Development of a T2D App for Elderly Users: Participatory Design Study via Heuristic Evaluation and Usability Testing

Abstract

Diabetes management applications effectively help patients monitor blood glucose levels and adjust diet and exercise plans. However, most diabetes management apps, including the SugarShift application developed in 2023, use generalized designs that often fail to consider the unique cognitive and physiological characteristics of elderly users, thereby impacting their acceptance and usage. This paper proposes an enhancement for the SugarShift application by introducing a mode specifically designed for elderly users. This proposed enhancement aims to directly involve elderly patients so as to meet their needs better through a participatory design approach. A heuristic evaluation conducted with six experts identified 126 usability issues and 179 heuristic violations in the current version, which has led to the recommendation to develop optimization strategies tailored for elderly users. Subsequent usability testing with 28 elderly patients and six experts emphasized the importance of interface simplicity, logical operation, and interaction quality in enhancing user experience. These factors improve usability, facilitate information processing, and reduce cognitive load. Despite the overall usability of SUS scores, significant challenges still need to be addressed in information display, system feedback, and user interaction. Recommendations for future enhancements include simplifying data entry and presentation, improving readability, and providing timely feedback to enhance usability and user satisfaction, ultimately increasing user retention.

1. Introduction

Experts expect the global prevalence of diabetes among the older adult population (ages 65–99) to reach 276.2 million by 2045, underscoring the substantial burden of this chronic disease in an aging society [1]. The World Health Organization reports a substantial increase in diabetes prevalence, from 108 million cases in 1980 to 422 million in 2014.

Diabetes exhibits a markedly elevated expansion rate in low- and middle-income nations, surpassing the rates observed in high-income countries. Specialists forecast that by 2030, diabetes will emerge as the seventh predominant cause of mortality globally [2]. Type 2 diabetes constitutes the majority of these cases (over 90%), predominantly occurring in older adults, especially those with insufficient physical activity, poor nutritional balance, and resultant obesity [3]. The impact of diabetes on diagnosed individuals combined with its associated economic burden establishes it as a significant public health issue [4,5]. Despite substantial investments in clinical care, public health initiatives, and research, the prevalence and burden of diabetes continue to escalate [6]. Evidence suggests that structured dietary and physical activity programs can prevent up to 60% of diabetes cases [7,8]. Therefore, as a chronic disease, effective management of diabetes hinges on integrating self-management support into the daily lives of patients to achieve optimal treatment outcomes.

1.1. Current Research on Diabetes Management Apps for the Older Adult

With the rapid proliferation of mobile health technologies and the widespread adoption of remote health programs, diabetes management is increasingly shifting towards digital mobile health solutions. Remote healthcare—through mHealth—has opened new avenues for diabetes self-management, demonstrating significant potential to enhance patients’ self-care capabilities [9]. Systematic reviews indicate that traditional face-to-face medical practices face multiple challenges in managing diabetes, such as ineffective prevention systems, uneven distribution of medical resources, and inadequate self-management [10]. Furthermore, recent studies highlight that digital health technologies (DHT), including artificial intelligence (AI), can significantly mitigate these obstacles. Integrating AI in tools for diabetic retinopathy detection and glucose monitoring enhances the precision and efficiency of treatments and fosters a shift toward personalized and predictive diabetes management [11,12,13]. This advancement suggests that future diabetes care will be more effective, substantially reducing patients’ burden.

Meta-analyses have examined the effectiveness of mHealth interventions for enhancing glycemic management in adults with Type 2 diabetes, reporting on their coverage, adoption, and feasibility [14]. An analysis of 56 studies involving 11,486 participants revealed that smartphone app interventions have higher coverage than SMS and web-based interventions. Utilizing mobile applications with real-time data monitoring capabilities on iOS and Android platforms, we analyzed the top five diabetes management apps in China based on download statistics. We collected essential information up to 9 August 2024, including each app’s release date, latest update, user reviews, and feature descriptions, as detailed in Table 1. Although the app stores contain over 3000 Chinese medical applications, only 50% have user ratings ranging from moderate to good [15]. Moreover, most apps in the current market are designed to cater to a broad diabetic population rather than targeting specific age groups or diabetes types, or focusing on features that are most relevant to patients.

In terms of functionality, as shown in Table 1, diabetes management apps in China primarily focus on data recording, followed by diabetes-related educational content and integrated information analysis. Researchers included 25 publications that involved smartphone apps as interventions, assessing measures that targeted two or more functionalities, such as monitoring blood glucose and uric acid, managing diets, adhering to medication, tracking exercise, and providing real-time feedback on medical advice. They observed only short-term effects, and the evidence for the expected success of some technologies, such as gamification in behavior change, remains limited [16]. Regarding interface interaction feedback, the predominant style of these apps is flat design; however, user feedback indicates issues with information overload, insufficient interactivity, and areas for improvement in information visualization and readability. Gong et al. searched for and evaluated over 2000 publicly available diabetes self-management apps in China, assessing 67 of them in depth. They found significant variability in app quality, with many apps showing low user engagement, poor adherence to guidelines, and a lack of evidence supporting health benefits [17]. Zhang et al. observed homogeneity in content and interaction modes among 20 Chinese diabetes apps, highlighting the need for a user-centered development approach to create content that professionally meets patients’ needs [18].

Table 1. Comparative Analysis of Diabetes Management Apps.

Icons & Names 1	Downloads (in 10,000 s) 2	Main Interface	Features 3	Interface Interaction	Expert Rating [19]
Sugar Nurse	1096.32		①④ ⑤⑥	Tab and grid navigation for clarity; minimalist icons; cool-toned colors; high-volume information disperses attention; dense information layout. [20]	3.7
Pioneer Bird	803.86		①④ ⑤	Tab navigation simplifies function transitions; multi-layered; interactive feedback via color and texture changes; mild interface colors [21].	3.12
Big Sugar Doctor	758.59		①④ ⑤	Tab navigation; 2–3 level structure; low recognizability of icons; text-heavy card layout leads to monotonous design [22].	3.12
Sugar Friend	731.99		①④	Tab navigation for easy switching; 2–3 level structure; cool-toned interface; recognizable launch icon, low for launch icon [23].	3.00
Sugar Circle	883.22		①④ ⑤	Tab and drawer navigation; information-dense homepage disperses attention; mild interface colors; high recognizability of function icons [21].	3.19

¹ “Sugar Nurse” is developed by Beijing Tanghu Technology Co., Ltd., located in Beijing, China; “Pioneer Bird” is from Jiangsu Wenxu Information Technology Co., Ltd., based in Jiangsu City, China; “Big Sugar Doctor” is provided by Huima Medical Technology (Shanghai) Co., Ltd., in Shanghai, China; “Sugar Friend” is developed by Fuzhou Kangwei Network Technology Co., Ltd., in Fuzhou, China; and “Sugar Circle” is from Shenzhen Aibowei Biotechnology Co., Ltd., located in Shenzhen, China; ² As of 9 August 2024, the data for platforms such as the App Store, 360 Mobile Assistant, Baidu Mobile Assistant, and Wandoujia—not every app is available for download on each platform; ³ The features include the following: ① Data Management and Monitoring, ② Medication Reminders, ③ Dietary Management, ④ Health Data Analysis and Medical Collaboration, ⑤ Education and Support, ⑥ Behavior Change and Motivation Enhancement.

Table 1. Comparative Analysis of Diabetes Management Apps.

Icons & Names 1	Downloads (in 10,000 s) 2	Main Interface	Features 3	Interface Interaction	Expert Rating [19]
Sugar Nurse	1096.32		①④ ⑤⑥	Tab and grid navigation for clarity; minimalist icons; cool-toned colors; high-volume information disperses attention; dense information layout. [20]	3.7
Pioneer Bird	803.86		①④ ⑤	Tab navigation simplifies function transitions; multi-layered; interactive feedback via color and texture changes; mild interface colors [21].	3.12
Big Sugar Doctor	758.59		①④ ⑤	Tab navigation; 2–3 level structure; low recognizability of icons; text-heavy card layout leads to monotonous design [22].	3.12
Sugar Friend	731.99		①④	Tab navigation for easy switching; 2–3 level structure; cool-toned interface; recognizable launch icon, low for launch icon [23].	3.00
Sugar Circle	883.22		①④ ⑤	Tab and drawer navigation; information-dense homepage disperses attention; mild interface colors; high recognizability of function icons [21].	3.19

¹ “Sugar Nurse” is developed by Beijing Tanghu Technology Co., Ltd., located in Beijing, China; “Pioneer Bird” is from Jiangsu Wenxu Information Technology Co., Ltd., based in Jiangsu City, China; “Big Sugar Doctor” is provided by Huima Medical Technology (Shanghai) Co., Ltd., in Shanghai, China; “Sugar Friend” is developed by Fuzhou Kangwei Network Technology Co., Ltd., in Fuzhou, China; and “Sugar Circle” is from Shenzhen Aibowei Biotechnology Co., Ltd., located in Shenzhen, China; ² As of 9 August 2024, the data for platforms such as the App Store, 360 Mobile Assistant, Baidu Mobile Assistant, and Wandoujia—not every app is available for download on each platform; ³ The features include the following: ① Data Management and Monitoring, ② Medication Reminders, ③ Dietary Management, ④ Health Data Analysis and Medical Collaboration, ⑤ Education and Support, ⑥ Behavior Change and Motivation Enhancement.

1.2. Limitations of Diabetes Mobile Health Apps in Elderly Patients

Despite the potential of mobile health applications to improve diabetes management, elderly populations exhibit lower engagement with digital technologies, especially when adapting to new technologies is required. According to the 53rd ‘Statistical Report on Internet Development in China’, only 10.3% of Internet users are 60 and above, and approximately 140 million older adults own smartphones but do not use the internet [24]. The primary reasons for these individuals’ lack of internet use include unfamiliarity with technology (60.7%) and age-related factors (19.8%) [25].

Furthermore, cognitive and physiological declines in older adults significantly restrict their ability to utilize these technologies. Studies indicate that about 44% of patients with Type 2 diabetes experience some form of cognitive impairment [26]. In terms of sensory functions, by the age of 65, approximately half of the older adult diabetic patients experience hearing loss, especially to high-frequency sounds, with losses reaching up to 60 decibels at 8000 Hz for those aged over 70. Visual impairments are shared among older adults, including corneal opacification and decreased visual acuity, with about 20% of individuals over 65 not achieving 20/40 vision even with correction. Moreover, a 30% reduction in tissue cells by age 75 leads to decreased muscle strength and coordination difficulties, while significant reductions in brain cell numbers affect cognitive and memory functions. These physiological changes directly impact older adults’ daily lives and information acquisition capabilities [27,28]. These cognitive and physiological declines directly affect their ability to interact with diabetes management applications, such as a diminished capacity to navigate complex interfaces or interpret complex information, thereby increasing the likelihood of abandoning these applications.

Optimizing the usability of interventions targeted at this demographic is critical. Nevertheless, if individuals find the management applications complex or misaligned with their needs, they might discontinue their use [29]. Consequently, numerous scholars have focused on the older adult as the target user group, investigating the impact of application interfaces, functionalities, and preferences on their adoption. Studies indicate that behavior change techniques can motivate senior citizens to use mobile applications, demonstrating the feasibility, necessity, and effectiveness of mobile interface designs based on the older adult’s cognitive load and perceptual abilities [30,31,32].

Moreover, the usability of applications significantly influences their efficiency in completing health-related tasks [33]. Previous systematic reviews of mobile application interventions have underscored the importance of enhancing the user experience during the development process, particularly regarding the challenges older users face when using mobile applications and the potential for integration into healthcare systems [33,34]. The literature focused on improving user experience highlights the need to consider incorporating alerts and reminders, usage tutorials, links to social media, and voice recognition for data entry [35,36]. Arnhold et al.’s meta-analysis highlights that diabetes applications often fall short in presenting information and providing functional operability, primarily because the design process does not fully consider the abilities, needs, and limitations of elderly users. Their analysis underscores the critical importance of ensuring accessibility in app usage for elderly patients, stressing that practical digital tools must accommodate the specific challenges faced by elderly patients to enhance usability and engagement [37]. Other findings indicate that older users report lower satisfaction and usage rates with multifunctional applications due to these usability issues [37,38].

2. Materials and Methods

Given this context, to further optimize the design and functionality of mobile health applications for elderly patients with Type 2 diabetes, and to enhance both their effectiveness and patient compliance, this study employs the most authoritative core methods of user experience research. By integrating a comprehensive framework of a Heuristic Evaluation (HE) and a System Usability Scale (SUS), the study continually improves the system interface and functionality based on feedback from experts and patients. This approach aims to evaluate and enhance the usability of diabetes management applications among elderly users, ultimately increasing their satisfaction and self-management efficacy [39,40].

Heuristic evaluation, proposed by Jakob Nielsen and Molich in 1989, is a method that rapidly identifies usability issues in interfaces based on best practices in user experience. It involves experts assessing natural systems or prototypes against usability guidelines (criteria) to detect and correct design flaws before extensive user testing, thus enhancing the overall user experience [41]. During the heuristic evaluation process, experts analyze usage scenarios and tasks to identify usability issues, categorize them into distinct heuristic categories, and assign severity ratings, providing diverse insights into usability [41,42]. Usability testing involves human–computer interaction, assessing interventions with potential target users to evaluate mobile applications [43,44]. This process ensures an understanding of the specific requirements and challenges of the user group, facilitating the identification of issues related to clarity of functionality, simplicity of navigation, and interface consistency. In a participatory design approach, research is conducted with potential end-users, making heuristic evaluation and usability testing essential tools in developing mobile health interventions. These assessments—conducted through user testing—evaluate comprehensibility, learnability, and appeal, thereby providing developers with direct feedback on which features are adequate or need improvement. Such evaluations ensure the effective integration of functions and interfaces, enhancing the application’s overall usability and user satisfaction [45,46].

2.1. Experimental Setup

2.1.1. System Description

Developers created the initial version of the SugarShift mobile health application for diabetes management in 2023. This tool is an economical and convenient resource for diabetes patients to self-manage and for the monitoring of diabetes patients, enabling efficient and rapid communication between patients and healthcare providers. The SugarShift project aims to enhance self-management among diabetes patients through features that support healthy eating, physical activity, blood glucose and blood pressure monitoring, consultations with doctors, and medication management. The intervention encompasses detailed information on healthy behaviors, provides insights on strategies for behavior change, and includes related activities.

In daily management, upon automatic login, patients are presented with a streamlined dashboard that displays the latest glucose and heart rate data, which can be updated to reflect current glucose control in real-time. The dietary module offers personalized daily meal plans with detailed calorie and nutritional information, allowing users to adjust according to their needs. The integrated exercise feature enables the setting of daily activity goals and automatically adjusts recommended calorie intake while also predicting changes in blood glucose levels post-activity to enhance users’ understanding of the impacts of exercise. Additionally, patients can manage medical appointments through the app, viewing and modifying upcoming medical consultations.

2.1.2. Usage Scenarios and Tasks

The scenarios and tasks delineate detailed procedures by which elderly patients engage with the diabetes self-management system, rooted in real-case scenarios that simulate the usage of the system by elderly patients for self-management in clinical or home settings. To ensure the scenarios closely mirrored real user experiences, the researchers assembled a task team comprising two diabetes experts, two UX design specialists, and four elderly diabetic patients. This team developed five usage scenarios and 26 distinct tasks. The team verified the content validity and accuracy of the tasks, achieving a Content Validity Index (CVI) of 0.93 out of 1.0. The five defined scenarios encompassed tasks such as viewing glucose and heart rate data, managing medications, organizing dietary plans, setting and tracking exercise goals, and handling medical appointments and consultations. Each scenario includes multiple specific tasks designed to comprehensively understand the challenges that elderly diabetes patients may encounter while using the application, considering their visual, auditory, and operational limitations. The medical appointment and consultation scenario and its associated tasks are depicted in

Table 2. Task Examples Used in Evaluation.

Scenario
Elderly diabetes patients use this management app for various daily health management activities. These activities include searching for and selecting doctors, locating nearby medical facilities, viewing other users’ reviews of doctors and medical facilities, scheduling appointments, receiving appointment reminders, consulting with doctors online, and purchasing medications.
No.	Task Description
1	Enter search criteria, filter results, and select a doctor.
2	On the doctor’s detail page, choose the time to schedule an in-person consultation
3	Use the “Nearby Doctors” feature to view recommendations for local hospitals and doctors and make an appointment.
4	View appointments and use the “Online Consultation” to text or video call with the doctor.

. An example of user usage of the “Consult” function module is shown in Figure 1. The app prototype interface has obfuscated content related to hospitals, doctors, and patient privacy.

2.2. Assessment Process

2.2.1. Phase One: Heuristic Evaluation

Nielsen’s studies indicate that three to five single-domain usability experts can identify between 74% and 87% of usability issues [41]. Research suggests that a practical usability assessment necessitates at least two to three evaluators with dual-domain expertise, who typically achieve a higher detection rate by identifying between 81% and 90% of usability issues [41]. These experts are exceptionally qualified to evaluate intricate systems like those in the healthcare sector due to their specialized proficiency in usability within the application domain and extensive foundational knowledge [47,48].

Based on prior research experiences and with the goal of effectively identifying usability issues, we established the following criteria for expert selection: dual-domain expertise, extensive practical experience, relevant academic background, medical practice experience, and a diverse team structure. According to these criteria, we invited four experts with PhDs in medical informatics and substantial experience in health informatics and usability design. Additionally, a seasoned user experience design expert and an endocrinology professional were included to ensure the comprehensiveness and depth of the evaluations. Combining these six experts guaranteed a thorough identification and resolution of usability issues within the application and a showcasing of its unique strengths, particularly in the complexities and specializations of the healthcare sector.

The evaluation process is detailed in Figure 2. In the initial phase, as detailed in

Table 3. Jakob Nielsen’s 10 Usability Heuristics [49].

No.	Heuristic	Explanation
1	Visibility of system status	Keep users informed with timely and appropriate feedback.
2	Match between system and the real world	Use familiar language and conventions from the natural world.
3	User control and freedom	Provide an ‘emergency exit’ for users to undo actions without elaborate dialogues.
4	Consistency and standards	Maintain uniformity in words and actions across similar scenarios.
5	Error prevention	Design to eliminate or check for errors before they occur.
6	Recognition rather than recall	Make information visible to reduce users’ memory load.
7	Flexibility and efficiency of use	Adapt designs for all users, allowing customization of frequent actions.
8	Aesthetic and minimalist design	Avoid irrelevant information in dialogues to focus on important content.
9	Help users recognize, diagnose, and recover from errors	Use clear language for error messages and offer constructive solutions.
10	Help and documentation	Provide easy-to-search, task-focused help and documentation when necessary.

, evaluators independently applied Nielsen’s ten heuristic principles to identify and describe usability violations [45,49]. This thorough process ensured that all potential issues were identified and recorded in a structured format, including the name, description, location, and the violated heuristic rule for each identified problem. Each expert conducted independent inspections of different system components, noting issues as they were identified (Appendix A).

After collecting data, one author (Z.L.) consolidated the issues that individual evaluators had detected and removed duplicates, resulting in a comprehensive list of unique problems. Both authors (Z.L. and X.Y.) verified this list for accuracy and validity to prevent any bias during the analysis. The list was then presented to evaluation experts to ascertain the severity of each issue. Severity assessments were based on the frequency of occurrence, impact on users, and persistence of the issues, employing descriptive statistics to summarize heuristic violations and their related severity ratings. The severity rating scale, divided into five levels from 0 to 4, is listed in

Table 4. Nielsen Severity Rating Scale [49].

Severity	Description
0	I disagree that this is a usability problem at all
1	Cosmetic problem only: need not be fixed unless extra time is available on the project
2	Minor usability problem: fixing this should be given low priority
3	Major usability problem: essential to fix, so should be given high priority
4	Usability catastrophe: It is imperative to fix this before the product can be released

. Ultimately, evaluation experts determined the overall severity of usability issues by calculating the average severity rating they assigned to each problem [44,49].

2.2.2. Phase Two: Usability Testing

From 11 to 19 August 2024, researchers reached out to the staff of the endocrinology ward at a public general hospital in Yantai, Shandong Province, China, to discuss this research project. All elderly patients diagnosed with Type 2 diabetes at this hospital were identified as potential subjects and invited to participate in a usability test. After thoroughly explaining the study details and confirming eligibility, patients who met the inclusion criteria were invited to participate in the usability testing.

Inclusion criteria

• Age ≥ 60 years;
• Diagnosed with Type 2 diabetes, as per the ‘Chinese Type 2 Diabetes Prevention and Treatment Guidelines (2017 Edition)’;
• Clear consciousness with normal communication and expression abilities.
• Previous experience with diabetes management apps and an expressed interest in participating;
• Able to read and write;
• Voluntary participation in the study.

Exclusion criteria

• Individuals with mental illnesses or cognitive impairments;
• Patients in critical or emergency conditions;
• Individuals utterly dependent on others for daily living;
• Patients with visual impairments that obstruct clear viewing of content.

Based on these inclusion and exclusion criteria, 33 elderly patients with Type 2 diabetes were initially recruited for the study. Throughout the follow-up, 5 participants were lost to follow-up, 2 withdrew voluntarily, and 3 were excluded for not meeting the research requirements. Consequently, 28 patients completed the study and were included in the data analysis. The demographic characteristics of these participants are presented in

Table 5. Baseline demographic characteristics of elderly diabetic patients.

Indicator	Category	Baseline M ± SD (%)
Gender	Male	10 (35.7%)
	Female	18 (64.3%)
Age (years)	-	67.12 ± 6.03
Educational level	Primary school and below	3 (10.7%)
	Junior high school	12 (42.9%)
	Vocational or senior high school	7 (25.0%)
	Junior college	5 (17.9%)
	University or above	1 (3.6%)
Living arrangement	Occasionally lives with others	18 (64.3%)
	Lives alone	10 (35.7%)
Monthly income (CNY)	<1500	2 (7.1%)
	1501–3000	13 (46.4%)
	3001–4500	10 (35.7%)
	>4500	3 (10.7%)
Duration of illness (years)	-	9.15 ± 4.80
Complications	Diabetic nephropathy and neuropathy	3 (10.7%)
	Diabetic retinopathy	8 (28.6%)
	Diabetic neuropathy	6 (21.4%)
	Diabetic peripheral vascular disease	4 (14.3%)
	Diabetic foot disease	2 (7.1%)
	No complications	5 (17.9%)
Duration of phone use (years)	-	3.70 ± 2.45
Smartphone comfort level	Very uncomfortable	3 (10.7%)
	Uncomfortable	7 (25.0%)
	Comfortable	12 (42.9%)
	Very comfortable	6 (21.4%)

. Age, education, income, duration of illness, complications, and smartphone usage were statistically described using mean ± standard deviation (M ± SD) or frequencies and percentages.

Before testing commenced, the app’s functionalities and objectives were introduced to the patients. Participants independently executed tasks on a prototype app via a smartphone interface based on a predefined task list. During the testing phase, patients were encouraged to describe any issues they encountered, which the research team members recorded. Following the testing, patients completed the System Usability Scale (SUS).

2.2.3. Institutional Review Board Statement

All patients signed an informed consent form before participating in the study. The Declaration of Helsinki guided the study, and the protocol was approved by the Ethics Committee of Mouping District Dayao Community Hospital in Yantai, Shandong Province (Approval number: 20240216-08).

Usability testing employed the same prototype application used in the heuristic evaluation to assess human–computer interaction, focusing mainly on the perceptions and performances of elderly patients during the experimental process [50]. Researchers conducted this evaluation in February 2024 at the Dayao Town Community Hospital in Muping District, Yantai City, China. Participants completed 26 tasks and recorded their actions on video. During the testing, they used the think-aloud protocol to explain their actions [51], while the authors noted and documented any potential issues during task execution [52].

A professional transcription service meticulously transcribed the video recordings verbatim. Two authors (Z.L and XY) reviewed and verified all textual records. Analysis of the video recordings drew inspiration from grounded theory, considering verbal and visual behaviors [53]. To ensure the analysis encompassed all viewpoints and fostered a sense of inclusivity among the audience, one author (Z.L) conducted the initial coding, followed by collaborative discussions with the other author (XY) to refine the coding boundaries. The textual records underwent a rigorous thematic analysis using an iterative coding procedure, focusing on identifying features that required redesign in terms of functionality and design. Both authors (Z.L and XY) reviewed the textual and video records independently. The categorization process involved repeatedly reading the records to detect patterns and classify usability issues.

Participants then completed the paper-based version of the System Usability Scale (SUS), offering a comprehensive overview of their subjective usability assessments based on ten standardized questions, as shown in

Table 6. System Usability Scale (SUS) [43].

No.	Question	User Feedback
		0	1	2	3	4
1	I think that I would like to use this system frequently.
2	I found the system unnecessarily complex.
3	I thought the system was easy to use.
4	I think that I would need the support of a technical person to be able to use this system.
5	I found the various functions in this system were well integrated.
6	I thought there was too much inconsistency in this system.
7	I would imagine that most people would learn to use this system very quickly.
8	I found the system very cumbersome to use.
9	I felt very confident using the system.
10	I needed to learn a lot of things before I could get going with this system.

. [43]. The standard System Usability Scale (SUS) questionnaire for this assessment is presented in Table 4. The SUS consists of ten fixed questions, half worded positively and the other half worded negatively. The analysis of SUS scores followed Brooke’s method [53]. The average System Usability Scale (SUS) score, which assesses overall user satisfaction, is derived through this specific calculation process. Each item’s score in this method ranges from 0 to 4. The calculation involves subtracting 1 from the scale position for items with positive wording (questions 1, 3, 5, 7, and 9). For items with negative wording (questions 2, 4, 6, 8, and 10), the score is obtained by subtracting the scale position from 5. These individual scores are then summed and multiplied by 2.5 to normalize the SUS scores to a range of 0 to 100 [53]. According to research by Bangor and others [43], an average score of approximately 70 is generally considered good or acceptable.

3. Results

3.1. Heuristic Evaluation Results

The heuristic evaluation revealed 126 usability issues and 179 heuristic violations, highlighting significant areas for improvement. The results of the heuristic evaluation focused on elderly patients are presented in

Table 7. Heuristic Evaluation Results for Elderly Patients.

Nielsen’s Heuristic Principles	Common Issues Encountered by Elderly Users	UI (126) 1	HV (179) 2	SS 3
Visibility of system status	Feedback: System delays in displaying operation feedback create user uncertainty.	7	9	3.2
	User Cognition: Difficulty understanding system status due to lack of explicit feedback.	6	9	3.1
	Lighting: Screen brightness adjustments are unclear, hindering visibility for elderly users.	2	3	1.4
Match between system and the real world	User Cognition: Complex terminology and symbols complicate task understanding and completion.	14	17	3.6
	Color: Inappropriate color choices clash with elderly cognition.	9	11	2.8
	Sound Alerts: Unfamiliar alert sounds are hard to comprehend.	4	8	2.4
User control and freedom	User Control: Absent immediate feedback and undo functions complicate error correction.	7	11	3.4
Consistency and standards	Interface Style: Inconsistent interface elements affect operational anticipation, like button sizes and functions.	12	15	3.8
	Color: Color inconsistency across modules challenges adaptation.	7	10	2.8
	Sound Alerts: Inconsistent alerts across interfaces confuse.	4	7	1.7
Error prevention	Function Design: Complex operations without simplification lead to frequent errors.	4	6	1.9
	Information Feedback: Delayed error notifications lead to unintended wrong entries.	6	9	2.3
Recognition rather than recall	User Cognition: No autocomplete or suggestions, increasing cognitive load.	4	7	2.1
	Function: Re-entry of information due to lack of history or shortcuts.	5	7	1.9
Flexibility and efficiency of use	Interface Style: No support for personal interface adjustments.	3	5	2.6
	Interface Visual Design: Fixed, non-customizable layout.	3	5	1.6
Aesthetic and minimalist design	Interface Visual: Cluttered interface with unnecessary animations.	8	11	3.2
	Color: Excessive decorative colors distract from information clarity.	3	5	2.7
Help users recognize, diagnose, and recover from errors	Feedback: Unclear or complex error messages take time to understand.	5	6	2.1
	Function Aspects: Ineffective error alerts or providing solutions.	3	4	2.2
Help and documentation	Documentation: Complex language and unclear guidance.	4	5	2.1
	User Cognition: Absence of intuitive guides or tutorials.	4	6	1.9
	Sound Alerts: Lack of voice assistance in documentation.	2	3	1.2

¹ UI Usability Issues; ² HV Heuristic Violations; ³ SS Severity Score.

. These issues predominantly center around system feedback, user cognition, interface consistency, and error prevention. Elderly users primarily face challenges due to the system’s complex and non-intuitive design, and lack of timely feedback and support. These factors contribute to confusion, frustration, and inefficiency among elderly users when interacting with the system.

3.1.1. Evaluation Analysis across Different Views

Researchers distributed the identified issues and heuristic violations across various views within the application, as depicted in Figure 3. This figure illustrates the distribution of usability issues and heuristic violations across different modules. The medication management module not only had the highest number of usability issues (35) but also the most heuristic violations (49), with the highest severity score (2.66). This indicates that this module significantly impacts the user experience of elderly users. The diet management and medical consultation modules followed closely, displaying numerous usability issues and violations, highlighting the critical need for design improvements in these areas. In contrast, data monitoring and exercise management modules exhibited fewer problems, yet still required optimization to meet specific user needs.

Detailed severity ratings are presented in Figure 4, which shows the distribution of the severity of usability issues across the modules. The medication management module exhibited the most severe problems, characterized by the highest numbers of catastrophic and significant issues that significantly affected the older adult users’ experience in a negative way. The diet management and medical consultation modules also showed many major and severe issues, indicating these areas as priorities for design and functionality enhancements. Conversely, while the exercise management and user service modules also presented issues, their severity was comparatively lower.

3.1.2. Heuristic Evaluation Analysis

The synthesis from the heuristic evaluation reveals that most catastrophic ratings involve defects in tasks related to disease management and technical errors, particularly system-related deficiencies in presenting essential information to users. Examples of usability issues and comments provided by evaluators are as follows:

• Medication Management
  The medication management interface presents complex, multilayered information and instructions that designers have not optimized for easy use by elderly users. Interface elements such as button sizes and touch-sensitive areas do not accommodate decreased hand dexterity and reduced tactile sensitivity typical of older users. The display of critical information, such as medication names, dosages, and times, involves small fonts or insufficient contrast, which could be more user-friendly for those with diminished vision. Furthermore, the high information density needs more visual separation, leading to difficulties in information parsing.

• Dietary Management
  There is a lack of interactive and visual educational resources, such as video tutorials and step-by-step illustrations, to guide elderly users through healthy dietary management. Designers have densely packed the dietary recording and analysis interface. The design does not highlight critical information—such as calorie intake, blood sugar impact, and nutritional composition—lacks a clear visual hierarchy, and fails to meet the visual and cognitive requirements of elderly users.

• Medical Consultation
  The application lacks a rapid and intuitive emergency support function for elderly users who may face health crises, representing a significant oversight of their unique needs. The medical consultation content also includes extensive medical terminology, which can be challenging for elderly users to comprehend.

• User Services
  The interface lacks clear guidance and simplified step prompts, making it difficult for elderly users to independently complete device pairing and synchronization. The presentation of essential information like dates, medication names, and dosages does not use formats easily readable by elderly users, such as sufficiently large font sizes or adequate color contrast.

• Data Monitoring
  The presentation of charts and information in the data monitoring section is overly complex and not intuitive for elderly users. Chart font sizes do not accommodate their visual needs, color contrasts are insufficient to distinguish various data points, and there are no easy-to-understand legends, annotations, or dynamic aids, all of which affect elderly users’ ability to comprehend information.

• Exercise Management
  The application’s exercise recommendations do not sufficiently consider the physical conditions and capabilities of elderly users. There is a lack of features designed specifically for elderly users, such as demonstration videos of exercise movements or recovery advice post-exercise. Elderly users may encounter operational difficulties when interacting with data visualization interfaces, such as using sliders or clicking on small chart elements. These interaction designs do not adequately consider the reduced hand dexterity and diminished tactile sensitivity of elderly users.

3.2. Usability Testing Results

Usability testing involved 28 elderly diabetic patients aged 65 and above, identifying two primary barriers to usability: design and functionality.

3.2.1. Participant Feedback

Design

Design refers to the visual perception of the interface and information. The evaluation revealed that the prototype design significantly influences user engagement and interest, either facilitating or hindering user interaction. Design aspects include interface aesthetics, visual appeal, and the clarity of information, which is closely related to the structure and presentation of information. Excessive or complex text can reduce the processability of information, adversely affecting usability. Using technical terminology can also decrease the accessibility and comprehensibility of the interface. Participants in the usability testing emphasized the importance of readability, particularly advocating for large fonts and high-contrast color combinations. They recommended simplifying the interaction design to reduce complex operations, providing intuitive navigation and large buttons. Charts should be easy to understand and include clear legends and annotations. Participants also suggested enhancing interactive and educational resources, such as video tutorials and step-by-step instructions, to help elderly users better understand and perform health management tasks.

Examples of specific feedback on the design provided by elderly participants in the usability testing included:

• /…/ The font really needs to be bigger; sometimes, I have to find my glasses to see it clearly. Also, there should be a big contrast between the background and the text color, so it is easy to see at a glance.—Elderly Patient 8
• /…/ The interface felt too large; I had to scroll down to see other items, initially thinking there was a problem with my phone.—Elderly Patient 27
• /…/ I tried using this app to track my blood sugar, but the buttons are too small, and I keep missing them. I wish they could make it as simple and easy to use as a TV remote.—Elderly Patient 11
• /…/ The icons are not easily recognizable, and it’s easy to press the wrong one when there are other similar icons, which is also straining on the eyes.—Elderly Patient 27
• /…/ After recording my blood sugar, pressing the back button on my phone didn’t work, but the back button in the app did.—Elderly Patient 27
• /…/ I try to follow the health advice in the app, but some of the diagrams are too complex for me to understand. It would be helpful if they could simplify them or include an explanation of what they mean.—Elderly Patient 17
• /…/ Is there a way to turn these health tips into video tutorials? I might find it easier to learn by watching videos, because sometimes I don’t quite understand the text explanations.—Elderly Patient 24

Functionality

Functionality dictates how users accomplish tasks and process information through application interaction, aiming to enable efficient and effective task completion. The layout of functionality must consider the organization of information and the complexity of processes to prevent overly complicated steps that could hinder the user experience. Functionality includes assessing the rationality and appropriate arrangement of the application’s operational logic, which is closely interconnected. It also includes interaction efficiency, such as the ease of operation during task execution, transparent task workflows, and immediate feedback on actions. Based on participant feedback, here are four key usability issues in functionality that highlight areas requiring special attention during application design and development:

Logical Flow and Guided Navigation

Elderly patients emphasized the importance of logical flow among the application’s components. They highlighted that guided navigation is crucial because it enhances the application’s usability through intuitive step-by-step instructions and transition prompts. Logical flow should be coherent, ensuring users can seamlessly transition from one functional process to another, especially when managing complex tasks such as handling multiple medications or tracking various health data metrics.

• /…/ I need this program to tell me what to do next. Especially when managing so many medications, if it could guide me step-by-step, I wouldn’t feel so confused.—Elderly Patient 11
• /…/ I always have trouble finding the button to switch functions. It would be great if there were more obvious signs. When you’re older, you want everything to be simpler and more straightforward.—Elderly Patient 3
• /…/At first, I thought I knew what to do, and it seemed easy enough to use on my own, but when I got home, I still encountered parts that were unclear.”—Elderly Patient 23
• /…/ I would prefer a printed manual for clarity and ease of finding information.—Elderly Patient 13

Consistency in Operational Logic

Participants expressed a strong need for consistency in the application’s operational logic, particularly given the potential decline in cognitive abilities among elderly users. They noted that when the application’s user interface maintains consistent operational logic across all sections, it significantly reduces the learning burden for elderly users, allowing them to adapt to and master the application more quickly. Therefore, ensuring that every element and operational step in the application adheres to the same design principles and logical structure is crucial for enhancing overall usability for elderly users.

• /…/ I’ve noticed that the buttons and menus in each part of this program look similar, which makes it easier for me to remember how to use it. As I get older, my memory isn’t what it used to be, so having the same operations everywhere helps me learn quickly and avoid mistakes.—Elderly Patient 14
• /…/ I really like this app because no matter which part I’m in, the way to operate it feels the same. This helps me avoid mistakes and get things done faster.—Elderly Patient 27

Optimization of Error Handling Mechanisms

Elderly patients reported that the application’s support and feedback mechanisms are often insufficient when errors occur. An effective error-handling mechanism should include clear error messages and straightforward solutions that not only help elderly users quickly return to the correct operational path but also reduce the frustration caused by operational errors. Applications should employ error-prevention design strategies, such as disabling inappropriate options and validating the accuracy of data entries, to enhance usability and support for elderly users.

• /…/ Every time I do something wrong, this app should alert me and give me the right guidance instead of making me guess. I’m getting older and my memory isn’t great, so these little prompts really save me a lot of trouble.—Elderly Patient 11
• /…/ I’m not very good with electronic devices and often make mistakes. If this program could give me a prompt when I mess up, something like ‘That’s not right, do it this way instead,’ that would be a huge help.—Elderly Patient 16

Comprehensiveness of Interaction Feedback

Participant feedback underscores the critical importance of comprehensive and multidimensional interaction feedback for ensuring that elderly users correctly understand the outcomes of their actions. They stress the need for immediate and explicit feedback for successful operations and clear explanatory feedback for errors or inappropriate actions. Additionally, participants suggest that feedback mechanisms should encompass visual, auditory, and even tactile signals to accommodate potential visual and auditory impairments in elderly users, ensuring that each interaction receives an effective response and confirmation.

• /…/ I think whenever I enter information or complete a task, the app should have a very clear signal to tell me ‘success’ or ‘failure.’ This feedback could be a change in color or some simple sounds to help me understand what happened.—Elderly Patient 18
• /…/ I hope that whatever action I complete, I’m immediately told the result, whether it’s good or bad. It would be best to have both visual and auditory cues, because sometimes I can’t see the screen well, but I can hear sounds.—Elderly Patient 19

3.2.2. SUS Scoring

The study’s authors collected and analyzed data from 28 SUS participants, including the patient group, whose assessment results are shown in Appendix C, and the expert group, with results in Appendix B. An overview of the modified SUS rating table with inserted value ranges is shown in Figure 5. Detailed statistical analyses were conducted based on the SUS system usability scale evaluations for the patient group, presented in

Table 8. Statistical Analysis of SUS Scores for the Patient Group.

	Median (Min–Max)	Mean	Standard Deviation	t-Statistic	p-Value
Q1	4.0 (1–5)	3.5	1.4	1.89	0.07
Q2	1.0 (1–4)	1.64	0.99	−7.26	<0.01
Q3	4.0 (1–5)	3.79	1.2	3.47	<0.01
Q4	2.0 (1–4)	1.79	0.92	−7.01	<0.01
Q5	4.0 (1–5)	3.46	1.23	1.99	0.06
Q6	1.0 (1–4)	1.61	0.79	−9.38	<0.01
Q7	4.0 (1–5)	3.43	1.14	2	0.06
Q8	1.0 (1–4)	1.68	1.09	−6.41	<0.01
Q9	4.0 (1–5)	3.57	1.2	2.52	0.02
Q10	2.0 (1–4)	2.07	1.15	−4.26	<0.01
SUS	72.5 (65.0–85.0)	72.41	5.2	70.59	<0.01

, and for the expert group, shown in

Table 9. Statistical Analysis of SUS Scores for the Expert Group.

	Median (Min–Max)	Mean	Standard Deviation	t-Statistic	p-Value
Q1	3.5 (3.0–5.0)	3.67	0.82	2	0.1
Q2	2.0 (1.0–3.0)	1.83	0.75	−3.8	0.01
Q3	3.5 (3.0–4.0)	3.5	0.55	2.24	0.08
Q4	2.0 (2.0–3.0)	2.17	0.41	−5	<0.01
Q5	4.0 (3.0–5.0)	3.83	0.75	2.71	0.04
Q6	2.0 (1.0–3.0)	2	0.63	−3.87	0.01
Q7	3.0 (3.0–4.0)	3.33	0.52	1.58	0.17
Q8	2.0 (1.0–3.0)	1.83	0.75	−3.8	0.01
Q9	3.0 (3.0–4.0)	3.33	0.52	1.58	0.17
Q10	2.0 (1.0–2.0)	1.67	0.52	−6.32	<0.01
SUS	71.25 (65.0–75.0)	70.42	3.68	1.61	0.17

. An in-depth analysis of the participants’ evaluations of the system was performed by calculating the median (range), mean, standard deviation, t-statistic, and p-value. The median and mean values demonstrated the central tendency of the overall scores, whereas the standard deviation reflected the dispersion of the ratings. T-values and p-values were used to determine whether the differences between the scores and a neutral value (commonly set at 3) reached statistical significance. This comprehensive analysis assessed the system’s overall usability and highlighted consistencies and variations across user groups, assisting researchers in identifying strengths and areas needing improvement within the system. The findings provide a robust foundation for subsequent optimization efforts.

The analysis of data from both groups indicates that the overall usability of the system was rated as “good” by both the expert and patient groups. However, there are significant differences in their focus and in the stringency of their evaluations. Regarding the SUS total scores, the patient group had a median of 72.5 and a mean of 72.41, demonstrating broad acceptance among general users. In contrast, the expert group’s median was 71.25, with a mean of 70.42. Although these scores are also high, the difference from the neutral value of 68 was not statistically significant (p-value = 0.17), suggesting that experts were more cautious and stringent in their evaluations, reflecting higher expectations based on the unique needs of elderly users.

To better comprehend the significance of these scores, we consulted statistical data compiled by Veazie and colleagues, comparing it with findings from other diabetes management applications [54]. The ‘Diabetes Manager’ had a SUS score 68.5, slightly above the neutral value. Despite offering functionalities like insulin dosage suggestions and HbA1c calculations, its unclear privacy policy may undermine perceptions of trust and security in elderly users. The ‘Glucose Buddy Pro (GB+)’ had a SUS score 65.8. Although it provides similar glucose monitoring and diabetes educational features, its complex interface can confuse elderly users, impacting their user experience. The ‘BlueStar Diabetes’ had a notably higher SUS score of 85, integrating features such as dietary recommendations, glucose monitoring alerts, connections to Electronic Medical Records (EMR), and integration with wearable devices. These functionalities are particularly suitable for elderly diabetic patients who may require more daily support and seamless health data management. Both the ‘mDiab Lite’ and ‘mDiab’ applications had SUS scores of 47.5 and 48.3, respectively. The ‘mDiab Lite’ and ‘mDiab’ apps are more basic, lacking clear privacy policies and specific optimizations for elderly users, resulting in poorer experiences and lower scores among this demographic.

The assessments from both patient and expert groups demonstrate that the system is widely recognized for its usability among general users, particularly excelling in ease of use and user confidence. For instance, the scores for Q3 (System Usability) and Q9 (User Confidence) in the patient group were notably high, with means of 3.79 and 3.57, respectively, indicating positive user feedback on the system experience. Although the expert group rated these aspects highly, the slight differences suggest that the system performs well across diverse user groups. However, the expert group demanded higher standards in system complexity, consistency, and technical support, scoring lower on negative aspects such as Q2 (System Complexity) and Q4 (Technical Support and Needs). This highlights their heightened sensitivity to the system’s shortcomings in a professional setting. Despite this, the expert group rated the functionality integration highly (Q5 mean of 3.83), acknowledging the system’s effective feature integration. Overall, while the system exhibits good usability among general users, further enhancements in complexity and consistency are required to elevate evaluations from professional users.

Similarity between Groups

To analyze the similarity between the feature sets provided by the older adult patient group and the expert group [55], we employed three key statistical techniques: Cosine Similarity, Jaccard Similarity Coefficient, and Simple Matching Coefficient (SMC). Cosine Similarity measures the similarity between two sets in a vector space by evaluating their inner products. The Jaccard Similarity Coefficient compares the intersection ratio to the union of binary sample sets, while the Simple Matching Coefficient further accounts for the similarity of zero and one values within the sets. Results from these methods indicated good similarity, with indices exceeding 0.5, suggesting a substantial agreement between the groups on various aspects of the system, as detailed in

Table 10. Similarity indexes.

Metric	Cosine Similarity	Jaccard Index	Simple Matching Coefficient (SMC)
Average Value	0.81	0.56	0.67

.

The patient and expert groups demonstrated high concordance in their evaluations, with a Cosine Similarity of 0.81, indicating overall solid alignment in their scoring trends. This suggests that both groups reached a consensus on multiple aspects of the system evaluation. However, the Jaccard Similarity Coefficient at 0.56 and the SMC at 0.67 reflect some divergence on specific issues, particularly regarding system consistency and complexity. This divergence may arise because the expert group focuses on filtering and refining system functionalities rather than on proposing entirely new ideas. The primary role of the expert group may involve filtering and improving suggestions made by the patient group, thus ensuring the system’s overall quality and professionalism.

Despite these differences, the high similarity scores among the patient group indicate their strong positive reception of the system. Compared to the expert group, the patient group emphasizes the practical usability experience, while the expert group focuses more on detail and consistency in evaluating system functionalities rather than on broad innovations. Although there is considerable agreement in multiple dimensions, the expert group exhibits higher professionalism and stringency in system evaluations, implying that further system enhancements should more critically consider the needs and standards of professional users.

4. Discussion

As discussed in the literature, the increasing global aging population and widespread use of mobile devices present new challenges for developing mobile health applications, particularly those designed to optimize usability for senior citizens in diabetes management applications [56]. This study investigates the mediating factors in the “Sugar Shift” application tailored for elderly diabetes patients, aiming to explore and integrate key design factors that enhance usability. Previous studies recommend using diverse usability methods to provide developers with insights into potential usability issues [57].

The heuristic evaluation identified significant usability issues and catastrophic ratings associated with ineffective information display and system feedback, indicating the need to mediate interventions aligning with users’ cognitive and logical habits. This alignment aids in making effective decisions and controlling actions for tasks relevant to patient care. Usability testing results demonstrate that design features such as simplicity and clarity, high-quality content, and interactive functionality that emphasizes ease and logical flow can enhance usability. Previous research also highlights the importance of intuitive usability, especially when users need more technological proficiency or when training is not feasible [58]. In designing mediating factors for future applications, carefully considering the anticipated volume of data input is crucial. Research findings suggest that dissatisfaction with system adaptability and information overload are among the most common complaints from elderly users, potentially leading to complete application abandonment [59].

User engagement is closely related to usability, encompassing how users interact with technology and their emotional responses [60]. For instance, participants noted that small numbers on the timeline of blood glucose records discouraged continued reading. Previous studies have emphasized that interface layouts should be straightforward, intuitive, and logical, featuring clearly labeled, direct access to standard functions, large fonts, and high-contrast colors to enhance the readability of information. This approach ensures that health information is visual and comprehensible [61,62,63,64]. Eye diseases are among the primary complications of diabetes [65]. Therefore, in addition to incorporating features in mobile apps that promote active management of exercise, diet, and medication to control blood sugar and blood pressure strictly [65], it is recommended that mobile applications use large icons and adjustable font and icon sizes to ensure users can easily read historical data, trend charts, or medication reminders. Therefore, mobile applications must offer flexibility, providing elderly users with personalized options based on their preferences.

Moreover, excessive data entry steps can increase confusion among elderly app users, whereas simplifying interaction processes can help reduce operational difficulties, enhancing the likelihood of successful use [64,66]. Ho et al. noted that elderly individuals are less concerned with the logic and principles behind interactions but are more focused on achieving goals quickly. Interaction processes need to be easy to manage and should enhance error tolerance [67]. Specifically, interaction processes simplify everyday tasks such as entering blood glucose levels or logging dietary intake by breaking them into straightforward steps, minimizing the need for precise mouse movements, reducing distractions, and guiding users with wizard-driven interactions that provide clear feedback for each action.

Effectiveness and satisfaction are critical metrics of usability [68]. Despite the use of visual charts, participants still reported issues such as “incomprehensible information” and “overly complex data”. The evaluation revealed that even with simplified graphical presentations, users still made execution errors. Previous research has also highlighted that familiarity with contemporary products and successful interaction decreases with age and prior experience [69]. Thus, we concur with earlier researchers that mobile health applications for senior citizens should consider the physical, emotional, and social aspects of human–computer interaction in order to adjust other design attributes [70]. Such applications could incorporate memorability features based on Nielsen’s usability model. According to Muqtadiroh et al., memorability attributes can aid elderly users in recalling application options, functionalities, and operational steps, especially when revisiting the application after a period [71]. Specifically, developers can present complex medical information in formats familiar to elderly users, such as images or illustrations. This approach enhances the efficiency of information transmission and user comprehension, particularly when explaining complex medical directives or health advice. User acceptance, consistently reported in the literature as a challenge, correlates with better technology acceptance when multiple modes of communication are offered, including auditory cues, video instructions, and tactile feedback [72]. Based on individual health data and habits of elderly users, applications can deliver personalized health reminders, such as medication, blood glucose testing, and appointment reminders, helping patients to maintain regular management routines and prevent episodes of hypoglycemia and thus enhancing self-management outcomes for diabetes.

Given the complexity of diabetes management, technology plays a critical role in facilitating patient self-management and education. Future designs should integrate the latest technologies to provide real-time feedback on patient lifestyle data and seamless integration into mobile applications. This approach not only alleviates the management burden on patients, but also offers healthcare teams the following new options for remote monitoring and reviewing health data:

• Digital Blood Glucose Monitoring: CGM systems, with subcutaneous sensors, enable precise real-time blood glucose monitoring, providing instant feedback and hypo/hyperglycemia alerts via apps. IGMS enhances healthcare efficiency by syncing blood glucose data with hospital systems [73].
• Smart Insulin Delivery Systems: Closed-loop insulin pumps and the InPen system automate insulin dosing by integrating CGM data and connecting to mobile apps via Bluetooth, streamlining glucose management [74].
• Digital Dietary Management: Apps and online platforms improve adherence to dietary guidelines by optimizing personalized meal plans and enabling direct caloric measurement [75].
• Digital Exercise Intervention: Smartphones and wearables provide personalized exercise plans, monitor outcomes, and use gamification to increase patient engagement and adherence [76].
• Digital Health Education: Digital platforms like telehealth, video, social media, and dedicated apps deliver flexible and accessible diabetes education, supported by remote monitoring for continuous health management [74].

This study also underscores the significance of dual-domain experts in the evaluation process. These experts are adept at identifying technical usability issues and possess insights into the unique needs associated with chronic disease self-management [77,78,79]. For instance, in the medication reminder feature, UI/UX design experts consider multimodal reminder methods and enhance the attractiveness and readability of crucial information through visual design. Additionally, these experts assess considerations for polypharmacy in elderly patients and potential drug interactions. Experts recommend that future chronic disease management systems continue to employ dual-domain experts to leverage their expertise in enhancing user experience design. Through their comprehensive understanding of technical usability and disease management needs, these experts can propose innovative solutions, ensuring that applications are technologically advanced and truly meet patients’ practical needs. Their collaboration can aid in designing more intuitive, easy-to-use, and fully functional applications, enhancing patients’ self-management capabilities and ultimately improving health outcomes. By implementing standardized evaluation processes and detailed assessments tailored to specific scenarios and tasks, this interdisciplinary collaboration model increases the innovation and efficacy of solutions and ensures the repeatability and universality of evaluation results [47,48,80].

4.1. Future Research Directions

In the “SugarShift” project, we conducted expert-based usability testing to address a critical question: “How should an application be designed to support effective self-management for diabetes patients aged 60 and above?”. This study combined a comprehensive literature review, market research, heuristic evaluation, and System Usability Scale (SUS) analysis to examine current application usage, identify factors that enhance or inhibit user acceptance, and outline the specific design features necessary for developing diabetes management applications.

By the end of this year, we plan to develop a senior mode for “SugarShift”, explicitly tailored for users over 60 years old. This prototype application will be user-centered, considering the cognitive characteristics of the older adult—including alteration of color schemes, font choices, and methods of information feedback—to ensure a coherent interface and reduce the visual and reading burden on patients. To ensure the professionalism and practicality of this senior version, we will collaborate with experts from interdisciplinary fields—including geriatric chronic disease care specialists, nutritionists, and psychologists—forming an online advisory team. Additionally, we will involve computer science and artificial intelligence experts to ensure technological advancement and feasibility.

Potential users and usability experts will be involved from the early stages of product development to guarantee high usability and a user-driven approach. We will regularly conduct usability testing with users and experts, continuously integrating the results into the application’s optimization process until the project’s completion. This iterative development model ensures that the product evolves to meet changing user needs while maintaining the latest technological advancements and high efficiency.

4.2. Limitations

A major constraint of this research is the limited number of participants and the possibly non-representative sample selection, necessitating careful interpretation of the results. Additionally, although this study incorporated heuristic evaluation and usability testing to enhance understanding of potential usability issues, these experiments were conducted in a controlled laboratory setting rather than in natural environments, such as the participants’ homes, which could have alleviated stress and optimized engagement. This difference in setting may affect the validity of the data, as it was impossible to control the testing environment strictly. While this flexible approach facilitates increased participation and comfort, we may need more support to identify and address usability issues effectively.

Experts recognize that heuristic evaluation reveals the most significant usability issues in specific contexts, providing valuable directions for design improvements. However, adding more dual-domain evaluators could uncover more nuanced issues, thus providing broader insights for developing mobile health applications. Although we strive to identify and resolve usability issues through these assessments to improve diabetes patients’ self-management, the cost and complexity of these testing methods are also significant factors to consider in implementation. Future research should explore more cost-effective evaluation methods to facilitate broader application across diverse patient populations and various mobile health applications.

5. Conclusions

In this study, we employed a method that combines heuristic evaluations by dual-domain experts with subsequent usability testing to thoroughly explore the experiences of elderly diabetic patients using a management application. Implementing this approach enhanced our ability to identify and address real usability issues, ensuring our findings’ depth and practicality. Experts precisely assessed whether the data input interface of the application suited the cognitive and operational habits of elderly users and determined whether the backend data processing meets clinical requirements. Heuristic evaluation provided unique insights related to prominent tasks in patient self-care, identifying potential issues that could affect elderly users and how the application adheres to Nielsen’s usability guidelines. Usability testing validated the heuristic evaluation results and revealed some nuances that might not be apparent in a laboratory by observing actual behavioral feedback from elderly users. Overall, through heuristic evaluations by dual-domain experts and usability tests with actual users, this study not only enhanced its scientific and systematic nature but also ensured that the research outcomes genuinely improve the quality of life for the target user group in a user-centered manner. This research approach provides an essential reference framework for developing medical applications, emphasizing the importance of integrating interdisciplinary expertise in a user-centered design process.