimal fraction of applications demonstrated significant improvements in glycated hemoglobin (HbA1c), a key measure of clinical diabetes outcomes. ⁴ None of the studies showed long-term effects on patients' quality of life, blood pressure, weight, or body mass index. Many of the intervention studies had methodological problems, including short study duration, inconsistent reporting of randomization, masking and dropout analysis, and frequent use of cointerventions, which made interpretation of the results difficult. None of the included trials were considered to be of high quality.

Table 1.1

Reproduced from Rhee SY, Kim C, Shin DW, Steinhubl SR. Present and future of digital health in diabetes and metabolic disease. Diabetes Metab J. 2020;44:819–827, under the Creative Commons Attribution Non-Commercial License (CC-BY, 4.0, https://creativecommons.org/licenses/by-nc/4.0/).

In a recent umbrella meta-analysis conducted by our group, the number of studies meeting stringent inclusion criteria was limited and the results were unlikely to be clinically significant. This analysis presented in Table 1.2 showed that ICT intervention was associated with a decrease in HbA1c levels in patients with type 1 diabetes (SMD, −0.05; 95% CI, −0.08 to −0.02; moderate), type 2 diabetes (SMD, −0.06; 95% CI, −0.07 to −0.05; moderate), and gestational diabetes (SMD, −0.07; 95% CI, −0.10 to −0.03; low). To improve diabetes-related clinical outcomes and prognosis, there is an urgent need to develop applications that accurately reflect clinical evidence.

Table 1.2

CI, confidence interval; GDM, gestational diabetes mellitus; ICT, information and communications technology; SMD, standard mean difference; T1DM, type 2 diabetes mellitus; T2DM, type 2 diabetes mellitus.

From Park S, Lee H, Cho W, et al. Efficacy of information and communication technology interventions for the management of diabetes mellitus: An umbrella review and evidence map. Obes Rev. 2024;e13714. https://doi.org/10.1111/obr.13714. Article in Review.

Wearable devices are not without drawbacks. There may be issues with accuracy or reliability, and patients unfamiliar with the devices may have difficulty using them correctly. Other considerations include device cost and battery life. While widely used for individual health improvement, these devices are not yet widely used as components of comprehensive clinical interventions in conjunction with actual hospital settings. The efficacy of wearable devices in improving clinical outcomes is also not well established. A meta-analysis based on data from randomized controlled trials of wearable trackers found that the clinical utility of these devices was confirmed only for certain demographic groups, such as those over 50 years of age, males, and whites. ⁵ Further evidence is therefore needed.

Solutions

Digital health solutions offer the ability to deliver information tailored to individual user needs, increase the efficiency of self-management through personalized interventions, improve clinical outcomes, and enable early detection or effective treatment of complications. In addition, user data generated through the use of digital health solutions can be analyzed and used to generate evidence that informs and improves healthcare delivery. The following discussion highlights some key elements of digital health relevant to diabetes and metabolic disease.

Mobile applications

In recent years, mobile applications with a wide range of diabetes management features have been introduced, from blood glucose monitoring apps to nutrition and exercise trackers. Such applications can empower people with diabetes to improve their self-management skills.

In the area of blood glucose monitoring, mobile applications play a key role. People with diabetes can use these applications to enter and track their blood glucose data, enabling them to respond quickly to fluctuations in blood glucose levels. This facilitation of glucose management helps healthcare providers develop accurate treatment plans. ⁶ In addition, mobile apps contribute significantly to diet and exercise management. Users can record their diet, track physical activity, and monitor calorie and nutrient intake accordingly. This process helps to cultivate healthier eating habits and encourage physical activity, thereby supporting optimal glycemic control. ⁶ , ⁷ Mobile applications can also help improve medication adherence people with for people with diabetes. ⁶ Individuals can track their medication intake and schedule doses, improving medication management and reducing potential side effects.

Nonetheless, the efficacy of mobile applications in enhancing diabetes clinical outcomes is yet to be substantiated robustly. A report by the Agency for Healthcare Research and Quality (AHRQ) revealed that only a minimal fraction of applications demonstrated significant improvements in glycated hemoglobin (HbA1c), a crucial metric for diabetes clinical outcomes. ⁴ None of the studies indicated long-term effects on patient quality of life, blood pressure, weight, or body mass index. Many of the intervention studies exhibited methodological issues, including short study durations, inconsistent reporting of randomization, masking, and dropout analysis, and frequent employment of cointerventions, which complicated result interpretation. None of the included studies were deemed high quality. Differences in user perceptions and experiences with application interfaces, as well as differences in individuals' receptivity to the use of information and communication technologies, may also contribute to this lack of scientific evidence.

Wearable devices

Wearable devices, including smartwatches and activity trackers, are designed to be worn on the body to track lifestyle data such as physical activity, heart rate, and sleep patterns. Most commonly presented as wristwatches or bands, these devices also come in several other forms, including necklaces, rings, and belts. In recent years, wearable devices have gained popularity as valuable health management tools for people with diabetes and other chronic conditions.

By providing real-time feedback on physical activity and blood glucose levels, wearables can motivate patients to maintain a healthier lifestyle. ⁸ These devices can also be integrated with advanced diabetes management technologies, including smartphones and continuous glucose monitors, to provide a more comprehensive view of an individual's health. In addition, wearable devices are highly beneficial for data analytics, as they can collect data on blood glucose levels, exercise habits, dietary patterns, and more. This data can then be analyzed to provide clinically meaningful insights. This real-world data has catalyzed innovative research designs aimed at improving health outcomes for people with diabetes.

One example is a study involving participants in the All of Us Research Program. ⁹ Using activity tracker and electronic health record data from participants in the All of Us program, researchers established thresholds for the level of activity required to reduce the risk of several chronic diseases. They identified activity levels that reduced the risk of diabetes (8160 steps), essential hypertension (8290 steps), GERD (8260 steps), major depressive disorder (8210 steps), obesity (8280 steps), and sleep apnea (8220 steps). ⁹ They also found that activity levels above these thresholds provided no additional benefit for disease prevention.

While widely used for individual health improvement, these devices are not yet extensively deployed as components of comprehensive clinical interventions in conjunction with actual hospital settings. The efficacy of wearable devices in improving clinical outcomes also remains underevidenced. A meta-analysis based on data from randomized controlled trials of wearable trackers found that the clinical utility of these devices was only confirmed for certain demographics, such as those over 50, male, and white. ⁵ Thus, further evidence accumulation is required. With the proliferation of wearable devices, medical waste is an emerging issue. There is a growing interest and effort in more eco-friendly wearable devices to help in the effective management and treatment of chronic diseases, including diabetes. ¹⁰

Advanced glucose monitoring

Continuous Glucose Monitoring (CGM) is an advanced technology that allows people with diabetes to continuously monitor their blood glucose levels. Unlike traditional self-monitoring glucose meters, this technology measures glucose levels in the subcutaneous interstitial fluid. ¹¹ While this data is somewhat different from blood glucose concentration, it can play a critical role in improving the health status and treatment plan of individuals with diabetes.

CGM allows people with diabetes to monitor their blood glucose levels in real time. The sensor, which is inserted into the subcutaneous tissue, continuously transmits glucose levels to a wirelessly connected receiver or smartphone. This real-time visibility allows patients to see their current blood glucose levels. The system can also analyze glucose level trends and historical data, alerting the user to potential rapid fluctuations in glucose levels. This preemptive notification facilitates proactive responses to glucose changes, particularly in the prevention of hypoglycemia in people with diabetes. ⁷ Current clinical guidelines recommend the routine use of real-time CGM to improve clinical outcomes in all adults with type 1 diabetes, adults with type 2 diabetes on multiple daily insulin injections, and pregnant women with type 1 diabetes. ⁷

Standardized summaries, such as the Ambulatory Glucose Profile (AGP), have been proposed to enable the effective use of CGM in clinical practice for patients with diabetes. ¹² These summaries are considered part of formal treatment goals for people with diabetes in clinical practice guidelines. ¹³ Generally recommended goals for most nonpregnant adults with diabetes include Time in Range (time with blood glucose levels between 70 and 180 mg/dL) greater than 70%, Time Above Range (time with blood glucose levels above 180 mg/dL) less than 25%, and Time Below Range (time with blood glucose levels below 70 mg/dL) less than 4%. ¹³ More recently, the glycemia risk index (GRI), which provides a single-number summary of glycemic quality, has been developed and is being explored for clinical application. ¹⁴ While GRI is not yet widely used in routine diabetes care, it is believed that it can provide a more detailed picture of an individual's clinical situation in conjunction with existing clinical indicators, so it is expected that its use will increase in the future.

In addition, CGM can help promote healthy lifestyle behaviors such as diet and exercise. By monitoring blood glucose levels in real time, patients can correlate the nutrient content and quantity of their meals with their blood glucose levels, aiding in meal planning and glycemic control. CGM systems can also be used to monitor blood glucose levels before and after physical activity, helping to plan exercise more effectively. Recently, the clinical utility of CGMs in nondiabetic individuals has also been explored. ¹⁵ For example, the use of CGM is being explored to improve blood glucose patterns for diabetes prevention, improve mental or physical functioning, and promote motivation for healthy behavior change. More recently, there have been attempts to measure metrics other than glucose, such as ketones or lactate, using methods similar to CGM. If these attempts gain meaningful scientific support, it is expected that the development of wearable devices for monitoring continuous biomarkers will accelerate.

Telemedicine

Telemedicine is the delivery of health services through digital communication technologies. It includes various applications that use smartphones, two-way video, and other forms of ICT to deliver health services. Depending on the mode of delivery, these services can be provided to patients synchronously or asynchronously. Synchronous telemedicine is when a healthcare provider communicates with a patient in real time to provide a medical service, while asynchronous telemedicine is when a service, such as reading a patient's test results, is provided asynchronously at a later time. The term telehealth is also often used to refer to broader digital health services, such as remote patient monitoring, in addition to telemedicine. ¹⁶

In Korea, the implementation of telehealth services within the institutionalized healthcare system has been long delayed due to various different stakeholder positions. However, since the COVID-19 pandemic, telehealth services have become officially accessible and are now widely used. According to the Ministry of Health and Welfare of the Republic of Korea, from February 2020, when telemedicine services were allowed, to January 2023, 36.61 million telemedicine services were provided by 25,697 medical institutions to a total of 13.79 million individuals. ¹⁷ Although the program was implemented under the umbrella of COVID-19, most of the actual care was provided to individuals with chronic diseases, including diabetes. However, telemedicine cannot completely replace face-to-face medical visits, especially for first-time patients, and there is a problem that telemedicine fees for primary medical centers in Korea are about 30% of face-to-face fees, which does not compensate doctors