Digital Health Technology in Diabetes Management in the Asia–Pacific Region: A Narrative Review of the Current Scenario and Future Outlook

Abstract

Diabetes is a growing global health concern with a high prevalence in the Asian and Western Pacific regions. Effective diabetes management mainly relies on self-care practices. However, glycemic control remains poor, especially in developing nations where healthcare access is limited. Low physician density and minimal healthcare funding exacerbate the challenges faced by people with diabetes in Asia. Digital health technologies offer promising solutions to bridge these gaps. These technologies enhance patient engagement, improve medication adherence, and promote healthier lifestyles. Mobile apps provide tools for self-management, such as monitoring physical activity and dietary intake, while telemedicine platforms and electronic medical records facilitate patient data management and remote consultations. Despite the advantages provided by digital health technologies in managing diabetes, barriers to their adoption include infrastructure limitations, regulatory challenges, and issues with data security. Some Asian countries have made major strides in the adoption of digital health tools with national strategies and regulatory bodies to manage digital health options; however, disparities in digital health readiness persist. Effective implementation of these technologies requires addressing these barriers, including enhancing infrastructure, improving app usability, and ensuring regulatory compliance. While digital health solutions present significant opportunities, their impact depends on overcoming current challenges and ensuring equitable access and effective use in managing diabetes. Future directions should focus on prioritizing app acceptance and efficacy, as well as integrating machine learning and artificial intelligence-powered digital solutions.

Key Summary Points

Diabetes affects a substantial global population, with severe economic impacts and poor glycemic control in developing regions.

This study aimed to evaluate the effectiveness of and barriers to implementing digital health solutions for diabetes management, particularly in low- and middle-income countries (LMICs) in the Asia–Pacific region.

Digital health solutions, including mobile apps and telemedicine, have shown promise in improving diabetes management outcomes by enhancing patient engagement and self-care.

Despite their potential, barriers, including infrastructure limitations and low health literacy, hinder the widespread adoption of these technologies. Future research and policy development are vital for overcoming these challenges, addressing safety and regulatory concerns, and optimizing digital health solutions for diabetes care.

Introduction

In 2021, diabetes affected 537 million people globally, with the top five countries located in Asia, while the Western Pacific region alone accounted for 38% of global cases. The numbers are expected to escalate to 152 million by 2045 in the Southeast Asian region alone [1]. Effective management of diabetes hinges on self-care practices by people with diabetes (PwD) [2]. Research indicates that glycemic control in PwD has been consistently poor in developing nations [3]. In addition, limited access to healthcare resources exacerbates the problem. In Asia, there are only approximately 1.4 physicians per 1000 people, underscoring the region’s significant healthcare workforce challenges [4]. The proportion of gross domestic product (GDP) allocated to healthcare in several Asian countries is also relatively low, averaging 5.5%, which contributes to inadequate healthcare infrastructure and resources [5]. High costs associated with diabetes care (including medications and physicians’ fees), geographical barriers, low levels of literacy (particularly health literacy), and limited access to knowledge on diabetes self-management, further exacerbate the issue [3, 5–9].

On a positive note, digital health technology is emerging as a crucial tool in bridging these gaps, transforming the landscape of diabetes care. Patient engagement can be enhanced by integrating digital solutions into routine care. Improved medication adherence and healthier lifestyles for PwD can be promoted through these tools [10]. Mobile applications (apps) are effective in reinforcing education and skills needed for diabetes self-management [11]. Additionally, electronic medical records (EMRs), telemedicine, and decision support systems are emerging as viable options for standardized interventions in primary care. These technologies enable data collection, expand access to resources, and provide flexible care options [2, 10]. With recent advancements in healthcare delivery technologies using machine learning (ML) and artificial intelligence (AI), there is substantial potential to promote greater patient engagement and optimize diabetes care [12, 13]. This paper reviews current digital health solutions, evaluates their benefits for diabetes management, and examines the barriers to their implementation in the Asia–Pacific region, particularly in low- and middle-income countries (LMICs). It also discusses strategies to enhance the usefulness of apps and explores future perspectives.

Methods

A comprehensive search was conducted across multiple databases, including PubMed, Google Scholar, and Scopus, using terms such as “diabetes”, “digital app”, “Mobile”, “Smartphone”, “mHealth” “CGM” (“diabetes”[All Fields] OR “diabetes mellitus”[MeSH Terms] OR (“diabetes”[All Fields] AND “mellitus”[All Fields]) OR “diabetes mellitus”[All Fields] OR “diabetes”[All Fields] OR “diabetes insipidus”[MeSH Terms] OR (“diabetes”[All Fields] AND “insipidus”[All Fields]) OR “diabetes insipidus”[All Fields] OR “diabetic”[All Fields] OR “diabetics”[All Fields] OR “diabetes”[All Fields]) AND (“mobile applications”[MeSH Terms] OR (“mobile”[All Fields] AND “applications”[All Fields]) OR “mobile applications”[All Fields] OR (“smartphone”[All Fields] AND “app”[All Fields]) OR “smartphone app”[All Fields])) AND (clinicaltrial[Filter]). The search was limited to articles published between June 2008 and June 2024. A grey literature search for relevant articles was also conducted using the Google search engine and Google Scholar.

Ethical approval was not applicable or required as this article is based on previously conducted studies and does not contain any new data from experiments on human participants or animals performed by any of the authors.

Results

A total of 78 articles were selected, with most studies demonstrating positive outcomes for digital health solutions in diabetes management, particularly in improving glycemic control and patient engagement. The review incorporated various types of studies, including observational studies, randomized controlled trials (RCTs), systematic reviews, and narrative reviews. Variables such as the type of digital health solution, effectiveness, implementation challenges, and region-specific data were incorporated. The desired features of digital apps were captured through a detailed review of published literature on app functionalities and user feedback. Limitations included variability in study designs, population diversity, and differences in the length of follow-up periods.

Discussion

Navigating Diabetes Care Challenges with Digital Solutions

Mobile health (mHealth) and electronic health (eHealth) interventions are primarily delivered via smartphones, wireless screening devices, or telemedicine platforms. Notably, phone-based apps have a larger patient reach, allowing for broader and more effective dissemination of health information and support [14]. The appeal of text message-based programs is evident, with approximately 8.6 million signing up for India’s mDiabetes program in 2018 [15]. The following section explores how digital health solutions can address key challenges in diabetes management.

Challenge: Limited Access to Healthcare and Geographical Barriers

Solution 1: Telemedicine and remote consultations: Mobile apps and telemedicine platforms can bridge the gap in regions with a low density of healthcare providers by allowing remote consultations [4, 16]. Virtual consultations reduce the need for frequent in-person visits, which lowers healthcare costs. Additionally, it helps overcome geographical barriers by providing support remotely [17–19].

Solution 2: EMR integration of diabetes data: EMRs allow physicians to track comprehensive patient data, while patient portals enable patients access to their test results and care plans in real-time, providing instantaneous feedback [20]. EMRs can facilitate the implementation of standardized clinical guidelines for diabetes management in primary care settings [21]. EMRs integrated with mobile apps can aid remote monitoring alongside personalized care, enhancing the capacity to deliver high-quality care [22].

Challenge: Inadequate Healthcare Funding and Infrastructure

Solution: Efficient resource utilization: Digital tools can enable preventive care and early intervention by providing real-time health monitoring, personalized health insights, and timely alerts for potential issues. They enable individuals to track their health status continuously, reducing the need for frequent in-person consultations, which can be expensive and logistically challenging in low-resource settings. Timely alerts for potential health issues enable early intervention, helping prevent minor conditions from developing into severe conditions that demand extensive treatment. This can help mitigate the strain on healthcare systems functioning on limited budgets by lowering the overall demand for intensive care services [5].

Challenge: Low Levels of Health Literacy

Solution 1: Educational content: Ensuring the availability of apps in local languages is crucial for increasing their accessibility and effectiveness in diverse populations [23]. Digital health applications incorporating culturally tailored educational resources can help users understand and self-manage their condition effectively [6, 24, 25].

Solution 2: Behavioral support: Apps provide behavior change strategies and actionable insights, making it easier for individuals with low health literacy to adopt and maintain healthier lifestyles. They also facilitate ongoing engagement and support for PwD [25–32].

Challenge: Suboptimal Self-care Practices

Solution 1: Glycemic monitoring and feedback: Digital tools monitor glucose levels and provide real-time feedback to PwD to help improve glycemic control. They also aid in assessing physical activity levels, nutrition, and weight control [32]. Research demonstrates that mobile apps and mHealth interventions significantly reduce glycated hemoglobin (HbA1c) levels, with excellent results in patients with a long-standing history of diabetes [18, 19, 22].

Solution 2: Engagement and adherence: Mobile apps support self-management by providing reminders for medication, lifestyle modification tips, and diabetes education, which may be especially beneficial to elderly PwD struggling with compliance [18, 33, 34].

Challenge: Constraints in Emergency Management

Solution 1: Emergency alerts and information: Apps are designed to provide critical information and customizable alerts for managing hypoglycemic episodes and other emergencies, ensuring timely interventions [25–30, 35].

Solution 2: Clinical decision-support systems: Embedded decision-support features in mHealth apps promote PwD–clinician interactions and aid in clinical decision-making, improving treatment outcomes [36].

By leveraging digital health technologies, the challenges associated with diabetes care can be effectively addressed, leading to improved outcomes and enhanced quality of life for PwD [36]. To successfully harness the potential of digital health solutions, it is essential to understand the key elements that contribute to the effectiveness of diabetes management apps.

What Makes for an Effective Diabetes Management App?

The design and content of health apps should involve an interdisciplinary team, including local and international health professionals and organizations, to ensure comprehensive and culturally relevant solutions [33]. Standardizing patient-centered outcomes using mHealth evidence reporting and assessment (mERA) guidelines is crucial to ensure consistent and reliable reporting [19].

The following features (Fig. 1) can be considered for an effective diabetes app [23, 29, 34, 35, 37–41]. The parameters for selecting app features were derived from a comprehensive analysis of multiple data sources. These included references based on observations and surveys from a randomized trial of diabetes apps, systematic reviews addressing mobile health barriers, feedback from nearly 400 users of a decision support application, assessments from cluster-randomized trials with healthcare providers, and a systematic search of qualitative and mixed-method studies on technology-assisted diabetes self-management education (DSME). This diverse approach ensured that the chosen features effectively addressed psychological needs, usability concerns, and practical considerations in diabetes management.

Fig. 1.

Features required for an effective diabetes app [23, 29, 34, 35, 37–41]. GDMT guideline-directed medical therapy, HCP healthcare professional

An Overview of Diabetes Apps

Diabetes apps offer a range of digital solutions designed to assist with the management and monitoring of diabetes. Table 1 presents details of apps based on their evidence of effectiveness and varied functionalities, addressing different aspects of diabetes management [13, 22, 42–59].

Table 1.

Overview of digital health apps for diabetes management: features, study designs, and outcomes

App name	App features	Monitoring/education/dose titration	Study type	Population (T2DM/ T1DM)	Study sample size (n)	Age (years)	Baseline parameters assessed	Study duration	Study outcomes	Reference	User satisfaction
Smart Glucose Manager	The Android-based app offers reminders for glucose checks, medication, meals, and exercise. It also offers real-time glucose data storage with export options and bolus insulin dose calculations based on carbohydrate ratio and target glucose levels	Monitoring/dosing	RCT	T2DM	67	52 ± 11.7	Age, gender, BMI, diabetes duration, HbA1c	6 months	From 0 to 3 months, significantly lower HbA1c (− 1.36%) in cases vs. (− 1.13%) in controls From 3 to 6 months, HbA1c further improved (− 0.96%) in cases but no significant improvement in controls HbA1c improvement was positively correlated with the usage of the app	[42]	Yes, good user engagement
DIABEO® system	A telemedicine solution integrating a patient mobile app with a web portal for healthcare providers, enabling real-time monitoring and therapeutic decisions for insulin therapy	Monitoring/dosing	Randomized, parallel-group, open-label trial	T1DM = 610 T2DM = 55	665	38.5 ± 13.8	Age, gender, diabetes duration, HbA1c, usage of app, hypoglycemia, quality of life	12 months	Statistically significant reduction in HbA1c (− 0.51%) vs. control	[43]	Beneficial; familiarizing PwD with the app by nurses led to better usage rates
Diabetes Interactive Diary (DID)	It includes a carbohydrate and insulin bolus calculator and promotes patient–provider communication via a short message system	Monitoring/education/dose titration	Open-label, international, multicenter, randomized, parallel-group study	T1DM	127	36.9 ± 10.5	Age, gender, blood pressure, HbA1c, diabetes duration, cholesterol profile, BMI, fear of diabetes complications, quality of life	6 months	Similar efficacy in reducing HbA1c levels (decreased by  0.49 in the intervention group and 0.48 in the control group, which involved traditional carbohydrate counting)	[44]	Yes (due to improvements in QoL)
OneTouch Reveal®	Contains color-coded tools, personalized reminders, and graphics for glucose metrics	Monitoring/education	RCT	T1DM = 79 T2DM = 49	128	44.6 (19–71)	Age, gender, diabetes duration, HbA1c, type of diabetes, self-monitoring of blood glucose frequency, and diabetes treatment the patient was receiving at baseline	12 and 24 weeks	The decrease in HbA1c was greater in participants using the meter + app after 12 weeks (− 1.04%) than in participants using the meter alone (− 0.58%) The use of the OneTouch™ Verio Flex glucose meter alone or in combination with the OneTouch Reveal® diabetes app was associated with significant improvements in glycemic control after 12 and 24 weeks	[45]	Yes, in terms of usage and control of blood sugar
Welltang	Provides real-time, contextually relevant coaching and education, utilizing integrated data to target behavioral coaching	Monitoring/education	RCT	T1DM = 18 and T2DM = 82	100	53.5 ± 12.4 (control); 55.0 ± 13.1 (intervention)	HbA1c, blood glucose, hypoglycemic events, blood pressure, low-density lipoprotein cholesterol, weight, the satisfaction of patients to use Welltang, self-care behaviors, and diabetes knowledge of patients	3 months	The average decrease in HbA1c was − 1.95% in the intervention group vs. − 0.79% in the control group	[46]	Yes
Diabetes Manager + iBGStar®	Installing this app helps the user’s iPhone serve as a glucometer	Monitoring	Single-center, prospective, randomized, open-label study	T1DM	100	38 ± 11 (iBGStar group); 39 ± 12 (control group)	Age, gender, BMI, diabetes duration, HbA1c, insulin dose, hypoglycemia fear scale, behavior scale, worry scale	6 months	Improvement in insulin dose; significant decrease in HbA1c) (− 0.51%) in the iBGStar group vs. control (− 0.16%)	[47]	Yes
NexJ® systems Health Coach	It includes a mobile phone-based health coaching protocol	Education	Non-inferiority, pragmatic RCT	T2DM	97	53.2 ± 11.3	HbA1c from baseline to 6 months, waist circumference, weight, BMI, satisfaction with life, Positive and Negative Affect Schedule, quality of life (Short Form Health Survey-12), and depression and anxiety (Hospital Anxiety and Depression Scale)	6 months	Both groups reduced their HbA1c levels; however, there were no significant differences between the groups in the changes in HbA1c levels at 6 months Reduction in HbA1c at 6 months was (− 0.84%) in the intervention group vs. (− 0.81%) in the control group, along with significant decreases in weight and waist circumference	[48]	Yes, in terms of QoL and weight reduction
One Drop® + activity tracker	Includes a variety of features for diabetes management; stores and monitors blood glucose readings, medication doses, physical activity, and dietary intake	Monitoring	RCT	T1DM	95	40.9 ± 10.7	Age, gender, diabetes duration, BMI, digital usability, HbA1c, and hypoglycemic events	3 months	Significantly lower HbA1c at 3 months in the intention-to-treat analysis group (− 0.5%) from baseline	[49]	Yes
Glucose Buddy™	A free iPhone app that allows users to input blood glucose levels, insulin doses, medications, food intake, and physical activity. It features a customizable graph for data visualization and enables data export via email	Monitoring	RCT	T1DM	72	35.20 ± 10.43	Age, gender, HbA1c, insulin pump use, and diabetes duration	9 months	Reduction in HbA1c was − 1.28% in the intervention group vs. + 0.11% in the control group	[50]	Not mentioned
Diabetes Pal®	It includes a hypoglycemia guide	Dose titration	Randomized, open-label, parallel-group trial	Insulin-naïve patients with T2DM	66	53.3 ± 7.4	Age, gender, diabetes duration, HbA1c, BMI	24 weeks	Reduction in HbA1c (− 0.39%) from baseline in the intervention group compared to the control group, with no significant difference between groups	[51]	Yes
BD™diabetes care	Contains features enabling data logging, reminders, and educational content	Education	Multicenter, open-label, parallel-group RCT	T2DM	58	58.3	Age, gender, race, ethnicity, duration of diabetes, duration of insulin therapy, mean basal insulin dose, information on pen needle length, and the current use of the mobile app	8 weeks	Reduced glycemic variability in the intervention group (p = 0.06) vs. the control group	[52]	Yes
VoiceDiab	It includes three remote servers for automatic speech recognition, text analysis, and insulin dose calculation	Monitoring	RCT	T1DM and intensive insulin therapy	44	16.8 ± 10.2	Age, gender, BMI, duration of diabetes, HbA1c	1 year	With VoiceDiab support, postprandial blood glucose was in the target range (70–180 mg/dL) 58.6% of the time vs. 46.6% in controls	[53]	Not mentioned
WellDoc®	Provides real-time feedback on blood glucose levels, displays medication schedules, and includes treatment plans for hypo- and hyperglycemia	Monitoring/education	Non-blinded RCT	T2DM	30	51.04 ± 11.03	Age, gender, diabetes duration, HbA1c, BMI, diabetes comorbidities, medication treatment regimen, physician specialty	3 months	Decrease in HbA1c (− 2.03%) vs. control (− 0.68%). Improved self-care behavior and knowledge of food choices. Self-reported satisfaction for confidence in control of diabetes and provider care	[54]	Yes
D-Partner	Includes a bolus wizard with a food database, a basal insulin titration manager, and automated notifications	Monitoring	Single-center RCT	T1DM	24	37.4 ± 11	HbA1c	3 months	The intervention group showed a significant reduction in HbA1c at 3 months from baseline compared to the control group (p = 0.01) Values were not mentioned	[13]	Not mentioned
Tidepool®	It is a unified software platform that consolidates data from various devices	Monitoring/education	Pilot study	T1DM	15	20–60	Gender, age, years in practice, provider type, and average number of patients seen per month	6 months	The number of times the HCP reviewed the patient data increased from a mean of 2.8 to 6.1 HbA1c decreased by − 0.2% from baseline at 6 months	[55]	Yes, easy to use
Wellthy CARE™	It enables individuals to record their diabetes self-care data; individuals receive real-time educational, behavioral, and motivational feedback using an AI-powered chatbot The program uses a digital persuasion model to boost patient motivation, simplify tasks, and deliver action triggers through culturally relevant content like lessons, videos, pro-tips, quizzes, and storyboards	Monitoring/education	Observational study	T2DM	102	50.8 (49.2–52.4)	Gender, age, weight, BMI, baseline HbA1c	16 weeks	PwD showed a significant decrease in HbA1c, fasting blood glucose levels, and postprandial blood glucose levels after 16 weeks HbA1c, FBG, and PPBG levels decreased progressively with higher program engagement Mean change in HbA1c from baseline at the end of 16 weeks was − 0.49%	[56]	Beneficial (positive app engagement demonstrated by improved glycemic control and a good retention rate)
My Star Dose Coach™	Includes an insulin dose calculator that provides automated dosing suggestions	Dose titration	Multicenter RCT, treat-to-target study	T2DM	151	62.1 ± 9.5 years	Age, sex, race (Caucasian), body weight, BMI, eGFR, previous insulin use	16 weeks	The percentage of participants reaching the FSMPG target range without confirmed (≤ 70 mg/dL [≤ 3.9 mmol/L]) or severe hypoglycemia was 34.3% vs. 14.5%, respectively. The time at which 50% of the participants achieved the FSMPG target was less in the device-supported than the routine titration arm. Change in HbA1c from baseline to week 16 was − 1.12%	[57]	Yes (ease of use of the device was ≥ 6 on a scale of 1–7)
mySugr® Bundle App	SMBG and continuous glucose monitoring data can be automatically uploaded via Bluetooth or Apple Health and synced through Cloud and certified diabetes educator-led coaching	Monitoring/education	Retrospective observational study	T1DM = 29, T2DM = 19 (3 had latent autoimmune diabetes in adults [LADA], and 3 had an unreported diabetes type)	52	30.8 ± 15.3	Age, gender, HbA1c	6 months	Sustained improvements in HbA1c (− 1.3%) from baseline	[58]	Yes, high user satisfaction
iCareD system	Self-monitoring of blood glucose with automatic transfer of data for glucose levels, diet, and physical activity counseling	Monitoring/education	RCT	T2DM	269	52.5 ± 12.3	Age, gender, HbA1c, duration of diabetes, BMI	52 weeks (ongoing)	The significant mean decline in HbA1c levels for the mobile diabetes self-care group was − 0.86% and that of the group with personalized, bidirectional feedback from physicians was − 1.04%	[22]	Yes, high user satisfaction
TangPlan + WeChat	A combination of web-based software for T2DM management and a mobile app for providing guidance on self-management	Monitoring/education	RCT	T2DM	120			6 months	The mean change in HbA1c in the TangPlan + WeChat group was − 1.6% and − 0.11% in the control group	[59]

BMI basal metabolic index, FSMPG fasting self-monitored plasma glucose, HbA1c glycated hemoglobin, PwD people with diabetes, QoL quality of life, RCT randomized controlled trial, T1DM type 1 diabetes mellitus, T2DM type 2 diabetes mellitus

Selection Criteria for Apps

The selection of apps was based on the following key criteria:

• Type of study: Use in high-quality studies, such as randomized controlled trials (RCTs). Some single-arm studies were included if outcome data was presented. Studies with data on real-world outcomes were also included.

• Outcome data: Demonstration of outcome data related to HbA1c improvement, user satisfaction, healthcare practitioner (HCP) reporting, or improvements in quality of life.

• Features and focus: Assessment of various functionalities, including monitoring, education, dose titration, and engagement.

• Approval status: Apps with strong RCT data were favored, regardless of formal regulatory approval such as from the US Food and Drug Administration (FDA) and European Union (EU).

Table 1 gives an overview of various diabetes apps that are currently available [13, 22, 42–59].

Assessment Summary of Digital Apps

Digital health apps address multiple facets of diabetes self-management. Users of OneTouch Reveal® experienced a 1.04% reduction in HbA1c within 12 weeks, proving its efficacy in glucose monitoring [43, 45, 50]. Apps that focus on delivering educational content, including real-time health coaching and goal-setting features, as seen in NexJ® Systems Health Coach and Wellthy CARE™, exhibited improved glycemic control and quality of life [48, 56]. Apps providing tools for insulin dose adjustment showed a reduction in HbA1c by as much as 1.12% [51, 57]. Overall, the use of apps resulted in a reduction in HbA1c levels, ranging from 0.39% to 1.95%. Apps like mySugr®, Wellthy CARE™, My Star Dose Coach™, and Welltang demonstrated sustained user engagement [46, 56, 60, 61]. Moreover, Welltang (China) and Wellthy CARE™ (India) provide culturally relevant features and educational content that may further enhance user engagement. Glucose Buddy™, an app with customizable graphs and data export features, may also appeal to users seeking adaptable diabetes management tools. While this assessment summarizes the efficacy of these apps, factors such as group heterogeneity, differing study designs, and differences in app applicability should be considered when interpreting these findings.

Barriers to the Utilization of Digital Health Solutions

The utilization of mHealth in developing countries of Asia faces significant barriers, including low faith in the credibility of eHealth services [62], age-associated barriers, and costs [62, 63].

Infrastructure limitations, inadequate equipment availability, and technology gaps are also significant barriers to the uptake of digital health solutions [10, 33, 37]. Lack of services such as stable internet connectivity hinders the effective implementation of digital health technologies in developing countries. Many privately developed apps lack essential self-management functionalities and provide limited interactions between disciplines [64]. Most apps fall short of providing comprehensive educational resources, either due to nonadherence to clinical guidelines or due to the lack of a patient-centered outlook [25, 33]. Apps offering clinical decision support are still in the nascent stages of development and need improved integration and device compatibility [35]. The use of incompatible healthcare systems makes it difficult to share patient records efficiently, leading to errors and inefficiencies in care [65].

Key Strategies to Boost Utilization of Digital Solutions

Enhancing access, including funding support (low-priced mobile phones, apps, or internet subscriptions) [37], primary HCP training for effective use, and incentivization for sustainable tool use can increase the adoption of digital health tools [10, 66]. Individualization of mHealth apps, taking into account cultural needs [24], age, and cost [25, 67], is an important factor in ensuring their success. Community-based health education involving motivational and problem-solving sessions can facilitate both young and older adults in self-management [25, 68, 69].

Safety and Regulatory Issues

Regulatory frameworks often struggle to keep pace with the diverse tools and rapid advancements in digital health technologies [70, 71]. Inadequate cybersecurity and data privacy regulations pose serious threats, especially in developing nations [70]. Successful digital health promotion includes addressing unfair data practices and improving responsible technology governance, especially in LMICs [72]. The Global Digital Health Partnership (GDHP) was established internationally in 2018 as a collaborative effort among 33 countries and territories, the World Health Organization (WHO), the Organization for Economic Co-operation and Development (OECD), and the International Digital Health & AI Research Collaborative (IDAIR). The partnership seeks to improve healthcare safety, quality, and effectiveness through digital health services [73].

In addition to this, the Centre of Regulatory Excellence (CoRE), Singapore 2020 focuses on efforts to address healthcare data policies, data security, and digital health regulation in ASEAN (Association of Southeast Asian Nations) and Asia–Pacific regions. It prioritizes establishing a secure framework for managing healthcare data, including standardized data-sharing protocols and improved transparency. It also seeks to keep pace with the rapidly evolving digital health landscape, while promoting international cooperation to align standards and streamline regulatory processes. Among Southeast Asian countries, Malaysia stands out with a well-administered legal framework for data security. This framework comprehensively addresses data storage, transmission, and utilization [74]. The Infocomm Media Development Authority of Singapore and the Personal Data Protection Commission are involved in data protection and governance in Singapore [74].

A software solution meeting the “medical device” definition is classified as Software as a Medical Device (SaMD) and must be further categorized. This classification is crucial as it determines the regulatory requirements before and after the product enters the market [75]. In most ASEAN countries, there is no specific guidance for SaMD, which is regulated using general medical device regulatory frameworks. However, regulatory authorities in Australia, Japan, and Singapore have been proactive in developing and advancing SaMD-specific regulatory frameworks. These agencies have implemented international best practices, providing opportunities for pre-submission consultations with innovators and developing regulatory review approaches tailored to the unique needs of digital health products [74].

As an example, the SaMD classification approach by the Singapore Health Sciences Authority (HSA) is illustrated below [75]:

1. Low risk: Software/app that does not measure patient parameters but only displays parameters derived from another monitor (e.g., patients’ monitors)

2. Moderate risk: Apps that measure heart rate, echocardiogram, and episodes of hypoglycemia

3. High risk: Apps used for direct continuous monitoring

The Ministry of Health, Singapore and HSA regulate the approval of health apps as medical devices before release. This information is publicly available on the HSA and HealthHub websites [76].

Additionally, the International Medical Device Regulators Forum (IMDRF) AI Working Group is involved in developing guidance documents on standardized AI terminology and definitions for AI medical devices. It includes WHO representatives and regulatory agencies from China, Japan, Singapore, and South Korea, among other nations [74].

Implementing secure data storage solutions like private clouds (as seen in Singapore), educating users and healthcare providers about data security practices, and protecting personal health information are needed for the successful adoption of digital health technologies [10, 77]. Cybersecurity practices could include applying endpoint management tools and perimeter defenses like antivirus and firewalls, restricting technology use to comply with regulations such as the Health Insurance Portability and Accountability Act (HIPAA), and securing remote environments with multifactor authentication. Raising security awareness through comprehensive training and updating policies to support new technologies is also crucial, alongside implementing incident reporting frameworks like the Care Computer Emergency Response Team (CareCERT) alerts for effective cyber threat management [78].

Future Directions for Diabetes Digital Health

Further research is needed on parameters such as acceptance of app features and perception of its efficacy and design, especially in low-resource settings [79]. Clinical decision-making functions require refinement and evaluation for effectiveness [80]. ML-enabled web applications for diabetes management could be adopted in remote areas as part of community health initiatives [81]. Automatic retinal screening and emergency alerts powered by AI algorithms enhance precision, prediction, and personalization, thereby increasing the impact of digital health [82].

Conclusion

Digital health technologies offer promising avenues for enhancing diabetes management through improved patient engagement, personalized care, and streamlined self-management tools. These technologies, ranging from smartphone apps to wearable devices, empower individuals with diabetes to monitor their health more effectively and make informed decisions about their lifestyle and treatment. AI is set to revolutionize diabetes care by enabling more precise, accessible, and personalized treatments. Despite significant advancements, challenges such as poor technological literacy, regulatory hurdles, and varying levels of acceptance across different demographics and regions persist. By focusing on collaborative development, comprehensive research, and inclusive implementation strategies, the full potential of digital tools to enhance healthcare delivery and patient outcomes could be harnessed.

Author Contributions

All authors (Daphne S.L. Gardner, Banshi Saboo, Jothydev Kesavadev, Norlaila Mustafa, Michael Villa, Edward Mahoney, and Shailendra Bajpai) contributed to the study conception and design. Definition of intellectual content and literature search were performed by Daphne S.L. Gardner, Banshi Saboo, and Jothydev Kesavadev. The first draft of the manuscript was written by Daphne S.L. Gardner and all authors further contributed and commented on previous versions of the manuscript. All authors read and approved the final manuscript.

Funding

Funding for the journal’s Rapid Service Fee, as well as for medical writing and editorial assistance, was provided by Becton Dickinson/Embecta (formerly BD Diabetes Care), Singapore.

Data Availability

Data sharing is not applicable to this article as no datasets were generated or analyzed during the current study.

Declarations

Conflict of Interest

Daphne S.L. Gardner received speaker fees and honorarium from Roche, Sanofi, BD/embecta, Abbott Diabetes Care, Medtronic, and NovoNordisk. Received advisory fees from BD/Embecta, DKSH and NovoNordisk. Michael Villa received speaker fees and honorarium from Bayer. Edward Mahoney and Shailendra Bajpai are employees and stockholders of BD/embecta (formerly BD Diabetes Care), manufacturer of a range of safety-engineered devices. No potential conflict of interest relevant to this article was reported by Banshi Saboo, Jothydev Kesavadev, and Norlaila Mustafa.

Ethical Approval

Ethical approval was not applicable or required as this article is based on previously conducted studies and does not contain any new data from experiments on human participants or animals performed by any of the authors.