Design,implementation,and evaluation of complex telenutrition interventions used for diabetes self-management and monitoring

Abstract

Background

The global increase in diabetes prevalence has heightened the need for interventions focused on health promotion, prevention, management, and clinical monitoring. The challenges for affected individuals and health professionals are numerous, including monitoring diet, physical activity, medication adherence, and blood glucose levels.

Objective

This article aims to describe the current state of knowledge on the processes of design, implementation, and evaluation of complex telenutrition interventions used for diabetes self-management and monitoring with a health professional, with the goal of improving adherence, usability, and sustained use.

Methods

Research was conducted using the MEDLINE, CINAHL, and Embase databases, limited to articles in English and French on the development or evaluation of complex telehealth interventions for type 2 diabetes or pre-diabetes. Three reviewers independently selected each of the articles, and the principal researcher analyzed them.

Results

The findings reveal strong end-user involvement, including patients and various health professionals, in the design and evaluation of digital tools, promoting multidisciplinary diabetes management. Telehealth interventions were designed to be used across various platforms and devices, enhancing accessibility. The analysis highlights the importance of ease of use of monitoring technologies, with a trend toward automation and integration with wearable devices for simplified monitoring and rapid adjustments. The analysis also emphasizes the need for rigorous usability evaluations to ensure that technologies meet user needs. Interventions that incorporated theoretical models of behavior change tended to show high levels of user satisfaction and adoption and encouraged patient engagement in managing their condition.

Conclusion

This review reveals that complex telenutrition interventions represent a significant advancement for diabetes management. They enable close collaboration between patients and health professionals, enhancing effective diabetes self-management through more accessible and user-friendly digital platforms. The results highlight the importance of user-centered design of telehealth solutions for the success of these initiatives.

Introduction

The global increase in diabetes prevalence has led to a need for interventions in health promotion, prevention, management, and clinical monitoring by health organizations. According to the International Diabetes Federation (IDF) in 2019, diabetes accounted for 11.3% of deaths worldwide, and it is projected that by 2045, 700 million adults will be diagnosed with diabetes globally. ¹ In Canada, nearly 11 million individuals live with or are at risk of developing diabetes. ² It is crucial to note that nutrition plays a central role in the onset, maintenance, and exacerbation of diabetes. ³ Managing diabetes presents numerous challenges for both the affected individuals and health professionals. Patients need support and tools to manage their diet, physical activity, medication, and insulin dose calculations when necessary. ⁴ To this end, health professionals must provide motivational and tailored support to each patient based on their needs, goals, and health outcomes. They must also interpret health indicators and make the necessary medication adjustments to ensure proper glycemic control and prevent complications associated with diabetes. ⁴

Nowadays, digital tools are an essential lever for managing and monitoring diabetes by facilitating communication between patients and caregivers and promoting interdisciplinary collaboration.^5,6 In this context, telenutrition refers to the use of digital and communication technologies to provide remote nutrition care. Telenutrition is especially relevant in diabetes care, where nutrition plays a central role in both prevention and management. These interventions can range from simple remote consultations to more complex systems that integrate dietary tracking, personalized feedback, educational content, and continuous follow-up by a health professional. ⁷ This broader understanding is consistent with recent literature and professional definitions, which emphasize the importance of integrating lifestyle, nutritional, and clinical components in chronic disease care.^7,8 When thoughtfully designed, complex telenutrition interventions can improve the accessibility and personalization of care, while enhancing patient engagement and treatment adherence. ⁹ Although digital health technologies are increasingly used to support diabetes self-management, the role of registered dietitians remains insufficiently integrated into many telenutrition interventions. As regulated health professionals specialized in evidence-based nutrition care and behavioral support, dietitians are uniquely positioned to guide patients through the complex dietary adjustments required in the management of type 2 diabetes.^10–12 Evidence suggests that when dietitians are actively engaged in the development and delivery of digital interventions, outcomes such as dietary adherence, glycemic control, and patient satisfaction are significantly. ¹¹ Their contributions are also critical to the development of high-quality educational content and the design of user-friendly nutritional tracking tools.^11,12 Their involvement during the design and evaluation phases can also help optimize user interfaces, adjust feedback algorithms, and ensure clinical coherence throughout the intervention.^13,14

The autonomous management of chronic diseases, especially diabetes, has become a cornerstone of modern public health strategies, driven by the rapid evolution of digital technologies. These technologies have introduced a variety of tools aimed at enhancing glycemic self-monitoring, follow-up, and patient education. ¹⁵ For example, continuous glucose monitoring systems provide real-time surveillance, while digital educational platforms offer personalized information for disease management. ¹⁶ Although the overall effectiveness of these devices still requires rigorous scientific validation, preliminary studies suggest they could significantly improve health outcomes for people living with diabetes.^17,18 Research shows a correlation between the use of these tools and better management of HbA1c, a reduction in hypoglycemia episodes, and improved quality of life. These digital tools, such as mobile apps for blood glucose tracking and telemonitoring systems, transform traditional consultation practices by offering an interactive interface where patients and caregivers can monitor progress towards predefined health goals together. ¹⁹ This technological synergy fosters a holistic and personalized care approach, making the patient a central and engaged actor in the therapeutic process. By enabling individuals to monitor their own health accurately and share this information in real time with their care team, health technologies strengthen the patient-centered care model, leading to better alignment between medical interventions, individual needs, and preferences. ²⁰ Moreover, this collaborative approach encourages patients to make informed decisions about their health and adjust their behavior to achieve optimal outcomes. ¹⁹ However, it is crucial to continue evaluating the effectiveness of these tools in clinical practice, understand the factors influencing their adoption and usability by patients and health professionals, and ensure these innovations are accessible to all population segments.

However, few devices have actively sought the involvement of health professionals and patients in their development.^20,21 The lack of knowledge about existing technologies, resistance to change, cost, accessibility of these technologies to the elderly and those with cognitive impairments, integration with local software, data security, device reliability, and potential technical problems related to internet connectivity are all obstacles expressed by users. ²⁰

Concerns related to the use of health technologies, literacy levels of patients, and other challenges, underscore the importance of considering the overall architecture of the technologies developed as well as their context of use. The quality of web platforms and applications is conditioned by their ergonomics, the accuracy of their algorithms, and the data entry methods. ²² Coupling these technologies with clinical expertise and the social support of health professionals could enable more tailored therapeutic management of people living with diabetes.

Ease of use is a critical element in the design of health technologies. It refers to the simplicity with which a user can employ a tool, an app, or a device without encountering difficulties or complications. ²³ Ease of use is a crucial factor in user adoption and satisfaction, as it directly affects their experience and ability to effectively use the product. ²³ In the context of health technologies used for monitoring of people living with diabetes by clinicians, diabetes being a chronic condition requiring constant monitoring, calls for various technological tools to ensure optimal patient follow-up. The use of well-designed technological tools can improve the quality of care for people living with diabetes by providing clinicians with accurate and timely information. ²⁴ Conversely, if the tool is poorly designed, it can lead to errors in clinical decision-making and compromise patient safety. Monitoring diabetes often requires the use of blood glucose measuring devices, such as glucometers, and apps for tracking nutrition, medication, and physical activity. Devices for monitoring blood glucose, such as glucometers, require absolute precision and reliability to avoid incorrect readings that could result in inappropriate treatments. Similarly, an intuitive user interface for these devices is essential. A study revealed that clinicians and patients often face usability issues with poorly designed systems, which can lead to misinterpretations of data and, consequently, inappropriate treatment adjustments. ²⁵ Privacy and security also remain concerns. Clinicians must be able to access data quickly and easily while ensuring the security of sensitive patient information. The complexity of managing diabetes is such that errors in data entry or interpretation can have serious consequences, as illustrated by a study highlighting the implications of technological errors in diabetes monitoring. ²⁶

The use of theory in the development of interventions aimed at changing behaviors is associated with high user satisfaction.^27–30 This is based on the fact that integrating psychological, sociological, or behavioral theories into the design of interventions allows for a better understanding of the motivations, needs, and reactions of the individuals concerned. By understanding the psychological mechanisms underlying a given behavior, designers are better prepared to anticipate potential obstacles, resistance to change, and emotional reactions of users. This foresight allows for the design of telehealth interventions that are more acceptable, personalized, and less intrusive, contributing to better end-user satisfaction.

Usability and adherence are crucial variables as they determine the success of a telehealth intervention. ²³ Furthermore, it is necessary to precisely study how each component of an intervention that combines self-management and monitoring with a health professional should be developed to improve adherence, usability, and retention. In this article, “adherence” means the alignment between the patient's behavior and the proposed treatment. “Usability” is characterized by the user's satisfaction with the technology and its ease of use.

We use the term “digital tools” as defined by WHO classifications, which include digital and mobile technologies, such as websites, mobile apps, telehealth, and wearable devices. ³¹ We define complex interventions as those that include multiple interacting components and are characterized by the number and difficulty of behaviors required by those who deliver or receive the intervention, the different organizational levels targeted by the intervention, the number and variability of outcomes, and the degree of customization allowed by the intervention.

Research is therefore needed to guide the successful development of telehealth solutions and to gain a better understanding of the key elements and fundamental components of their design, implementation, and evaluation for diabetes management. A scoping review of the literature is presented to describe the scope of the knowledge concerning the development and consideration of factors that may influence their use and to map the evidence to inform future directions.

Objective

The aim of this scoping review is to describe the current state of knowledge describing the processes of design, implementation, and evaluation of complex telenutrition interventions used for self-management and monitoring of diabetes with a health professional, with the goal of contributing to the improvement of adherence, usability, and sustained use.

Methodology

Our scoping review was organized based on the five stages outlined by Arksey and O'Malley, ³² with enhancements suggested by Levac et al. ³³ These stages include 1—formulating the research question, 2—identifying relevant literature, 3—selecting studies, 4—extracting and mapping data from the literature, and 5—synthesizing and presenting the findings. The process is described in detail in the sections below, following the guidelines provided by the PRISMA Extension for Scoping Reviews (PRISMA-ScR). ³⁴ Furthermore, we incorporated additional recommendations from Levac et al., ³³ including independent review of full-text articles by two reviewers and collaborative development of the data mapping form with the research team.

Research question identification and study selection

This review aims to address the following research questions: “What are the fundamental components of complex telenutrition interventions used for self-management and monitoring of diabetes with a health professional that contribute to the improvement of adherence, usability, and observance? How should these components be developed, implemented, and evaluated?” These questions arise from the lack of consensus in the scientific literature regarding the design, implementation, and evaluation of telenutrition solutions. Enhancing knowledge on this topic could contribute to the future development of more efficient telehealth interventions. The research questions and objectives were developed based on the expertise of the research team and a preliminary analysis of the literature on the subject. According to the scoping review methodology, this review includes studies using different approaches and research designs.

Identification of relevant studies and selection of studies

The search strategy was developed in collaboration with a librarian from the Université. de Montréal specializing in health. Keywords based on the three concepts of telehealth, nutrition, and diabetes were identified by reviewing relevant articles, their references, and associated keywords. A systematic search was conducted in the scientific databases Medline, CINAHL, and Embase, covering the period from January 2010 to March 2022. We reviewed the bibliographies of the included documents to identify additional studies. Only articles published in peer-reviewed scientific journals were examined. In accordance with the framework established by Arksey and O’Malley, no quality assessment was conducted. ³³

In order to best align the selected studies with the research question, we used an iterative process of developing inclusion and exclusion criteria throughout our research. The searches were limited to scientific publications in English and French, describing the design or evaluating the effects of a complex telehealth intervention on the therapeutic management of people living with diabetes or pre-diabetes, and involving self-management and monitoring by a health professional. In this article, complex interventions had to be digital, interface with the patient, address the prevention, management, or monitoring of type 2 diabetes or prediabetes, and potentially undertake an analytical treatment in addition to simple data collection and display. The retained articles had to describe in detail the design process of these systems, including theoretical foundations and user feedback, strategies and challenges of practical implementation, as well as evaluations measuring clinical impact and user acceptability. Only articles published since 2010 were selected to account for the widespread adoption of smartphones. Single-component studies and/or those not incorporating a health professional, populations studied outside the age range of 18 and older, publications concerning type 1 diabetes, publications lacking empirical data, and literature reviews were excluded. However, articles on literature reviews served to support the framework and discussion. All search results were merged into the reference management tool Covidence, and duplicates were removed. ³⁵

Two reviewers decided on the inclusion of relevant studies based on the title, abstract, and full text (the review team was composed of CM, DK, and BV for selection based on the title and abstract, and CM and KV for selection based on the full text). Differences were fully discussed until a consensus was reached. Full texts of relevant studies were retrieved for further analysis.

Data mapping

A data extraction form was developed by the research team and included the following elements: characteristics of the study (title, participants, outcomes of interest, and effectiveness), characteristics of the intervention (a brief description of the intervention, self-management components, and health professional monitoring components), and the advantages and limitations of the intervention and research according to the authors or reviewers.

Collecting, summarizing, and reporting results

In line with the framework of Arksey and O'Malley and Levac et al, ³³ descriptive numerical summaries and thematic analyses using Nvivo 1.7 software were employed for data analysis, resulting in an approach akin to a “narrative review.” All articles were coded in their entirety through multiple rigorous iterations during which the first version of the codebook was developed and continuously refined.

Theoretical model of Technology Acceptance Model 3

The Technology Acceptance Model 3 (TAM3) was used to analyze the data as it enables the assessment of users’ receptiveness to information technologies. It focuses on the factors influencing the acceptance and use of technologies, such as perceived usefulness and perceived ease of use. The TAM 3 model is particularly relevant for analyzing the ease of use of technologies because it extends the original TAM by incorporating additional explanatory variables that affect intention and actual system use. By integrating factors such as social influence and facilitating conditions, TAM 3 provides a holistic evaluation of technology adoption in the healthcare field.^36,37 It also allows for an assessment of how the features of an application impact user experience and their commitment to adopting and using the technology consistently.

Ease of use analysis

In this paper, “ease of use” refers to the simplicity with which a user can interact with a tool or technological application to perform a specific task in the health domain.^38,39 This encompasses not only the ease of learning and initial use but also the ability to use the tool or application effectively and satisfactorily over the long run.

We classified the ease of use of technologies for the monitoring of people living with diabetes on a simple scale: easy, moderate, and difficult. “Easy” refers to features where data is collected passively or automatically, without requiring direct user intervention (e.g. connected scales for weight measurement). “Moderate” refers to features that require minimal or semi-automated interaction. The user provides some input or interacts partially, but the process is assisted by tools (e.g. meal planning tools using predefined data). “Difficult” refers to features where data must be manually entered, requiring effort and precision from the user (e.g. manual logging of consumed foods or physical activities). This classification was applied to the essential features of the technologies in question.

Results

Study characteristics

Database searches identified 3755 articles, with 995 duplicates removed. The remaining 2760 articles were part of the initial selection of abstracts and titles, of which 2313 articles were excluded. In total, 15 articles were included in this review (see figure 1 for the flow diagram of selected studies). The qualitative analysis of the 15 articles in Nvivo 1.7 resulted in the encoding of 626 text segments, spread over 67 codes.

Figure 1.

Flow diagram of study selection.

The studies describing information on complex telehealth interventions used for self-management and monitoring of diabetes with a health professional originated from 10 countries. Four studies were conducted in the United States,^27,40–42 three studies came from South Korea. A single studied examined came from Germany, ²⁸ Iran, ²⁹ de Norway, ⁴³ Singapore, ³⁰ South Africa, ⁴⁴ Spain, ⁴⁵ Switzerland, ⁴⁶ and the United Kingdom ⁴⁷ (Figure 2).

Figure 2.

Country.

The publication year of the selected studies ranged from 2012 to 2022. Most were published recently, with 10 articles published in the last 5 years.^{29,30,40,42,43,46–50} Nearly half were feasibility studies with interventions lasting between 3 months and 1 year.^{27,30,40,43,46,47,49} One third were design studies lasting 6 months^{28,29,41,44,45}; one cohort study ⁴³ and one pre-post study ⁴⁸ were also retained.

Theoretical foundations of complex telenutrition interventions

Six of the 15 interventions clearly articulated the use of at least one theoretical model that informed their design. Among them, three highlighted that the development of the intervention is based on a theory of behavioral change including the Theory of Planned Behavior, the Transtheoretical Model, Social Cognitive Theory, and the Health Belief Model.

Berman et al. relied on the Theory of Planned Behavior, Social Cognitive Theory, and Behavioral Economics. ⁴² Koot et al. used Motivational Interviewing and the Health Belief Model as intervention approaches, ³⁰ Salari et al were inspired by the Transtheoretical Model, ²⁹ while others were based on technology acceptability models including the Iterative Self-Management Pathway, ²⁸ the Model for Assessment of Telemedicine, ²⁸ and the TAM. ²⁸ In the remaining 10 studies, no theoretical foundation was reported, or suggested the use of a theoretical model.^{40,41,43–50}

Design strategies

The design strategies involved conduction user research such as through interviews, surveys, and usability tests to gather information on user behavior and preferences. Based on this research, designers created usage scenarios to guide the design process, as in Hidalgo et al., where different devices and systems were tested for the development of gluUCModel, a multi-platform web application made accessible via various devices and browsers, eliminating device type restrictions. ⁴⁵ Berman et al. described the development of an application based on human-centered software design principles, with preliminary usability tests. ⁴² The application evolved throughout the study, including an intelligent conversational bot to enhance user interaction. Jeon et al. designed the application based on user feedback and focus groups, incorporating personalized interventions and various features for a complete user experience. ⁴⁸

Another strategy employed by Kim et al. involved the use of the information-motivation-behavioral (IMB) skills model to enhance diabetes management. This approach integrated components of IMB skills to provide a comprehensive solution. ⁵⁰ Modave et al. adopted an Agile methodology for the development of DiaFit, involving end-users and stakeholders in an iterative development process. ⁴¹ This approach is characterized by focusing on the incremental delivery of the technological solution, each iteration building on the previous one, ⁵¹ allowing for quick feedback and adjustments, thus swiftly responding to evolving needs or organizational conditions. ⁵² Schmocker et al.'s project involved various partners and incorporated feedback throughout the development of the GlucoMan app. ⁴⁶ These studies highlight a shift towards adaptive, participatory, and evolutionary design processes, aiming to provide accessible and customized technological solutions for diabetes management.

Behavioral change strategies

Studies revealed that behavioral change strategies for diabetes management used a multifactorial approach that integrated various psychological and behavioral models and techniques. Berman et al. emphasized the practice of lifestyle medicine and the use of health coaching based on motivational interviewing. Jeon et al. ⁴⁸ and Kim et al. ⁵⁰ developed an application based on the IMB-DSC theoretical framework to influence IMB skills, which in turn affect self-care behaviors and physiological outcomes. Lamprinos et al. addressed patient empowerment through an Information and Communication Technology-based framework for self-management, ²⁸ while Salari's team uses the Transtheoretical Model of Stages of Change to assess and encourage self-management behaviors in diabetes. ²⁹ These approaches demonstrate the importance of placed on active patient engagement in managing their disease, using both direct interventions and the support of digital technologies to foster lasting and significant changes in health behaviors.

Stakeholders involved

The reviewed studies reveal substantial variety in participant profiles, underscoring the complexity and heterogeneity of the people living with diabetes. Participants are generally middle-aged, ranging between 50 and 54 years of age, with a notable predominance of women. From a demographic standpoint, Koot et al. illustrate ethnic diversity with Chinese, Malay, and Indian participants, while Berman et al. show a broad geographical spread with participants from 38 US states. Moreover, the studies analyzed were conducted in ten different states. This geographical and ethnic diversity can influence diabetes management strategies, emphasizing the importance of a personalized approach that takes into account local public health strategies.

In the field of telehealth, collaboration with stakeholders has allowed for the gathering of diverse expertise, leveraging resources, and ensuring that research and scientific projects meet the real needs of society. The diversity of professionals involved such as nurses, doctors, dietitians, and psychiatrist, indicates a holistic approach to diabetes management and highlights the importance of a multidisciplinary approach. Thirteen studies involved patients in the design and/or usability evaluation of digital tools. Seven studies involved doctors,^{27,28,41–43,45,46} four studies involved nutritionists,^30,40,42,44 three studies involved a nurse,^42,43,50 and one involved a psychologist. ⁴² This plurality of specializations reflects the complexity of diabetes management and underscores the need for a coordinated, multidisciplinary approach. Collaboration among different specialists is essential to address the multiple facets of diabetes, as illustrated by the range of professionals involved in the examined studies.

Usability

The analyzed studies highlight the diverse functionalities of diabetes management applications, spanning medication management, nutritional monitoring, physical activity tracking, health record digitization, and glycemic control. While most tools offer moderate usability, the integration of advanced technologies like barcode scanning, connected devices, and real-time data sharing enhances their effectiveness and user engagement. However, variability in implementation and limited focus on certain areas, such as weight tracking and remote clinical support, underscore the need for greater standardization and comprehensive design to optimize patient self-management and healthcare provider support. Finally the absence of features classified as “difficult” in the results of the analyzed articles indicates that no functionalities were deemed overly complex (see Multimedia appendix 2: Usability analysis of monitoring technologies for people living with diabetes in the selected studies).

Medication

Compiled studies provide insights into digital features aiming to optimize therapeutic compliance and medication management. Most of the applications described had moderate usability featuring manual entry of medications by the patient from a medication directory in seven studies.^{27,29,41,43,46–48,50} This functionality, often complemented by reminder systems, helped users track their medication intake and maintain treatment adherence. The act of manual entry may also enhance patient awareness of the importance of medication in controlling diabetes. Some studies present a notable advancement in usability with the capability for medication entry via barcode scanning, thereby reducing the risk of entry errors and simplifying the process for the patient.^44,46 Other studies do not provide details on medication management functionalities, indicating either a focus on other aspects of diabetes management.^{28,30,40,42,45,49} Consideration of medication in diabetes management applications is essential, as it has a direct impact on glycemic management and the prevention of complications. However, the diversity of approaches and ease of use vary significantly between applications.

Nutrition

Nutritional monitoring through mobile apps is a crucial element of integrated diabetes management. Analysis of available studies reveals a variety of features offered by apps to assist patients in tracking their diet. The apps mentioned in the majority of studies represent moderate usability, including nutritional databases that allow users to log their food intake.^{28,29,41–50} These databases, often linked to automated calculations of caloric and nutrient intake, provide essential feedback for a balanced diet tailored to the needs of people living with diabetes. The food entry systems, whether manual or via barcode scanning, play a fundamental role in facilitating the recording of dietary intake.^30,40 They provide accurate monitoring while also helping to educate patients about the nutritional content of the foods they eat. They not only enable accurate tracking but also educate patients about the nutritional composition of consumed foods. The study by Jung et al. ⁴⁰ brings additional ease of use with a feature that allows dietitians to view images of food transmitted by patients’ smartphones, which could improve the accuracy of nutritional tracking and provide personalized dietary advice.

Physical activity

Tracking physical activity in diabetes management applications aims to support adherence to an active lifestyle while providing clinicians with accurate data for monitoring and adjusting therapeutic plans. Data from selected studies illustrate different approaches and technologies used to record and analyze physical activity. Seven applications use built-in phone pedometers and connected devices like the Apple Watch to measure steps automatically and passively.^{30,41,43,46–48,50} This technological integration allows for effortless data collection for the user and provides an accurate estimation of energy expenditure. Four other studies highlight the necessity for the user to manually enter their physical activities into the application.^28,29,45,49 Although this method can enhance patient engagement in tracking their activity, it requires consistent discipline and is likely to be influenced by the accuracy of the data entered by the user. While this method may reinforce patient engagement in monitoring their activity, it requires constant discipline and is susceptible to being affected by the accuracy of the data entered by the user. Although some studies did not integrate a physical activity tracking feature into their devices,^27,40,44 the results reflect a trend towards automation and integration with wearable devices for activity monitoring.

Health management

The digitization of health records represents a major evolution in healthcare management. Integrating elements of the health record into applications and telehealth platforms facilitates integrated and patient-centered management, allowing quick and accurate access to crucial medical information. Features relating to the health record are found to be of moderate usability due to the complexity of managing additional information on diabetes such as lab values or medical examinations. Three studies present moderate ease of use characterized by manual entry of medical data and laboratory results,^44,45,48 while only one study had a high ease of use through automatic synchronization of clinical results. ²⁷ The examined studies demonstrate a variety of methods for managing digital health records. These systems require active patient participation and can enhance understanding and engagement in managing their health. On the other hand, some applications seem to integrate features for manual entry of medical data, without specifying whether these are complementary to automatically collected data or entirely dependent on the user. Nevertheless, the integration of features that refer to the health record was only concerned in four studies, possibly indicating an area for improvement for the design of future applications. These features offer considerable benefits such as consolidating health data, facilitating communication between patients and healthcare professionals, and personalizing care. However, variability in the implementation and use of these systems indicates the need for increased standardization to maximize their effectiveness.

Weight tracking

Weight tracking plays a crucial role in diabetes management, given its close link to glycemic control and cardiovascular risk. In some applications, the integration of connected scales allows for automatic and passive recording of weight, offering high ease of use.^30,42,46 This feature reduces the cognitive burden for the patient and ensures the accuracy of the data entered. Conversely, applications analyzed in other studies require manual weight entry by the patient.^43,45,48 While this method involves greater patient involvement, it may also reinforce awareness of the importance of weight in diabetes management. The majority of studies (n = 9) do not provide information on weight tracking, which might indicate a focus on other aspects of diabetes management but represents a missed opportunity to integrate relevant data into the overall health profile of the patient.

Glycemic control

Glycemic control is a crucial pillar in diabetes management, and current digital technologies provide innovative tools to enhance this process. Analysis of various applications revealed a diversity of approaches for glycemic monitoring, from manual data entry to the integration of connected devices. Manual glucose entry, although widespread in the studied applications (n = 9), involves the user actively measuring and recording their glucose levels via a glucometer.^{27–30,41,43,45,49,50} This method requires constant patient involvement but can also be a source of errors and forgetfulness. More advanced applications connect directly to continuous glucose monitoring devices like Dexcom6, allowing passive and automatic data recording.^46–48 This integration reduces the workload on the patient and facilitates more accurate and real-time glucose monitoring. The study by Jeon et al. highlights the use of glucometers connected via Bluetooth, which not only automatically record data but also provide feedback on glucose trends to enhance users’ behavioral skills. ⁴⁸ Finally, the approach by Ku et al. illustrates the use of personalized text feedback based on a comprehensive evaluation of the daily glycemic profile, allowing for personalized and timely intervention, potentially optimizing glycemic control and adjusting treatment based on daily fluctuations. ⁴⁹ These digital tools transform glycemic control into an integrated process, enabling patients to better understand and manage their condition.

Therapeutic education tools

The incorporation of therapeutic education tools in diabetes management applications represents a significant advancement in disease self-management. The analyzed applications offer various educational features that foster the acquisition of knowledge and skills necessary for effective diabetes management. Seven studies provided accessible and user-friendly educational material in their applications to support therapeutic education of patients and facilitate knowledge transfer.^{27–30,45,46,53–55} This educational content was available in the form of article libraries, e-learning lessons, and learning videos, promoting nutritional and therapeutic education for users. The study by Quinn et al. illustrates the use of learning libraries and digital logbooks, enriched with over a thousand self-management messages, integrated into a personalized feedback algorithm. This suggests an adaptive learning strategy where the patient receives information specific to their clinical and behavioral profile. ²⁷ Ku et al. introduce a methodology of social engagement through social network groups and personalized text feedback. ⁴⁹ This collaborative education model can enhance understanding and application of diabetes management knowledge in a social and supportive context. Hidalgo et al. and Salari et al. demonstrate the effectiveness of online courses, personalized recommendations, and automated interventions based on the analysis of patient-entered data.^29,45 These strategies provide tailored educational content that can be integrated into the patient's daily routine. Bradway et al. utilize electronic questionnaires and follow-up messages to create ongoing dialogue between patients and care providers. ⁴³ This regular communication can lead to better adherence to prescribed treatments and timely adjustments of care plans. Berman et al. focus on improving cooking skills, offering educational content to encourage the preparation of healthy meals at home, to support behavioral changes in eating and strengthen self-monitoring of dietary habits. ⁴² Finally, Jeon et al. integrate evidence-based personalized recommendations, a diary, and recording of self-care behaviors to support theoretical learning and practical implementation of specific behaviors. ⁴⁸

Remote clinical support

This feature allows health professionals to remotely monitor patients’ glycemic levels, treatment adherence, and lifestyle modifications. This enables quick identification of potential issues and intervention before they worsen. Most studies provided high ease of use for clinical support features by allowing clinicians to monitor their patients’ health data in real time.^{27,28,40,42,43,45–50,56} This access facilitates longitudinal monitoring of patients and enables care providers to better understand the progression of the disease and the effectiveness of treatment. This could be particularly useful for real-time treatment adjustments and ensuring continuous monitoring.^{28,43,45–47,49,50} Other studies go further by introducing a behavioral support dimension through instant messaging features and scheduling coaching calls. This not only allows for rapid and effective communication but also provides more personalized and responsive support that enhances patients’ behavioral skills and communication with care providers.^27,42,48,50 These interventions can prove useful for patient adherence and offering emotional and educational support. Feedback on glucose trends can help patients better manage their condition.

Usability evaluation

Usability evaluation of digital tools is an assessment of ease of use and user experience. It involves analyzing the user interface, user interactions, and user feedback to determine if the technology is easy to use and meets the needs of end-users. Methods used for usability evaluation in the studies included tracking technology navigation,^27,43,53 System Usability Scale (SUS) questionnaire,^28,43 post-consultation questionnaire, ⁴³ acceptability questionnaire, ⁴⁴ telephone recordings, ⁴⁴ and a pre-study survey to healthcare professionals. ⁴³ Interventions that used theoretical models of behavioral change demonstrated high levels of satisfaction and adoption. The study by Berman et al. shows that engagement with the application is measured by the number of daily interactions, indicating active and regular use of the application with an average of 4.3 engagements per day. ⁴² This study also highlights a positive correlation between the level of engagement and the improvement of health markers such as HbA1c. In Bradway et al., the use of the application reflects patients’ self-management habits, and the time spent in the application suggests significant interaction with various features. ⁴³ Jeon et al. highlight user feedback through an open-ended questionnaire, offering a qualitative perspective on usability. ⁴⁸ Koot et al. present quantitative data on participant engagement in using the application and online health courses, showing a decrease in engagement over time but maintaining certain essential interactions. ³⁰ Lamprinos et al. differentiate the use of applications between patients and doctors, with higher use of web portals and mobile apps for tracking health data and recommendations, although some features are less used. ²⁸

The absence of features classified as “difficult” in the results of the analyzed articles reflects a deliberate emphasis on usability and accessibility in the design of telenutrition interventions. These interventions are specifically tailored for diabetic patients with varying levels of digital literacy, necessitating solutions that are easy to use. This focus aligns with the recent literature which highlight the importance of reducing users’ cognitive load through automation and intuitive design.^57,58 Most of the analyzed interventions incorporate features categorized as “easy” (green) or “moderate” (orange), illustrating efforts to simplify the user experience while maintaining functional effectiveness. For example, nutritional tracking and blood glucose monitoring tools frequently rely on automated or semi-automated processes, such as connected glucometers or smart scales, thereby minimizing manual data entry and reducing the risk of user error. Additionally, the iterative development processes employed in many of these studies, including usability testing and feedback from end-users, likely played a critical role in refining the features to enhance simplicity. Features requiring significant cognitive or physical effort, which might discourage patient engagement, were likely identified, and subsequently modified or removed during these stages of design. Furthermore, the focus on features categorized as “moderate” reflects a strategic balance between encouraging patient engagement and maintaining accessible interactions. Such features, including therapeutic education and physical activity tracking, generally require minimal user input, promoting active participation without overwhelming users. Moreover, technological advancements such as mobile health applications, wearable devices, and integrated platforms have further facilitated the development of accessible and effective interventions. These innovations enable the implementation of complex functionalities while remaining intuitive for users. Overall, the absence of features classified as “difficult” underscores a commitment to designing interventions that are both effective and accessible, ensuring sustained patient engagement and adherence.

Discussion

Summary of results

The included studies focus on complex telehealth interventions designed for self-management and monitoring of diabetes in collaboration with healthcare professionals. The review highlights the importance of considering telenutrition as a complex and multifaceted intervention, rather than a simple digitization of dietary counseling. Complex telenutrition interventions combine multiple interactive components, such as automated monitoring tools, behavior change techniques, personalized education, and clinical supervision. When designed based on theoretical frameworks and adapted to users’ contexts and capacities, these interventions can support long-term adherence and improve diabetes self-management outcomes. The analysis highlights a geographical distribution across 10 different countries, with a predominance of research conducted in the United States and South Korea. The diversity in cultural frameworks and public health policies across countries may influence the implementation and effectiveness of telehealth interventions. For instance, cultural differences, healthcare infrastructure, and levels of technological accessibility among these countries could affect the generalizability of the findings. Moreover, with most studies published within the last five years, there is evidence of a growing trend toward telehealth adoption, raising questions about the maturity and long-term sustainability of the interventions studied. The types of studies included vary, encompassing feasibility studies, design studies, cohort studies, and pre-post studies. Feasibility and design studies predominate, suggesting that many telemedicine programs for diabetes management are still in their early stages of development. This underscores the need for long-term research to evaluate the effectiveness and efficiency of these interventions. In terms of theoretical foundations, six out of the 15 interventions were based on at least one theoretical model, while others did not reference any theoretical framework or demonstrated limited use of such models. The analysis emphasizes the involvement of end-users in the development process as an emerging practice. Indeed, the design strategies highlighted in the studies include user-centered approaches, such as needs analyses, usability testing, and scenario-based development.^41,42,45,46 Patient participation in the design and evaluation of digital tools is a key element, as is the involvement of various healthcare professionals in a multidisciplinary approach to diabetes care.^59,60 The applications developed are often accessible across multiple platforms and devices, eliminating restrictions related to the type of device used. ⁶¹ Some studies employed agile development methodologies, enabling rapid adjustments based on user feedback. ⁶²

The analysis also underscores the importance of the usability of diabetes monitoring technologies, with applications offering diverse functionalities for medication management, nutrition tracking, physical activity monitoring, health record integration, weight tracking, and blood glucose monitoring. Although usability levels vary, there is a clear trend toward automation and integration with wearable devices to enable simplified and accurate monitoring.

The absence of features classified as “difficult” in the results of the analyzed articles reflects a deliberate emphasis on usability and accessibility in the design of telenutrition interventions. These interventions are specifically tailored for people living with diabetes with varying levels of digital literacy, necessitating solutions that are easy to use. This focus aligns with the recent literature which highlight the importance of reducing users’ cognitive load through automation and intuitive design.^57,58 Most of the analyzed interventions incorporate features categorized as “easy” (green) or “moderate” (orange), illustrating efforts to simplify the user experience while maintaining functional effectiveness. For example, nutritional tracking and blood glucose monitoring tools frequently rely on automated or semi-automated processes, such as connected glucometers or smart scales, thereby minimizing manual data entry and reducing the risk of user error. Additionally, the iterative development processes employed in many of these studies, including usability testing and feedback from end-users, likely played a critical role in refining the features to enhance simplicity. Features requiring significant cognitive or physical effort, which might discourage patient engagement, were likely identified, and subsequently modified or removed during these stages of design. Furthermore, the focus on features categorized as “moderate” reflects a strategic balance between encouraging patient engagement and maintaining accessible interactions. Such features, including therapeutic education and physical activity tracking, generally require minimal user input, promoting active participation without overwhelming users. Moreover, technological advancements such as mobile health applications, wearable devices, and integrated platforms have further facilitated the development of accessible and effective interventions. These innovations enable the implementation of complex functionalities while remaining intuitive for users. Overall, the absence of features classified as “difficult” underscores a commitment to designing interventions that are both effective and accessible, ensuring sustained patient engagement and adherence.

Finally, the analysis highlights the importance of therapeutic education tools and clinical support, which allow healthcare professionals to remotely monitor patient data and intervene proactively. Usability evaluations were conducted using indicators such as navigation tracking and various questionnaires, including the SUS, which facilitated the development of technologies that meet user needs. Applications incorporating theoretical behavior change models encouraged patient engagement in managing their condition and showed high levels of user satisfaction and adoption.^28,29,48,50 Additionally, the absence of clearly articulated theoretical foundations in 10 out of the 15 studies highlights the need for better conceptualization and theoretical justification of telehealth interventions, which could also impact the ability to replicate and adapt these interventions across different contexts. While the involvement of end-users is a promising strategy, it is unclear whether the populations studied are representative of the diversity of people living with diabetes. For instance, the focus on middle-aged participants and a predominance of women in some studies may limit the understanding of the needs and outcomes of other demographic groups, particularly those who are underrepresented or marginalized. Finally, although the studies demonstrated various design strategies and behavior change approaches, the diversity in methods used to evaluate usability and technology adoption suggests that more standardized evaluation frameworks may be needed for greater consistency. High satisfaction and adoption rates for interventions based on theoretical behavior change models indicate that such approaches could be prioritized in future developments.^14,63

Recommendation for future designs

Data from international studies on telehealth interventions for diabetes management offer valuable insights for designing future platforms that support diabetes self-management and facilitate monitoring by health professionals. These platforms should be designed based on proven theoretical models such as the Theory of Planned Behavior and the Transtheoretical Model to achieve behavior change.^64–66 Design strategies should include the participation of end-users to ensure that the tools precisely meet their needs and preferences. Adopting agile and user-centered approaches, which allow for quick adjustments based on user feedback, is essential.^59,62,67 This includes establishing multi-platform systems that are accessible, simplifying the entry of medical data, and integrating therapeutic education tools, which together promote comprehensive and effective disease management. ⁶² Features should promote active self-management of diabetes, with tools to track medication, nutrition, physical activity, and blood glucose levels. ⁶⁸ The use of connected devices for automatic data collection and the possibility of immediate feedback on health data can enhance patient adherence and the accuracy of monitoring. Moreover, remote clinical support must be integrated into everyday technology to facilitate patient adoption while ensuring a high level of data security. ⁶⁹

However, this review has identified several limitations in the existing literature that warrant particular attention in future research. Moreover, studies do not adequately address mechanisms for maintaining long-term patient engagement. While digital tools may initially generate strong interest, factors such as technological fatigue, user attrition, and the need for continuous engagement strategies remain underexplored. Dynamic approaches, such as personalized content and reward mechanisms, are necessary to support long-term adherence. ⁶⁹ Expanding the scope of investigation to include diverse populations and more in-depth contextual analyses could provide a more comprehensive understanding of the opportunities and challenges associated with telehealth interventions. ⁷⁰ Finally, the organizational and economic dimensions of interventions, such as implementation costs, system interoperability, and resistance to change, are crucial factors to explore in order to ensure the sustainability of these interventions within healthcare settings. ⁷¹ Additionally, particular attention should be given to usability evaluation, using validated instruments, to ensure that platforms are intuitive and meet user expectations. ⁷²

Scientific rigor and study limitation

The systematic search strategy employed in this scoping review was developed in collaboration with a specialized librarian, which allowed for the identification of precise keywords, thereby optimizing search results and ensuring exhaustive data collection. Additionally, reviewing the references of included articles facilitated the identification of supplementary studies, broadening the scope and completeness of the collected data. The adoption of the methodological framework outlined by Arksey and O'Malley, ³² enhanced by the recommendations of Levac et al., ³³ ensured a systematic and coherent process for formulating research questions, selecting studies, and synthesizing results. The involvement of multiple reviewers in the independent evaluation of titles, abstracts, and full texts using specialized management software (Covidence) minimized selection bias. Furthermore, data extraction and synthesis were conducted collaboratively, supported by an analysis strategy utilizing Nvivo 1.7 software. This approach provided analytical depth through multiple coding iterations that refined thematic categories and ensured consistent interpretation of the data. A significant methodological strength is the theoretical anchoring provided by the TAM 3. ³⁶ This framework offers both analytical and practical perspectives for assessing the usability and adoption of telehealth technologies. By addressing behavioral determinants of technology acceptance, TAM 3 sheds light on strategies to improve the design and implementation of telenutrition interventions.

However, in accordance with the Arksey and O'Malley framework, no formal critical appraisal of the quality of the included studies was conducted. ³² The absence of quality assessment limits the ability to weigh findings based on the scientific rigor of individual studies and may introduce bias into the overall interpretation of results. Another limitation lies in publication bias due to the restriction to articles published in peer-reviewed journals. This decision may have excluded valuable contributions from technical reports or theses. Similarly, the linguistic restriction to English and French reduces the international applicability of the findings, potentially overlooking relevant studies published in other languages, particularly from regions where diabetes prevalence is high. The inclusion period, limited to articles published since 2010, aims to reflect recent advancements in mobile and digital technologies but may exclude foundational studies that established the basis for complex telehealth interventions. Additionally, methodological variability among the included studies poses challenges for direct comparisons. Differences in design, implementation strategies, and evaluation methods complicate the generalization of results and the formulation of standardized recommendations.

Another limitation of this review lies in the insufficient exploration of critical dimensions necessary for the success of telehealth interventions. Few studies assessed the accessibility of technologies for vulnerable populations, such as older adults or those with limited digital literacy. Similarly, key aspects like data security and integration with existing systems were insufficiently addressed, despite their importance for the sustainability and effectiveness of digital health interventions. Multidisciplinary collaboration, while identified as a strength, was not consistently applied across the included studies. Limited stakeholder involvement could restrict the diversity of perspectives considered during the design of interventions. The review also lacks a longitudinal perspective on the analyzed interventions. Most included studies focus on short-term evaluations, limiting available insights into the long-term impacts of telehealth technologies on adherence, clinical effectiveness, and user satisfaction. As noted by Eysenbach et al., the effects of digital interventions can evolve over time due to factors such as technological fatigue or changing user preferences. ⁷³ Moreover, organizational and economic implications of implementing telehealth interventions were insufficiently explored. Implementation costs, challenges related to system interoperability, and institutional resistance to change are critical factors for ensuring the sustainability of digital health solutions. Greenhalgh et al. emphasize the importance of examining these dimensions to enable the successful adoption and diffusion of health innovations. ⁷¹

Finally, the review does not adequately address mechanisms for maintaining long-term patient engagement. While digital tools may initially generate strong interest, factors such as technological fatigue, user attrition, and the need for continuous engagement strategies are underexplored. Dynamic approaches such as personalized content and reward mechanisms are necessary to support long-term adherence. ⁶⁹ Expanding the scope of investigation to include diverse populations, foundational studies, and more in-depth contextual analyses, while integrating ethical and organizational considerations, could provide a more comprehensive understanding of the opportunities and challenges associated with telehealth interventions. ⁷⁰

Conclusion and future research directions

This scoping review highlights promising research avenues to enhance telehealth interventions for diabetes management that encourage proactive and personalized diabetes management. Firstly, this integration allows for simplified data collection and proactive healthcare professional interaction, enabling effective and personalized remote clinical support. The ease of use and automation of features such as medication management, nutrition, and blood glucose monitoring appear as key aspects for effective diabetes management. However, it is imperative to broaden research to include longitudinal and geographically diverse studies to capture the long-term effectiveness of telehealth interventions in various cultural contexts and health systems. Special attention should be given to applying solid theoretical models in designing interventions to strengthen adherence and behavior change in people living with diabetes. Furthermore, increased customization of interventions based on participatory and user-centered design could ensure that technologies align with the specific needs and preferences of end-user subgroups, thus improving engagement and usability. Concurrently, exploring the standardization of health data, interoperability, and system integration is essential for coordinated patient management. Future research should also compare different care delivery models in telehealth to establish the most effective ones and evaluate the impacts of interventions not only on clinical outcomes but also on psychosocial, behavioral, and economic aspects. Finally the incorporation of emerging technologies such as artificial intelligence and connected devices represents a significant pathway for optimizing remote diabetes management, thereby offering substantial potential to improve care and patient self-management.