The self-management of gestational diabetes (GDM) usually includes a series of complex lifestyle behavioral changes such as diet, exercise, and continuous self-testing of blood glucose [1]. This systematic approach aims to help women maintain optimal blood glucose levels and minimize complications for both mother and baby. However, previous research highlights the challenges for women with GDM, who find it hard to understand or follow the instructions from doctors [2]. Women have to make decisions on food choices daily, and this is done through effective cooking strategies [3]. It also challenges women to understand the labels and nutritional values of food [4]. Guidelines across countries emphasize the importance of strengthening physical activities and regard aerobic activities as acceptable approaches [5]. However, this requires carefully choosing the level of intensity and the types of exercise [5].

GDM digital technologies have covered various areas of GDM management, providing features including aspects of educational information, health behavior coaching, data recording, and communication interfaces for women [6]. By analyzing the overall effectiveness, previous literature found that GDM digital technologies could provide comparable quality of care to face-to-face visits with health care professionals or midwives, manifesting in heightened maternal health conditions and increased rates of natural childbirth, alongside diminished occurrences of adverse maternal and neonatal outcomes [7-9].

There are various forms of digital technologies, such as mobile apps, webpages, and digital devices, among which mobile apps are one of the most common types of technology used [9]. The rise in global smartphone penetration has led to the evolution of mobile apps, which are now one of the most representative forms of GDM digital technologies [8-10]. Mobile phones possess advantages in terms of accessibility as they are portable devices, thereby facilitating women’s access to digital services irrespective of their geographical locations [11,12].

However, unlike the overall optimistic results from the evaluation of effectiveness, our previous review and other studies found a variation in pregnant women’s views on the app’s usability [13,14]. Women had positive views of apps when they supported self-monitoring and viewed them as beneficial tools for obtaining information about GDM, making informed dietary decisions, and facilitating exercise regimens [13,15-17]. The technologies also led to increased levels of adherence, concentration, and satisfaction among women [7,14]. The apps offered platforms for the women and clinicians to keep continuous interaction, fostering the client-clinician relationship [16,18]. This mitigated various concerns that women encounter in managing their condition, demonstrating efficacy in facilitating women’s self-management practices [7-9]. In the meantime, apps added to women’s emotional burden with problematic features and content [13]. For example, redundant processes and frequent failures in data recording reduced women’s motivation for using the technology [18,19]. The differences between the information provided by apps and clinicians confused the women [16]. Evidence suggests that mobile health (mHealth) apps might unintentionally lead to increased unhealthy behavior [20]. mHealth apps designed inappropriately may contain inaccurate or incomplete information that misleads users or provide insufficient support for managing high-risk behaviors, which could affect users’ safety [21].

To ensure that women are offered apps of the best quality, health care practices face challenges in evaluating and selecting suitable options. The current era is facing the results of unregulated mHealth markets, which flourish in the number of emerging mHealth apps but vary in quality [22]. The increasing study of app quality evaluation has triggered a discussion on the definition of app quality. Studies found that the criteria used for evaluating app quality were heterogeneous [23,24]. In addition, established criteria lack precise definitions, which hinders the assessment of app quality.

Two previous reviews have evaluated the functionality and usability of GDM mHealth apps [6,25]. Both reviews primarily relied on a single quality criterion: the Mobile App Rating Scale (MARS). However, neither study addressed how well app features are adapted to GDM self-management content. Similarly, they offered no practical insights to assist health care professionals in selecting apps that are both functional and user-friendly. What is missing from the current literature is an evaluation of how these apps contain features suitable for GDM care, and guidance that helps health care professionals when selecting usable apps.

This study reviewed commercially available mobile apps to explore app quality and multifunctionality, targeted at antenatal care for women with GDM or pre-existing diabetes during pregnancy, available in the UK app market, and in the English language. This systematic evaluation only includes free-download apps and analyzes the available free-of-charge features, providing insights into the types of features that are most accessible for women with GDM. We assessed the quality of the apps using standardized evaluation scales. We evaluated the app’s multifunctionality based on app features and content tailored specifically to the GDM population, as well as general features that affect the normal use of apps.

This study followed the TECH (target user, evaluation focus, connectedness, and health domain) framework proposed by Gasteiger et al [26] (2023), which is tailored for the context of reviewing and evaluating the quality of commercially available mHealth apps as opposed to the systematic literature review methodology. Informed by the traditional systematic review process, the TECH framework aims to help conduct reviews on mobile apps by establishing search strategies, the use of eligibility criteria, and a comprehensive quality appraisal. It provides guidance to conduct a search, and has been applied in recent studies to evaluate mHealth apps, including risk assessment and self-management tools, by examining both their functionality and content [27-29].

This study consists of 2 major domains of evaluation: app quality and app multifunctionality. Figure 1 shows the structure of the evaluation elements and the evidence supporting the evaluation. The app quality evaluation in this study aimed to explore the type of features being used in the included apps and used generalized tools to quantify the level of quality. The measurement was informed by the TECH framework and guidance by introducing both the MARS and the IMS Institute for Healthcare Informatics Functionality Score (IMS). The aggregation of the 2 quality appraisal tools helps to evaluate mobile apps more comprehensively.

The multifunctionality evaluation aimed to conduct a GDM context-tailored assessment on the behavior change techniques (BCT) used in the apps, as well as other essential features needed to be enabled for normal app usage. Therefore, there were two elements in the multifunctionality evaluation: (1) the GDM-adapted features and techniques, which refer to the app features aligned with BCTs, along with the digital techniques used to enable the features; and (2) the general features, which refers to the essential features that enable the normal usage of apps, including instructions, data security, customization, and technical issues.

Following the guidance of the TECH framework, a scoping search on Apple App Store and Google Play in the UK context was run in March 2024 to identify eligibility criteria [26]. This study aims to identify apps that include GDM as one of the targeted user groups and that eligible apps must refer to GDM in their description to qualify for inclusion in this study. From this process, three potential types of mobile apps were identified: (1) mobile apps for GDM management, (2) mobile apps for pregnancy management with additional features for blood glucose control, and (3) mobile apps tailored for diabetes patients but claimed to be suitable for women with GDM. Additionally, the mobile apps were available in the UK context and the English language, free to download, running on Android (Google LLC) or iOS (Apple Inc). The eligibility criteria are listed as follows (Multimedia Appendix 1):

The search keywords "gestational diabetes," "GDM," and "pregnancy diabetes" were applied in the search strategy. The final search was conducted by 3 reviewers on the Apple App Store and Google Play in June 2024. Due to the Apple App Store’s online page design, it was not possible to export the search results into screening software to remove duplicates. Additionally, the mobile app search results were subject to frequent changes over time. To avoid unnecessary duplication in the search, all 3 reviewers searched and performed screening on the same day.

The screening process was guided by the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) checklist and the TECH methodology [26,30]. Two steps of screening were applied, including title screening and full mobile app screening. The mobile app name and description screening process was conducted on the search results pages to identify potential mobile apps for inclusion. We contacted the app companies to access those apps with limited access during the retrieval process. A subsequent full mobile app screening was conducted to check the relevance and feature operability. To include the mobile apps which were usable at the technological level as well as suitable for the targeted population, the full mobile app screening process made a decision of inclusion when (1) the screened mobile app was available in English; (2) no constant technical glitch or crash occurred when operating; and/or (3) the mobile app mentioned GDM at least once in options, information, or features. A snowballing strategy was used after the full mobile app screening process to check the relevant-recommended apps generated within the search results pages of the final-included mobile apps, such as the “you might also like” and “similar apps” pages, to reduce missing mobile apps.

During the screening process, 2 pairs of authors independently screened the search results in Google Play and Apple App Store. The level of agreement between the reviewers was determined using Cohen κ to calculate interrater reliability. To minimize the potential for duplication inherent in app store searches, the analysis was based on the results from a single search term. Discrepancies between the paired authors were discussed and resolved. The results of eligible mobile apps were checked and agreed upon by the other authors. The devices used to search and screen the apps were iPhone 11 Pro Max (iOS 17.5), iPhone 11 (iOS 17.5), iPhone X (iOS 16.7.8), Samsung Galaxy J5 (Android v8.1 Oreo), and Huawei Mate 10 (HarmonyOS 4.0).

The extracted features and techniques of apps were analyzed and presented in aggregated data. Categorical data, including app characteristics items, GDM-adapted features, GDM-adapted techniques, and general features, were summarized as frequencies and percentages. Continuous data, including app size, MARS score, and IMS score, were presented as means with SDs. App characteristics, MARS score, GDM-adapted features and techniques, and general features were summarized in tables to provide a clear presentation of results. The IMS score was presented in a radar chart to illustrate the distribution of IMS functionalities across 10 domains.

This study did not involve human participants, animals, or any collection of personally identifiable information. All data analyzed were publicly available in the Apple App Store and Google Play Store. Accordingly, institutional review board approval was not required. As no human subjects were enrolled, informed consent was not applicable.

The search identified 1066 apps, 330 from the Apple App Store and 736 from Google Play. Two paired authors independently screened the search results in both app stores, the Apple App Store, resulting in 28 eligible mobile apps from the Apple App Store, and 21 in Google Play. The interrater reliability in both pairs indicates a minimum of substantial agreement, with Cohen κ values of 0.62 and 0.91. The title and description screening process resulted in 49 potentially eligible mobile apps in the two app stores. The mobile apps available in both operating systems were not combined at this step.

The 49 eligible mobile apps were retrieved and installed before the full mobile app screening process. This process led to 10 exclusions due to activation (n=7), subscription (n=1), and technical crash issues (n=2). The retrieving, installing, and checking process led to 39 (n=21 in iOS, n=18 in Android) inclusions for the full mobile app screening. The full app screening led to 10 exclusions (n=5 in iOS, n=5 in Android). The reasons for exclusion included irrelevance (n=9) and language (n=1). Perfect agreement was identified between the 2 pairs of authors. Mobile apps available in both operating systems (n=8) were then combined in number, resulting in 21 eligible mobile apps.

A subsequent snowballing process generated the inclusion of 2 mobile apps (available in both operating systems). The search and screening process contributed to the eventual inclusion of 23 eligible mobile apps for this study. Figure 2 illustrates the PRISMA diagram for the screening process (Multimedia Appendix 3 for the enlarged PRISMA diagram).

Of the 23 reviewed apps, 56% (n=13) were adapted to both operating systems (iOS and Android), followed by 22% (n=5) of the apps which were exclusively for iOS, and 22% (n=5) of the apps which were exclusively for Android. Most (n=16, 70%) of the apps were updated in 2024. On average, the apps were of mean 66.43 and SD 53.46 MB, ranging between 6.2 and 223.4 MB. All the apps were free to download, with 48% (n=11) free to access all features, and 52% (n=12) required a subscription. The major targeted user groups of the apps were people with diabetes (n=12, 52%), women with GDM (n=7, 30%), followed by women who are pregnant (n=3, 13%) and women with diabetes who are pregnant (n=1, 4%).

We categorized the apps into 4 major types based on the characteristics of features: multipurpose, self-monitoring, educational, and communicational. Self-monitoring apps were identified by their ability to assist women’s behavior for self-monitoring, such as recording data and reminding them of self-tests, with a total of 5 apps. Educational apps were identified by their ability to offer women information, with a total of 5 apps. Communicational apps were identified by their ability to provide a platform for women to communicate with their peers and health care professionals, with a total of 2 apps. The apps that used more than one of the mentioned characteristics of features were identified as multipurpose apps (n=11). All apps (n=23) were free to download, with all apps that were tailored for GDM or pregnant users (n=11) being free to access all features; meanwhile, the other 12 apps for general diabetes users needed to subscribe for full access. Table 1 shows the characteristics of the reviewed apps.

Overall, most reviewed apps used general features which ensured normal app usage, with 23 apps addressing data security, 22 apps addressing instructions, and 19 apps addressing personalization. There were 5 apps identified as having technical issues that affect normal use, or distractions that disrupt user interactions. General features and technical issues are summarized in Multimedia Appendix 6.

Some general features were applied universally by the apps. These include consenting to privacy policies (n=23) and asking permission to access data (n=23), providing a quick step-by-step introduction (n=16) when users launch the apps. For personalization, collecting users’ demographic data (n=16), asking their preferences for measurement units (n=14), and enabling customized features (n=16), such as reminder timing, were used. Very few apps mentioned the location of data storage (n=5) and the techniques (n=4) used to protect user data.

Additionally, of the 23 evaluated apps, there were 7 with issues that could affect normal user experience, including 5 multipurpose apps (MyManis, hug + u, myFetalLife, Blood Sugar - Diabetes App, and Diabetes:M) and 2 self-monitoring apps (Carbs & Cals and Diabetic diary-Glucose tracker). There were three major issues among these apps: (1) frequent technical glitches, (2) data management issues, and (3) frequent advertisements. Data management issues existed in 4 multipurpose apps, with frequent failure or errors of data recording (MyManis and Diabetes:M), limitations in the data recording feature when inputting decimals and time points (hug + u), and failure of displaying the recorded data in 1 app (myFetalLife). Frequent advertisements took place in 3 apps (Blood Sugar - Diabetes App, Carbs & Cals, and Diabetic diary-Glucose tracker), with advertisements showing up frequently when moving between screens or inputting data, leading to a sense of distraction. Technical glitches and crashes happened to 1 multipurpose app (myFetalLife). While this app contained various features suitable for the management of GDM, technical issues took place frequently and caused crashes and failures when recording the data.

This review evaluated the app functionality based on the generated lists of GDM-adapted features and techniques, as well as general functionality, collectively with 45 evaluation items on features and content. On average, the apps applied a mean of 17.95 and SD of 7.31 of all 45 items across domains. For GDM-adapted features and techniques, the apps applied a mean of 10.82 and SD of 5.31 of the 32 items. For general functionality evaluation, the apps applied a mean of 7.13 and SD of 2.58 of the 13 items.

We considered apps usable when they contained features and content higher than the average value in each domain and the total average value, with no distractions or technical issues (Table 4). These 6 apps were either categorized in the multipurpose (n=4) or the self-monitoring (n=2) app type in this review.

This study aimed to provide a comprehensive evaluation of the quality and multifunctionality of commercially available apps on their cash-free features for GDM self-management. By evaluating app quality using the MARS quality and IMS scales, the results showed that the overall app quality is at a satisfactory level. However, regarding individual MARS domains, the MARS engagement domain, which contains coaching and personalized content, was scored lowest among other domains. Following this, the multifunctionality of apps was evaluated by two elements: (1) GDM-adapted features and techniques, namely the app features that fitted into GDM management context, and the corresponding digital techniques used to enable the features, explored the variety of features used by the reviewed apps; (2) general features, which included instruction, data security, customization, and technical issues, explored the app features regarded as essential for basic app usage.

The reviewed apps predominantly focused on GDM educational content and blood glucose data recording. In addition, the digital techniques used to enable such features were relatively basic. Most apps relied on text-based approaches for content presentation and required manual operation for data management. Moreover, their content reflected only the most fundamental requirement for GDM, focusing primarily on blood glucose management rather than addressing the broader pregnancy experience.

One of our primary findings is the narrow focus on blood glucose monitoring in the design of many apps. Many of the apps included in this review offered features that could support blood glucose self-management, such as educational resources on self-test skill education and dietary guidance, as well as data recording, digital feedback, and graphic visualization. These features were also identified in our previous study as being preferred by women with GDM, as they facilitated self-management [13]. For example, having educational information in a single app improved easy access and helped women recall information [13]. Initial digital feedback using colored labels, combined with graphic visualization, helped the women interpret their data and adjust their diet according to blood glucose trends [13]. However, most apps lacked a multidimensional approach to GDM management. Even though diet and physical activity are both essential for stabilizing blood glucose levels, very few apps in this review included relevant content and features.

Our findings suggest that features relevant to pregnancy progression management were lacking in available apps. In this study, only a small number of apps provided information or features to support pregnancy management and fetal monitoring. There were only 5 apps that included features or information relevant to pregnancy, such as informing about baby growth over time. The features and information about pregnancy and fetal well-being were more likely to be available in those apps designed for pregnancy care. However, the functionality of these pregnancy-tailored apps in blood sugar control was relatively underdeveloped, typically due to technical issues related to data recording and display. These findings indicated an imbalance in focus in app design, with limited integration of broader pregnancy-relevant needs.

In addition, automated features, including automated data transmission and app-generated behavior change suggestions, were used by only a few apps. Similar patterns are also observed in other mHealth apps for pregnancy care, where basic features and techniques still dominate the feature composition [39-41]. These apps tended to focus on either collecting users’ personal and physiological data or providing normative and comprehensive information, rather than proactively facilitating women’s behavior changes [39-41]. Automated features, such as rewarding on milestone achievements, identifying barriers for behavior change, or providing feedback for behavioral adjustment, are rarely incorporated into the design of pregnancy-related apps [39,40]. Instead, these tasks are sometimes managed through interactions with health care professionals and handled manually [41].

This indicates that the apps currently available and free to download appear to lag behind the evolving needs of women and health care professionals. Women seek tailored guidance for managing their blood glucose, which enhances women’s self-awareness and supports their autonomy in self-management [13]. This is usually achieved through effective communication with health care professionals, where the women can receive one-to-one advice and personalized feedback on their blood glucose levels, diet, and lifestyle adjustment [42,43]. However, limited opportunities for antenatal consultations and the heavy workload of health care professionals have often resulted in limited advice on continuous behavior changes [42,43].

Our findings highlight a gap between the potential of technology and its real-world application. Frontier automated techniques are advancing at an astonishing pace, such as deep learning, decision trees, and reinforcement learning, which have been explored for predicting blood glucose levels and providing personalized treatment recommendations [44,45]. These tools can continuously process data such as blood glucose levels, ketonuria, and dietary adherence, identifying patterns and making real-time adjustments [44]. However, our review revealed that many frontier technologies remain largely inaccessible to women with GDM.

To address this gap, it is essential to integrate broader pregnancy progression management into a condition-specific management app that requires thoughtful design to ensure relevance across different stages of pregnancy and to balance general maternal health management with condition-focused support. As recommended by the 2015 NICE guidance, pregnancy progression management is an important component of GDM management [37]. This requires additional attention on fetal growth and well-being, indicating that additional screening test results and fetal self-observation data recording should be considered [37]. Additional evidence found that women with GDM preferred mHealth technologies that were closely aligned with their pregnancy needs, such as reminders for antenatal visits and pregnancy stages, and they particularly valued educational information related to their pregnancies [16,46].

The findings emphasize the potential of digital tools to provide support for individualized care. It calls for greater integration of automated technologies into app design. This requires co-design approaches involving women and health care professionals to ensure that future apps move from basic features to personalized, interactive, and context-based support for GDM management. Additionally, further exploration is needed to evaluate the automated app feature designs regarding their usability, effectiveness, and safety in real-world contexts.

Given the wide variation in features and techniques available in current GDM apps, and the range of factors that may influence app development, we adopted an evidence-based approach to summarize broadly applicable recommendations for the development of apps for women with GDM (Textbox 1). These recommendations outline options for core functional design, while also identifying optional advanced features that may be incorporated where appropriate.

The strengths of this review include its standardized process and comprehensiveness of evaluation. Standardized evaluation tools were used to measure the quality and multifunctionality of mHealth apps for GDM. In addition, it extends the TECH framework by Gasteiger et al [26] by introducing a more context-based, user-informed, and evidence-informed approach, integrating BCTs, NICE recommendations, and women’s preferences, enabling a detailed evaluation of app multifunctionality.

However, several limitations should be considered when interpreting the findings. First, the search was conducted only once on a single day. While this minimizes duplication of included apps, it may have resulted in an incomplete capture of all available apps due to the dynamic nature of app store listings, which can be influenced by advertising and other factors. Second, the decision to restrict the search to UK app stores, while ensuring consistency with UK-based NICE guidelines, may have limited the scope and breadth of the review by excluding insights from apps available in other regions. Third, our findings reflect the status of apps in June 2024. Given the high turnover and frequent updates of apps, their scores, features, and technical issues may have changed postreview, which may affect reproducibility.

This study explored the quality and multifunctionality of commercially available apps for GDM self-management, focusing on the cash-free features, following the TECH and PRISMA frameworks. Our findings highlight that GDM apps should provide more interactive features and support comprehensive pregnancy care, rather than being limited to glucose management and using basic digital techniques. Valuable insights can be drawn from other models of remote pregnancy care, especially those implemented under unavoidable circumstances, such as in rural settings or during the COVID-19 pandemic, where integration of data was required to collect from remote monitoring approaches [47,48]. On one hand, when designing digital technologies for a specific pregnancy complication, it is important to integrate general pregnancy-related data, such as self-reported symptoms, so that they can be managed alongside the target condition [47]. On the other hand, efforts should be made to enable the in-person services and advice to be available online, ensuring that women receive adequate support while engaging in self-management [48]. Taken together, the evidence highlights an emerging research and design agenda for future mHealth app design—one that prioritizes the integration of comprehensive pregnancy progression management and data interconnectivity to better meet the complex requirements of GDM management.