Of the included studies, 73% (19/26) contributed qualitative data to the thematic analysis. The views of women and HPs were integrated and reported together throughout the analysis. A total of 4 themes were identified: benefits of using technology, engagement with people via technology, usability of technology, and discouragement factors for the use of technology (definitions of the themes and subthemes are available in Multimedia Appendix 4). Furthermore, 2 subthemes were identified, as outlined in Figure 4.

This theme included 2 main components including engagement with peers and engagement with health care professionals.

Women with GDM indicated that accessing “peer support” by a digital GDM self-management system would be useful [75] as “somebody may know something more” [68]. Although some women had little or no experience with web-based group communication, they were still interested in communicating with other women with GDM via technology [68]. Peer support provided an opportunity for women to access “other people’s experiences” [73] for sharing and exchanging information [68]. As a woman with GDM indicated, peer support provided “a better overview of risks associated with GDM, what could go wrong potentially, and the good stories as well” [73]. Overall, women perceived that peer support empowered them with a broader scope of GDM knowledge than other women who were experiencing the same condition [73]. In addition, peer support reassured women that they were “not alone” [73] and offered them a “constant feeling of support” [73]. Furthermore, it enabled women to talk about their condition and experience in a “safe space” without being judged by other people [73]. Women indicated a lack of peer support in the current care system that might be addressed using technology [75].

Regarding engagement with health care professionals, women appreciated the possibility of receiving additional support using technology. They valued sharing their data and having regular GDM monitoring by HPs via technology [25,66,68,72,73,76], specifically for benefiting both their own health and that of their baby [72]. Women expressed how sharing data with HPs was “reassuring” and gave them a “safety net” [25] owing to a feeling of being monitored more closely by the HPs [73]. Similarly, some HPs believed that sharing data would provide an opportunity to review and monitor the data frequently [64,66], detect any changes or problems at an early stage [25,48] and thereby allow the early application of treatment or interventions for women with GDM [25,64,66].

Although some women and HPs felt comfortable communicating via technology [30,66,71,73,75], others were concerned about a lack of physical and emotional support [27,62,66,75] and a poorer quality of conversation [25,27,66]:

Nevertheless, women still felt there was a need to provide more interaction and communication between HPs and women via mobile app technology [76].

Women and HPs provided various perspectives on the usability of digital GDM systems in this theme. The content of the systems, including the quality of information and format and presentation of patients’ data, was the main usability aspect discussed by women and HPs in the included papers.

When women and HPs found the GDM systems “easy to use” [25,30,66,72,74], “simple” [25,66], “intuitive” [66] and “straightforward” [25,74], these impressions were influenced by the presence of simple language and images [63] and the simplicity of information presentation, such as displaying all data on one screen [62,72,77].

When usability concerns arose, they were also related to the data format and layout. Women and HPs suggested improving the layout and format of the information by changing the size of images or the amount of text [49], using videos [27], improving the data summary presentation [74,77], changing the data format to be similar to that of a paper logbook [25,48,63,74], and distinguishing different degrees of BG severity [48]:

In the study by Pustozerov and Popova [76], HPs also indicated that improving the data format would help them review the data more easily.

Discussions on usability were also directed at the effectiveness of GDM apps in fulfilling the needs of women. Participants in different studies provided opinions about the lack of functionality in their GDM self-management systems. Some of their diverse suggestions included an option to scan barcodes of food [74], a time-alerting function for entering data [13], an educational or coaching feature [48], the ability to add a note to BG readings, and the ability to record the type of physical activity they have performed [65].

Women were also interested in having pop-up messages [65], informing them about any changes in their data [48], their condition [73], or any new activities in the forum [68] on the system:

A final aspect of usability concerned the effectiveness of information content. Women and HPs found the GDM information in both older technology [73] as well as that introduced by the studies [63,65,77] to be insufficient and simplistic [63,65,73,77]. Personalized information was considered vital [63] for diet [48,49] and in-depth information about GDM [65,77]. In addition, some women had issues with the clarity of the content and wanted simple, clear [49,77] and commonly used language [63,77] such as using “tablespoon” or “bowl” as familiar measurement units used by their dietitian rather than imperial measurements that were used to display food quantities on the app [63].

Despite the clear views of some women and HPs that using GDM self-management technology was more efficient in monitoring [48] and recording [68], other women were concerned about the inefficiency of their GDM systems [73,74]. Women with GDM found it was time consuming to use the system, particularly to retrieve information from food databases [74]. Postpartum women who had GDM perceived that the apps they used for GDM were overcomplicated and required too much commitment to complete a task [73]:

The apparent disinterest of HPs was a cause of discouragement for women with GDM. Some said HPs lacked interest [63,65,71] and knowledge [65] in using technology. Indeed, their HPs’ preference for a paper logbook discouraged women from using digital GDM self-management tools [65], particularly those who were already unfamiliar with such technology [27]:

Similarly, HPs were concerned about women’s abilities to use technology:

Confirming this, some women reported little or no experience of using “message boards and things of that nature” [68]. Therefore, women themselves believed that some training might be needed to increase their confidence to use such technologies [27,72]. Some women with GDM were also concerned about the privacy of personal health information recorded on the systems [25,77].

In addition, HPs were concerned about the time required to use the systems and thought it would increase their workload [25,48,66,75]. They were also concerned that some women might not be able to afford the technology [75]:

Women with GDM and HPs also experienced technical problems as barriers to the use of GDM self-management technology. Both women and HPs reported some difficulties with data transmission [62,65,71], problems with accessing technology [75], and poor access to the local internet [25,30].

Quantitative studies used various measurements to gather data. Of these, only 12% (3/26) included a usability questionnaire to evaluate their systems, as summarized in Table 1.

Of the included studies, 8% (2/26) applied the system usability scale (SUS) developed by Brook in 1996 [85], with defined acceptability ranges for SUS scores (0-50 not acceptable, 50-70 marginal and 70-100 acceptable range) [86]. Jo and Park [13] reported a marginal score for their app, just below the acceptable threshold, (69.5 of 100). A similar, but acceptable, score was reported in the study by Gianfrancesco et al [74] for their web-based dietary system (70.9 out of 100) [74]. Pustozerov and Popova [76] included a custom questionnaire wherein women with GDM rated the “usefulness” and “convenience” of their GDM system on a 10-point scale. Usefulness was rated highly (8.7 out of 10), with convenience scoring somewhat lower (7.2 of 10).

In short, although these results suggest that previous GDM systems have usability challenges, it is impossible to draw any reliable conclusions with only 12% (3/26) studies providing results from a usability questionnaire.

The included studies used different measurements to evaluate user satisfaction. Given et al [25] used an adapted version of the Telemedicine Satisfaction and Usefulness Questionnaire by Bakken et al [87]. Of the studies that included user satisfaction questionnaires, 4% (1/26) did not make their satisfaction questionnaire available [21], 12% (3/26) used specially developed satisfaction questionnaires [30,34,50], and the rest (4/26, 15%) used satisfaction questionnaires without any information on how they were developed [21,67,69-71]. Studies by Hirst et al [30] and Mackillop et al [43] were the only ones to provide evidence of the validity and reliability of their developed questionnaires.

The included studies reported generally high user satisfaction in their evaluations of GDM systems [21,25,30,34,50,67,69-71]. However, their user satisfaction questionnaires evaluated many different aspects of GDM systems regarding the type of technology and its features, making it difficult to clearly summarize areas for improvement. Table 2 shows the key measures of the user satisfaction questionnaires in the included studies (the complete measures are available in Multimedia Appendix 5). Most questionnaires used a Likert scale rating to assess the degree of participants’ agreement with their statements about the GDM systems. Women in these studies interacted with the technology within the period from GDM diagnosis until childbirth (usually between 8 and 10 weeks). They all used and evaluated the real working prototypes. Miremberg et al [21] were not included in Table 2 because the questions or satisfaction items were not available in their study.

Assessment of the aspects of convenience was common. Caballero-Ruiz et al [34] highlighted the convenience of minimizing travel to centralized clinics as the strongest indicator of satisfaction (approximately, on average, 9.5 out of 10). In other studies, women rated GDM apps highly for factors such as not complicating their lives [34,69,70] and the ability of these apps to fit into their lifestyles [30].

Improvement of GDM self-management was a highly rated aspect of the studied systems, including helping women to record BG levels [71], reminding them to take medication and record BG levels, helping them eat healthier, encouraging them to be more active [67], and helping to improve their GDM knowledge [34]. Moreover, most women found SMS text messages helpful and motivated them to optimize their GDM self-management [50,67]. A total of 2 studies also reported a general increase in women’s confidence in the management of their GDM [34,70].

Confidence or trust in GDM systems was rated well. Women with GDM reported confidence that the health care team checked their BG levels on the GDM system [71]. Many studies reported high ratings of confidence in the GDM systems, with women recommending them to others [34,43,50,67,69] or planning to use them in their next pregnancy [25,43,50,67]. Similarly, a study reported a high degree of trust (average 9 out of 10) in the GDM system [34], while another study reported that the GDM system was reliable [30].

Slightly lower satisfaction scores were reported for other aspects of ease of use: clarity of visualization of changes to treatment was rated approximately 7 out of 10 [34], and Peleg et al [70] reported satisfaction with system response time as approximately 3.5 out of 5, and ability to assist with interpreting self-monitored data approximately 3.8 out of 5.

Overall, based on the usability results (scores just under or above the acceptable threshold), there is much room for improvement in the usability of GDM self-management systems. However, with the limited number of papers providing a quantitative usability evaluation and the heterogeneity of questions assessing satisfaction, more studies are needed to identify where the improvement of usability and user satisfaction should be focused.

The primary objective of this systematic literature review was to identify the views of HPs, women with GDM, and postpartum women who have had GDM regarding GDM self-management technology. The secondary objective was to investigate the usability and user satisfaction levels of existing technologies and quantitatively evaluate these factors.

Regarding the first objective, thematic analysis of the qualitative data in the selected papers identified four themes: (1) the benefits of using technology, (2) engagement with people via technology, (3) usability of technology, and (4) discouragement factors for the use of technology.

The thematic analysis of qualitative data revealed barriers to usability, including technical problems, data privacy, lack of emotional support, the accuracy of reported data, and adoption of the system by HPs. Convenience, improving GDM self-management, peer support, increasing motivation, increasing independency, and providing consistent monitoring were common facilitators of using this technology.

For the second objective, the narrative review of the quantitative data (usability and user satisfaction) showed that there is room for improvement in the usability of GDM self-management systems.

As mentioned earlier, this theme consists of 2 main components: “engagement with peers” and “engagement with health care professionals.”

The results of the thematic analysis demonstrated the benefits of peer support in digital GDM self-management systems [68,73,75] a finding supported by similar studies outside the scope of this review. Leziak et al [90] explored the experiences of low-income women with GDM and pregnant women with type 2 diabetes, using mHealth technology to support and improve diabetes self-management during pregnancy. Their results highlighted how women valued social interactions with other women and accessed their knowledge and experiences. McMillan et al [93] evaluated mHealth technology to support postpartum women with a history of GDM in maintaining postnatal activity and good dietary habits, finding that a discussion forum was a valuable feature in doing so [93]. As other previous studies have emphasized, such favorable opinions of women toward peer support stem from their ability to share or read stories about other women [91] and receive emotional support [94], which is an important factor in health communication [95,96]. Indeed, some HPs believed that pregnant women valued other women’s experiences more than HPs’ advice during their pregnancy [97]. However, Sherman and Greenfield [94] found that, when examining message boards for pregnant teenagers, some of the medical information posted by pregnant women was misleading because it was suitable for their specific condition and therefore inappropriate for others [94]. Furthermore, validation of posted information is also a major challenge [95], and further work is needed in this area to provide a reliable and validated communication path between women with GDM.

Our thematic analysis described women’s interest in sharing data with their clinicians by remote means, to obtain reassurance and to be monitored more consistently. This is also evident in some previous studies. Dalfra et al [31] found that women with GDM and pregnant women with type 1 diabetes appreciated their telemedicine system for sharing their data with HPs and their ability to communicate with them whenever needed. Similarly, Leziak et al [90] showed that women were also in favor of sharing data with HPs and receiving real-time feedback. However, in the included studies, some HPs found it difficult to trust women’s reported data [72,76]. In contrast, Kruger et al [98] found that HPs were satisfied with the accuracy of the data reported by women with GDM via a telemedicine system. Other studies have found that it is unlikely that women would misreport their records, as they are highly motivated to maintain BG control [31] for the sake of their baby’s health [57]. Further work is needed to examine the means of decreasing the possibility of reporting incorrect data.

Although the evidence available regarding the usability of digital GDM self-management systems is limited [99], the findings of our review are in line with those of previous studies on mHealth self-management systems for type 1 and type 2 diabetes. Katz et al [100] assessed 8 current diabetes self-management apps for adults with type 1 diabetes, discovering issues in the interpretability of data and high cognitive load. These results were corroborated by Fu et al [101] in an evaluation of 4 apps for type 2 diabetes management. Further studies have also found usability challenges with data format on mHealth self-management systems [102-104], such as difficulty interpreting or understanding data in its current format [104]. A useful digital self-management system should display data trends and patterns, specifically showing which data are normal or abnormal. Usability issues with data formats thus prevent patients from understanding their data [105,106], thereby limiting their self-management capabilities.

Our review also identified limitations in the functionality of the systems as another usability concern across the included studies. Previous reviews of general diabetes self-management apps have highlighted important missing functionality, including automatic transfer of BG data from a glucometer to a mobile app, personalized diabetes management advice [107], prevention of errors [108], freedom to edit or remove data entries and appointments, and the ability to automate common tasks [109].

The limited functionality of diabetes self-management systems can be considered a usability problem [109] and is likely to result in these systems failing to meet users’ needs [107]. Addressing these functionality limitations would mitigate some of the usability challenges and help users optimize their engagement and interaction with these systems.

Quantitative evaluation of GDM self-management apps in the studies by Jo and Park [13] and Gianfrancesco et al [74] yielded SUS scores just below and above the acceptable threshold, respectively. Unsurprisingly, previous studies that used the SUS questionnaire to evaluate diabetes self-management apps in different domains have received similarly poor ratings [101,110,111]. Similar to this systematic review, these previous studies used guidance from Bangor et al [85] to interpret the SUS scores, with most apps falling below the acceptable range.

Our quantitative analysis identified the need to improve the usability of GDM self-management systems. However, with the limited number of papers providing a quantitative usability evaluation, the heterogeneity of questions assessing satisfaction and the variation in systems being assessed, it is difficult for quantitative studies to identify where the improvement of usability and user satisfaction should be focused. Therefore, it is an aspect that needs further investigation.

Despite the perceived benefits of GDM technology, our analysis revealed technical problems as a prevalent barrier across the included studies. Previous studies have reported similar technical problems when using eHealth and self-management systems [102,112-115]. Moreover, a previous systematic literature review by Simblett et al [116] identified technical problems as one of the most significant barriers to using mHealth technologies. The most common technical problems in their review were app disappearance, loss of power, restarting without warning, not receiving notifications, receiving them at the wrong time, and having a difficult connection. Indeed, 2 participants withdrew from one of the included studies because of difficulties with internet connectivity. Parallel to the findings of this review, technical problems were the cause of reducing participants’ motivations [112,113] and even the cause of leaving the study by participants with other health conditions [114,116].

In addition to technical problems, the privacy of personal health information was a concern for some women. Simblett et al [116] also reported privacy concerns in one of the included studies. Although the use of advanced encryption algorithms and pseudoanonymization of personal data should address security and privacy challenges at the system level, it is important for future GDM systems to effectively communicate good security practices to reassure new users [117].

Although most women across all studies were interested in using self-management technology, some suggested that their HPs were disinterested. Similarly, Wake et al [118] recognized the lack of awareness and adoption of technology by HPs as an important barrier to using eHealth for diabetes self-management [118]. HPs’ difficulty to accept technology was experienced in previous studies [119-121], influenced by difficulty integrating it with their workflow [102,121], lack of integration with the medical record system [120], or a lack of technical knowledge [116]. Further work is required to involve HPs in the design and development of GDM technology more effectively to reduce this barrier.