Mobile health (mHealth) technology can be an effective tool to deliver behavioral interventions to improve health outcomes for individuals with diabetes. However, there is limited evidence on mHealth for pregnant women managing diabetes or for the role of health care providers (HCPs) in delivering mHealth tools. To prepare for the development of a pregnancy-specific mHealth intervention for diabetes, we sought to understand HCPs’ acceptance, utilization, and design recommendations for how mHealth technology can best be used to support the management of diabetes during pregnancy.
Focus groups and individual interviews were conducted with clinical HCPs (obstetricians, nurses, diabetes educators, dietitians, and lactation counselors) who care for low-income pregnant women with diabetes. Data were analyzed using the constant comparative method.
Participants (n = 29) generated five themes and 14 subthemes. Knowledge and acceptance themes included utilization of mobile resources and barriers to HCPs’ adoption. Knowledge of popular pregnancy applications (apps) dominated, while familiarity with and use of diabetes apps varied by profession and proximity to continued education. Technology adoption barriers included lack of credibility, dependence on Internet resources, and reliance on knowledge networks. Design themes addressed user-centered design, responsive and practical education, and clinical integration. Clinicians recommended patient-participatory design, customizable features, data-tracking features, motivational feedback, and bidirectional communication capabilities.
Although HCPs strongly embraced mHealth tools as an adjunct to obstetric care for women with diabetes, they also expressed limited knowledge and apprehension about patient engagement. HCP recommendations and concerns must critically inform the development of meaningful tools for this population.
Gestational diabetes mellitus (GDM) and type 2 diabetes during pregnancy are common and increasing perinatal comorbidities in the United States (1–3). Both types of diabetes pose significant threats to maternal and fetal health, as the relationship between glycemic variability increases risks of adverse perinatal outcomes (2,3). Pregnant women with diabetes require consistent clinical care, education, and self-management to reach blood glucose targets, which is essential to promote optimal maternal and fetal health outcomes. Although effective clinical management and medical treatment can reduce the morbidity associated with diabetes during pregnancy (4–6), the demands of such intensive management require advanced patient engagement and skills (7,8). The intensity and duration of diabetes self-management demands create burdens for women and impose even greater burdens on women of low socioeconomic status, who experience greater barriers to care (4,9–12).
A rapidly growing approach in the clinical management of diabetes is the adoption and integration of technologies, including smartphone applications (apps) to support independent self-management. Research has documented the success of mobile behavioral interventions for the management of chronic disease among diverse populations; additionally, mobile health (mHealth) technology has been perceived to be an accessible and scalable strategy to influence positive behavior and clinical health outcomes (13–15). Research specifically on mHealth and diabetes has identified the potential efficacy of mHealth for self-monitoring and making behavioral changes related to glycemic stability, diet and exercise, insulin dosage, and medication engagement in contexts outside of pregnancy (16–18). However, pregnancy is a unique period of amplified learning demands, greater barriers imposed by social determinants of health, and an abbreviated timeline for health learning and health behavior adoption. Thus, evidence-based practices designed for nonobstetric patients may not be suitable for the clinical needs of pregnant women. Yet, minimal data exist on mHealth tools for the management of diabetes during pregnancy.
Previous design studies of mHealth interventions have incorporated user-centered or codesign principles to understand the population context and promote health interventions (19–21). As provider stakeholders and potential endorsers of mHealth interventions, understanding the preferences and needs of obstetric health care providers (HCPs) is vital in conceptualizing clinician acceptability and feasibility of mHealth tools for use during pregnancy. Consideration of HCP attitudes toward mHealth technology is essential when clinical management requires a robust, coordinated care approach to promote patient self-management skills. For this reason, we aimed to study obstetric HCPs’ utilization and acceptance of and design recommendations for mHealth in support of diabetes during pregnancy to inform the development of an mHealth intervention for this population.
Research Design and Methods
This was a qualitative research study utilizing focus groups and face-to-face semistructured interviews. The primary collection of data through focus groups was chosen for their strength in generating ideas through interactivity and to obtain variant experiences on the topic of mHealth use in applied practice (22). Each focus group was composed of three to six participants to allow each participant an opportunity to share experiences within the whole group (23). For those who could not attend a focus group because of schedule conflicts, individual interviews were conducted. Participants were compensated with a gift card. Northwestern University’s institutional review board approved all research activities, and all participants provided written informed consent.
The study recruited HCPs serving low-income pregnant women from an academic antenatal clinic located in Chicago, IL. HCPs were identified by reviewing the institution’s clinical personnel records and recruited via e-mail outlining the purpose of the focus groups. HCPs who were English-speaking and experienced in treating or educating pregnant women about GDM or type 2 diabetes were invited to participate. To ensure diverse professional experience, a purposive sample of HCPs was sought, including individuals with one of the following professional titles: practicing physician (resident, fellow, or attending), nurse, diabetes educator, lactation peer counselor, social worker, medical assistant, patient services representative, or registered dietitian.
To understand how best to identify HCP acceptability and considerations regarding mHealth interventions, we reviewed the literature on stakeholder codesign frameworks and approaches to inform the development of an interview guide. We reviewed literature pertaining to participatory stakeholder methodology, sociotechnical codesign, and development of person-based approaches (21,24,25) Our intention was to capture HCPs’ organizational experience, perceived facilitators of and barriers to mHealth use, and engagement in identifying novel approaches to reduce health disparities related to diabetes management during pregnancy (Table 1).
Principles of Stakeholder Co-Design and Person-Based Approaches
| Principle . | Topic . | Example Interview or Focus Group Questions . |
|---|---|---|
| Stakeholder and organizational experience | Management resources | What outside resources do you recommend patients use to manage diabetes and/or pregnancy, if any? What makes them particularly helpful? |
| Identifying effective strategies | What do you think makes the most effective medium for engaging with patients and why? | |
| Defining key needs and challenges that mHealth interventions potentially can address | Knowing that patients are using smartphone apps for health, do you think they would be useful for diabetes? Why or why not? | |
| Feasibility and appropriateness | Contextualizing current practices | Do you have a sense of how patients already may or may not use smartphone apps to manage pregnancy and/or diabetes care? Do you speak with your patients about using any health apps? |
| Utility for patients | In your experience, how helpful do you believe these apps are to patients? | |
| Utility for HCPs | Are health apps useful to you as an HCP? How? | |
| Collaborative design solutions | Creating and identifying mHealth features to optimize interventions | Imagine the ideal smartphone app to help your patients manage their diabetes care during pregnancy. Describe it to me. What should it look like? What are its functions? |
| Defining opportunities in mHealth interventions to reduce health disparities | Do you have any other thoughts on how best to incorporate language, literacy, or cultural barriers into the development of an app? |
| Principle . | Topic . | Example Interview or Focus Group Questions . |
|---|---|---|
| Stakeholder and organizational experience | Management resources | What outside resources do you recommend patients use to manage diabetes and/or pregnancy, if any? What makes them particularly helpful? |
| Identifying effective strategies | What do you think makes the most effective medium for engaging with patients and why? | |
| Defining key needs and challenges that mHealth interventions potentially can address | Knowing that patients are using smartphone apps for health, do you think they would be useful for diabetes? Why or why not? | |
| Feasibility and appropriateness | Contextualizing current practices | Do you have a sense of how patients already may or may not use smartphone apps to manage pregnancy and/or diabetes care? Do you speak with your patients about using any health apps? |
| Utility for patients | In your experience, how helpful do you believe these apps are to patients? | |
| Utility for HCPs | Are health apps useful to you as an HCP? How? | |
| Collaborative design solutions | Creating and identifying mHealth features to optimize interventions | Imagine the ideal smartphone app to help your patients manage their diabetes care during pregnancy. Describe it to me. What should it look like? What are its functions? |
| Defining opportunities in mHealth interventions to reduce health disparities | Do you have any other thoughts on how best to incorporate language, literacy, or cultural barriers into the development of an app? |
A semistructured interview guide consisting of open-ended questions was used to elicit responses to queries regarding the development of an mHealth app for low-income pregnant women with diabetes. Given the similarities in management of type 2 diabetes and GDM, the interviews focused broadly on the management of both conditions during pregnancy. HCPs were asked to envision desired functions and features of a prenatal wellness and diabetes self-management mobile app for low-income women and were queried about their knowledge and opinions on current mHealth technology. HCPs were also asked to specifically consider the impact of social determinants of health in the management of diabetes during pregnancy and to consider any unique needs of low-income women. Focus groups and interviews were audio-recorded and facilitated by a trained research assistant who was not a supervisor of or directly involved in any clinical responsibilities of the participants. No personal information was asked of participants. Each focus group lasted 60–90 minutes; individual interviews lasted ∼60 minutes.
The focus group and interview recordings were professionally transcribed verbatim. All transcripts were uploaded to a secure, Web-based qualitative data analysis software (www.dedoose.com). Interview transcripts were analyzed using the constant comparative method to identify themes and patterns in the data (26). A random selection of transcripts was chosen for initial reading and generation of codes, or tags, of recurring concepts called subthemes by two members of the research team. All study team members met to discuss, negate, and combine preliminary codes and create a working codebook. The remaining transcripts were coded using the codebook, which was iteratively modified as the remaining transcripts were analyzed (27).
After all transcripts had been analyzed, all study team members met to collapse the subthemes into overarching themes in the data. The sample size was determined based on the concept of saturation, or the point at which the collection of further data no longer yielded novel results (26). After reaching saturation, the final themes and subthemes were analyzed. Themes and specific subthemes are discussed in detail below.
Results
Twenty-four HCPs participated in five focus groups, and an additional five HCPs underwent individual interviews. The professional disciplines represented are listed in (Table 2). All participants had experience in treating and/or educating pregnant women with diabetes and worked with both privately and publicly insured women. HCPs ranged in years of experience from 1 (first-year residents in obstetrics and gynecology) to >20 (an experienced perinatal diabetes-focused advanced practice registered nurse [APRN]).
Participant Characteristics
| . | n (%) . |
|---|---|
| Provider sex | |
| Female | 28 (97) |
| Male | 1 (3) |
| Provider type | |
| Physician | 14 (48) |
| Maternal-fetal medicine faculty | 6 (21) |
| Maternal-fetal medicine fellow | 3 (10) |
| Obstetrics and gynecology resident | 6 (21) |
| Registered nurse* | 6 (21) |
| Medical assistant | 2 (7) |
| Registered dietitian | 1 (3) |
| Licensed clinical social worker | 1 (3) |
| Lactation counselor | 1 (3) |
| Certified health education specialist | 4 (14) |
| . | n (%) . |
|---|---|
| Provider sex | |
| Female | 28 (97) |
| Male | 1 (3) |
| Provider type | |
| Physician | 14 (48) |
| Maternal-fetal medicine faculty | 6 (21) |
| Maternal-fetal medicine fellow | 3 (10) |
| Obstetrics and gynecology resident | 6 (21) |
| Registered nurse* | 6 (21) |
| Medical assistant | 2 (7) |
| Registered dietitian | 1 (3) |
| Licensed clinical social worker | 1 (3) |
| Lactation counselor | 1 (3) |
| Certified health education specialist | 4 (14) |
Including one APRN.
Thematic analysis of focus group data consisted of two broad domains: 1) HCPs’ knowledge, acceptance, and adoption of mHealth technology for diabetes during pregnancy and 2) HCPs’ mHealth design recommendations for diabetes education and support during pregnancy. Themes within the first domain included “utilization of mobile resources” and “barriers to HCP adoption” (Table 3). Themes within the second domain included “user-centered design,” “responsive and practical education,” and “clinical integration” (Table 4). Each theme encompassed two or more additional subthemes, discussed below.
HCPs’ Knowledge, Acceptance, and Adoption of mHealth Technology for Diabetes During Pregnancy
| Theme . | Subtheme . | Illustrative Quotes . |
|---|---|---|
| HCPs’ utilization of mobile resources | Knowledge of popular commercial apps | “I would say the large majority are … using apps to follow their pregnancy. Developmental stages, what stage your fetus is at, all of those things, but I can’t speak to how many are using it to manage their diabetes, but to follow their pregnancy, track it, and just have, you know, the What To Expect, Baby Center, [apps] all those things, are used pretty often.” (fellow physician) |
| Proximity to continued patient education on diabetes self-management | “CalorieKing.com is all the foods. We use it with [the Health Learning Center] … and then, we also use a lot of apps that really log glucose, like Glucose Buddy or GlucoseShare—anything that’s free on the phones, that we have seen and that really helps, and then on MyFitnessPal is a way to count carbohydrates as well as log your exercise.“ (health educator) | |
| Barriers to HCP adoption | Lack of credibility and/or evidence-based mobile resources | “But, to be honest, I think that in all of my practice, I don’t routinely rely on apps for patients …. I think that’s probably just because of lack of familiarity and lack of examples of seeing that integrated effectively into patient care.” (physician) |
| Confidence in Internet resources | “Sometimes, I’ll … print off some information that I get off of, you know, whatever [American College of Obstetricians and Gynecologists or American Diabetes Association] website, so especially if people are interested in some type of …Internet search, I at least try to steer them to a site that I know is reputable … but I don’t really recommend too much else.” (physician) | |
| Reliance on HCP knowledge networks | “I think that there are … trusted sites where there becomes this general consensus amongst providers or clinical teams like ‘Okay, yeah, that’s a trust site.’ Like, I don’t even have to see it, but everyone trusts it, so we can give that out, right?” (resident) | |
| Low patient engagement | “I think a smartphone app could be a useful adjunct, but I think also it’s like, is it gonna be useful for the people who already have a buy-in into using it? Right? …. I feel like, in our clinic, there are people who just won’t do it for whatever reason or can’t do it.” (fellow physician) |
| Theme . | Subtheme . | Illustrative Quotes . |
|---|---|---|
| HCPs’ utilization of mobile resources | Knowledge of popular commercial apps | “I would say the large majority are … using apps to follow their pregnancy. Developmental stages, what stage your fetus is at, all of those things, but I can’t speak to how many are using it to manage their diabetes, but to follow their pregnancy, track it, and just have, you know, the What To Expect, Baby Center, [apps] all those things, are used pretty often.” (fellow physician) |
| Proximity to continued patient education on diabetes self-management | “CalorieKing.com is all the foods. We use it with [the Health Learning Center] … and then, we also use a lot of apps that really log glucose, like Glucose Buddy or GlucoseShare—anything that’s free on the phones, that we have seen and that really helps, and then on MyFitnessPal is a way to count carbohydrates as well as log your exercise.“ (health educator) | |
| Barriers to HCP adoption | Lack of credibility and/or evidence-based mobile resources | “But, to be honest, I think that in all of my practice, I don’t routinely rely on apps for patients …. I think that’s probably just because of lack of familiarity and lack of examples of seeing that integrated effectively into patient care.” (physician) |
| Confidence in Internet resources | “Sometimes, I’ll … print off some information that I get off of, you know, whatever [American College of Obstetricians and Gynecologists or American Diabetes Association] website, so especially if people are interested in some type of …Internet search, I at least try to steer them to a site that I know is reputable … but I don’t really recommend too much else.” (physician) | |
| Reliance on HCP knowledge networks | “I think that there are … trusted sites where there becomes this general consensus amongst providers or clinical teams like ‘Okay, yeah, that’s a trust site.’ Like, I don’t even have to see it, but everyone trusts it, so we can give that out, right?” (resident) | |
| Low patient engagement | “I think a smartphone app could be a useful adjunct, but I think also it’s like, is it gonna be useful for the people who already have a buy-in into using it? Right? …. I feel like, in our clinic, there are people who just won’t do it for whatever reason or can’t do it.” (fellow physician) |
HCPs’ mHealth Design Recommendations for Diabetes Education and Support During Pregnancy
| Theme . | Subtheme . | Illustrative Quotes . |
|---|---|---|
| User-centered design | Participatory design | “I think having people who have been pregnant with diabetes themselves who are not medical people consulting on how to put the app together or what content goes in the app would be helpful.” (physician) |
| Customization and specificity | “It would have, first, … different settings based on their actual diagnosis. So, do they have preexisting diabetes? What kind? Do they have gestational diabetes? What kind? And then would … have to be further … catered to the appropriately managed, based on what type of diabetes diagnosis they have.” (physician) | |
| Interactivity | “I think with apps today, it’s important to have some type of social component to it … where you could talk to providers openly [or] you could talk to just peers who are dealing with the same … diabetes that you’re dealing with. I think there should be some type of social component to keep you engaged” (lactation counselor) | |
| Responsive and practical education | Diabetes education through data visualization | “Having the glucose levels in … graphical form, like you can look by a week and see, “Oh, okay, the weekends—that’s when my spikes are,’ or … nice ways of … looking at fever curves and stuff like that, so it would be a similar kind of thing for the glucose curves, just showing it in different ways.” (resident physician) |
| Real-time motivational feedback | “I think … especially for blood sugar specifically or like diet changes, … it’s an awesome place to store that information and then having that flipside that you can get some sort of response, whether it be … motivation to keep going…. I think so much about diabetes is about behavior change … having that motivation right at your fingertips could be very helpful.” (resident physician) | |
| Practical | “Because, once you get home with all this information, then they’re scared, you know? “How do I do this? How do I eat this? When do I eat it?” You know? If it’s on an app, then they can go, ‘I don’t know what to eat, so let me go to the app. Let me look at the diet.’ You know, ‘Let me look at how many carbs are in this.’” (health educator) | |
| Clinical integration | Provider engagement | “I think my role would be … helping them navigate the app and … education on it, encouragement to use it.” (social worker) |
| Data-driven to inform prenatal clinical visit | “It could be nice for that app to feed directly to whoever their provider is so … when they come for their visit, they can just download everything there, and everything is available, and … you almost get, like, real-time updates.” (resident physician) |
| Theme . | Subtheme . | Illustrative Quotes . |
|---|---|---|
| User-centered design | Participatory design | “I think having people who have been pregnant with diabetes themselves who are not medical people consulting on how to put the app together or what content goes in the app would be helpful.” (physician) |
| Customization and specificity | “It would have, first, … different settings based on their actual diagnosis. So, do they have preexisting diabetes? What kind? Do they have gestational diabetes? What kind? And then would … have to be further … catered to the appropriately managed, based on what type of diabetes diagnosis they have.” (physician) | |
| Interactivity | “I think with apps today, it’s important to have some type of social component to it … where you could talk to providers openly [or] you could talk to just peers who are dealing with the same … diabetes that you’re dealing with. I think there should be some type of social component to keep you engaged” (lactation counselor) | |
| Responsive and practical education | Diabetes education through data visualization | “Having the glucose levels in … graphical form, like you can look by a week and see, “Oh, okay, the weekends—that’s when my spikes are,’ or … nice ways of … looking at fever curves and stuff like that, so it would be a similar kind of thing for the glucose curves, just showing it in different ways.” (resident physician) |
| Real-time motivational feedback | “I think … especially for blood sugar specifically or like diet changes, … it’s an awesome place to store that information and then having that flipside that you can get some sort of response, whether it be … motivation to keep going…. I think so much about diabetes is about behavior change … having that motivation right at your fingertips could be very helpful.” (resident physician) | |
| Practical | “Because, once you get home with all this information, then they’re scared, you know? “How do I do this? How do I eat this? When do I eat it?” You know? If it’s on an app, then they can go, ‘I don’t know what to eat, so let me go to the app. Let me look at the diet.’ You know, ‘Let me look at how many carbs are in this.’” (health educator) | |
| Clinical integration | Provider engagement | “I think my role would be … helping them navigate the app and … education on it, encouragement to use it.” (social worker) |
| Data-driven to inform prenatal clinical visit | “It could be nice for that app to feed directly to whoever their provider is so … when they come for their visit, they can just download everything there, and everything is available, and … you almost get, like, real-time updates.” (resident physician) |
Knowledge, Acceptance, and Adoption of mHealth Technology
Theme 1: HCPs’ Utilization of Mobile Resources
HCPs’ knowledge of free and publicly available mobile resources for diabetes support, management, and education was diverse and varied according to profession. Utilization of and ability to recommend mobile resources in clinical practice aligned within the subthemes of “knowledge of popular commercial apps” and “proximity to providing continued patient education on behavior change.”
First, for physicians, awareness of popular commercial apps for nutrition, physical activity, and fetal development was greater than knowledge of apps designed specifically for diabetes support. Physician knowledge of resources was also less robust than that of other nonphysician HCPs. For example, a physician reflecting on encounters with app utilization among diabetes patients, said, “I can’t speak to diabetes care. For pregnancy, … I feel like the ones that most people have are ways to say, like, ‘You check, you put in your due date, and it gives you, like, a weekly update on like the size of your baby based on a fruit.’”
Second, greater proximity to continued patient education was noted to be associated with increased use of mHealth. The more HCPs’ roles included the actual provision of self-management education rather than referrals for such education, the more knowledgeable they had of diabetes apps. Health educators and dietitians referenced the use and availability of diabetes mobile apps as essential to providing ongoing patient education in diet instruction and monitoring of physical activity. Moreover, they noted that mHealth technology facilitated their ability to be sensitive to varying degrees of health and technology literacy. As one health educator detailed:
“There’s, like, at least 10 apps out there for glucose logging, and carbohydrate and exercise activity logging, and the ones that are good are easy [and] are usually on the cell phone platform already. Like, MyFitnessPal is already on the phone if you have a smartphone. It’s already there; you don’t have to pay a dime for it.”
Although health educators were aware of and likely to use mobile apps in practice, there was no reference to an optimal, centralized app for diabetes support. Health educators often spoke of a single app with a single function. For example, a health educator cited her frustration with the lack of a comprehensive diabetes mobile app: “I also have seen a lot of patients, not in this specific population, use apps like MyFitnessPal or LoseIt or something like that to track the food they’re eating. So, to put all their diabetes care into one app instead of, ‘Let’s track my food intake on here, and my glucose on this one, and my insulin on this one’ [would be good].” Participants also noted a need for a pregnancy-specific diabetes mHealth tool.
Theme 2: Barriers to HCP Adoption
Our analysis identified four prominent subthemes related to barriers to HCP knowledge and subsequent lack of recommendation of mobile apps for diabetes support to patients. These subthemes include “perceived lack of credibility and/or evidence-based mobile resources,” “confidence in Internet resources,” “reliance on provider knowledge networks,” and “low patient engagement.”
First, the most significant barrier to HCP adoption of mHealth for diabetes support was the perception of mobile apps having poor credibility or are not affiliated with trusted sources. Physicians in this cohort were more likely than other HCPs to express not knowing any evidence-based mobile resources and were often unable to suggest or recommend mobile apps to patients in support of clinical care. One physician expressed reluctance to endorse mHealth tools if they could not “vouch” for its medical accuracy:
“I mean… sure, there are tons of pregnancy apps, which I think are appropriate for … most lay information, but if I’m talking specifically to someone who has … [a] chronic medical condition and pregnancy, I would just want to make sure … that the information was appropriate and valid.”
A second subtheme was a shared sense of uneasiness regarding Internet-based resources serving as patients’ primary source of information outside of the clinical care setting and associated concerns about the reliability of such information. Although this perception held true across all focus groups and interviews, most HCPs also had greater confidence in recommending organizational Web resources and education than external resources or apps. For one HCP, it was not “natural” to recommend apps in comparison with Web-based sites:
“I think I recommend a lot of websites, forgetting that these websites also have apps. Like, I don’t know why referring them to … the app is not my first train of thought. I just give them the website, and I guess they can figure out later if there’s an app to it, but it’s never, like, my first thing. Like, there’s an app for that—you know? I’m more of, ‘There’s a website, or you can Google this,’ [or] something like that. I don’t necessarily think of an app.”
Another subtheme was related to reliance on HCP knowledge networks. Many HCPs described knowing or learning about external (i.e., noninstitutional) diabetes resources for patient dissemination through other HCPs, which limited a wider awareness of high-quality resources. Informal sharing of resources within the formal organizational structure, consequently, limited HCPs’ familiarity with reputable external diabetes resources for patient education and engagement. For resident physicians, knowledge of resources was passed down from senior staff; one resident described this experience with acquiring knowledge of diabetes resources for patient recommendation:
“I feel like I see what the goal is supposed to be for care of diabetes in pregnancy in the context of …speaking to the fellows … and to the attendings in diabetes clinic and that the resources, they clearly know like a million times more than I do, so … I tend to follow what their recommendations have been.”
Finally, while some HCPs emphasized the benefits of continued patient education via mHealth apps, others spoke to engagement challenges in low-income patients’ utilization of mobile technology in support of care. The subtheme of low patient engagement was not a prevalent barrier but did discourage a subset the interest of some HCPs in adopting mobile apps for patient care. When referring to specific patient adoption challenges, HCPs mostly cited low motivation and uptake as probable causes for disengagement. For one HCP, apprehension about patient uptake stemmed from personal experiences of patient nonengagement and therefore reluctance to use mHealth as a supportive tool for clinical care:
“I think it would be great if you could get … buy-in… like, if you can get people to want it and be engaged in their care and show them that this is a way that can get easier for them to engage in their care, but ultimately, I don’t know, unless you can get that … sense of self-efficacy, like, ‘I can do this, and I want to do this,’ and … you can get that sort of commitment from someone, it’s not necessarily going to fix it.”
HCPs’ mHealth Design Recommendations for Diabetes Education and Support During Pregnancy
Theme 3: User-Centered Design
HCPs across all focus groups and interviews agreed on the importance of user-centered design and specifically commented on leveraging patient experiences, knowledge, and preferences to enhance self-efficacy and diabetes self-management. Achieving the goal of user-centered design was suggested via three subthemes: “participatory design,” “customization and specificity,” and “interactivity.”
Participatory design encompassed the implementation of a peer-driven model, use of person-first language, and the provision of motivation for the whole pregnancy, not just diabetes-related concepts. Incorporation of a peer-driven model was believed to allow for inclusion of women’s lived experiences to inform the development of a mobile app in providing educational support for diabetes self-management. As one physician suggested, “I think it should be as much as possible written, or at least seem to be written, by peers—you know, maybe people who have been pregnant with diabetes themselves, and they are now kind of almost offering advice or support.”
As a component of participatory design, sensitivity to the appropriate and balanced use of evidence-based medical information was viewed as a necessity. For example, one HCP, reflecting on her day-to-day communication with patients led her to consider the importance of sourcing user experiences directly for relatability and health literacy:
“I think, at this point, I probably don’t realize how often I'm using ‘medical-ese’ that doesn’t make sense, and I know that someone else could more effectively … communicate without medical jargon in the app…. I’m sure it would be really interesting to interview patients and say, like, what has made sense to them, and what language they find most clear and helpful.”
The second subtheme of user-centered design was customization and specificity, which were considered essential to patient engagement. HCPs considered the ability to customize app functions and features as a patient-centric approach to promote engagement in diabetes education and support. Customizable features such as allowing patients to add journaling, tracking, alerts, or nutrition education features, would allow patients control over their management needs and preferences. HCPs perceived that the purpose of customization, as needed, was to recognize patients’ varying experiences with managing diabetes. As the APRN explained, not all patients require the same level of engagement:
“… Everyone is different, and so, if there was an app that had a function … you know, like not everyone forgets to eat, so if there’s an app that just has a function like [to] turn on snack reminders or whatever … or, you know, … like maybe you would get a list of all the things that you might choose to be notified about, and then you could pick what would be helpful to you.”
For other HCPs, customization related to the inclusion of features that recognized a patient’s preferred language, dietary customs, and diabetes type to individualize the app and lessen standardization. As one registered dietitian explained:
“I think a lot of people’s beliefs about food and about causes of diabetes could come into play in making different versions of the app for different communities, although I would not want to necessarily stereotype people just by what community they come from, but certainly I’ve had patients who are Hispanic who eat very differently from patients who are not and patients who are immigrants who eat differently than people who are born here, and, you know, I think you can have different versions to a degree based on someone’s background and resources and things like that.”
Another subtheme of user-centered design was inclusion of interactive communication features. HCPs discussed the importance of user interactivity to encourage social connection among pregnant women managing GDM or type 2 diabetes and suggested features that could address patients’ social, cultural, or medical concerns. Some HCPs desired a patient/provider communication feature to assist patients who experienced difficulties managing blood glucose levels or with a history of variability in recorded glucose results for extra support in reaching independent self-management.
Theme 4: Responsive and Practical Education
Responsive and practical education was viewed as an essential component in providing diabetes management education for low-income women via a mobile app. Ideas HCPs had regarding the components of a responsive and educational diabetes mHealth platform were organized into subthemes of: “provides diabetes education through data visualizations,” “contains real-time motivational feedback,” and “is practical for daily use outside of the clinical setting.”
The subtheme of data visualization was prominent. HCPs envisioned patients using mobile apps for daily logging and tracking (i.e., of glucose levels, dietary intake, and exercise activities) to receive immediate feedback via data visualizations. Visual elements such as graphs and summarized reports were considered significant in displaying relationships and providing visual cues to influence behavior change or maintenance. For example, a health educator cited her patients’ desires for a simple and straightforward way to track glucose levels in tandem with dietary behaviors:
“I think one thing … is a place to very easily and clearly track both what they’re eating and their corresponding blood sugar afterward … so they can look back. Like, I have a lot of moms who are interested in looking back and seeing patterns, and having that displayed really clearly to them [would be helpful].”
HCPs viewed the ability to visually display patient-entered data as a means of providing ongoing education about glycemic targets and of assisting patients in understanding day-to-day behavior patterns. One physician described the importance of an app being able to respond with advice and supportive notifications based on a patient's data reports, saying, “I think probably dietary advice would be great … based on their glucose. It could like show, like, oh, that five of your six values are above goal and flag it and give them some information.”
Related to data visualization, another emergent subtheme was the importance of mobile apps being able to provide real-time motivational feedback. HCPs expressed awareness and value of positive reinforcement as an impetus to diabetes self-management and self-efficacy. As one physician commented, “I think the other thing is just, again, like really congratulating people for … any level of accomplishment, because I think it’s really hard, and so … giving people positive reinforcement is really important for them.”
The final subtheme included providing practical education about managing diabetes during pregnancy. For one HCP, it was important for mHealth apps to provide suggestions about integrating health behavior changes into daily routines. For others, a mobile app should provide information relevant to pregnancy, as well as medical information for healthy clinical outcomes. AS one HCP said, “It would be great to have some resources for diabetes and pregnancy, obviously, and how that relates to the baby and outcomes for the baby as well [as] how-to videos.” Such perspectives highlighted the importance of mHealth tools serving as a repository of practical information and tips alongside the motivational elements described above.
Theme 5: Clinical Integration
The theme of clinical integration was posited as an important design feature for an mHealth tool for pregnant women with diabetes. The first subtheme was “provider engagement” as essential to promote mHealth for health behavior support. Across all HCP types, physicians were perceived as the primary professional to interface with a mobile app during clinical visits (e.g., by reviewing glucose logs in tandem with the patient), whereas nonphysician HCPs would encourage the continued use of mHealth devices to support ongoing health education and behavior change activities. For physicians, the use of a mobile app to augment diabetes education was a means to guide management conversations and inform long-term care education. As one physician envisioned the interface of mobile apps and patient care:
“They can walk me through the app and their day, like how the app fits into their day, and then, what would be helpful for me, is to have … summarized data that I can review beforehand to have some idea and then to sit down with the patient knowing that information and go through the app with them, and they … talk to me about things.”
The second subtheme of clinical integration was data review and collection to “inform prenatal clinical visits.” HCPs recognized the advantage of viewing patients’ logged health data via an app ahead of clinic visits to further understand blood glucose levels and formulate a plan of care and perceived data-driven mHealth apps to be of significant value. Furthermore, recorded data would alleviate physician dependence on patients bringing written logbooks.
HCPs envisioned the integration of a mobile app into routine prenatal care as being both utilitarian and advantageous in serving as a technology-enhanced resource. For some HCPs, the prospect of an app with the ability to transfer or download data would streamline administrative tasks associated with patient care such as managing changes in prescription regimens, updating medical charts, and completing clinical progress notes. One physician stated, “If there was a way to connect [a] patient’s app portal, … like, ‘This is your new regimen,’ ‘this is what your Levemir is,’ or ‘this is what your Lantus is’ … so that they have the updated regimen in their app and that the physician can kind of update that as they go.”
Discussion
The integration of mHealth technology into clinical care for individuals with diabetes is increasingly common, yet few data directly address the unique needs of pregnant women with diabetes. This investigation aimed to understand HCPs’ utilization and acceptance of and design recommendations for mHealth in support of diabetes care for pregnant women. A diverse sample of HCPs contributed varied perspectives and experiences of mHealth knowledge and application in the clinical care setting, while also offering design recommendations for the development and delivery of mHealth support tools for diabetes self-management during pregnancy. Critical themes addressed the supportive factors for the utilization of mobile resources, barriers to HCPs’ adoption, support for user-centered design, requests for responsive and practical education, and perspectives on mHealth clinical integration.
Themes related to barriers to adoption and recommendation of mHealth tools were consistent with other studies citing lack of evidence-based mHealth knowledge among physician HCPs and emphasized a greater reliance on Internet-based resources rather than mHealth resources among all HCPs (14,28). Although acceptance and acknowledgment of mHealth apps as adjunct support tools to clinical care were highly endorsed in our sample, HCPs expressed anxiety about long-term engagement among patient populations who may lack buy-in for the use of mobile technologies for health maintenance. Similar studies have reported challenges with patient engagement with and adoption of mHealth apps because of a lack of integration into patients’ personal routines (29), data security and privacy concerns (30), natural loss of interest in app content (31), and patients’ low health literacy (32), all of which support the HCPs’ concerns in this study. A novel finding in our study was that HCP-based barriers may also differ by profession; in this study, nonphysician HCPs were more likely to be aware of commercial diabetes apps to recommend to patients compared with physicians, whose knowledge was limited to pregnancy tracking apps.
In imagining the ideal mHealth app for women managing diabetes during pregnancy, HCPs desired the design and development process incorporate a participatory approach. For HCPs, a participatory design required sourced end-user experiences and preferences to develop an app providing appropriate diabetes messaging, while also centering on the whole pregnancy experience. This theme reflects a widely established and applied design feature of mHealth interventions, which include end-users’ lived experiences as expert knowledge to develop an efficient and relevant platform (33–35). HCPs also suggested that customizable features should not only translate to user health and lifestyle preferences (i.e., customization of language, diet, and diabetes type), but also that there should be the ability to filter behavior management features to meet individual patients’ needs and objectives. For example, it was believed that a patient may be sufficiently skilled at glucose tracking, but may need greater support in terms of engaging educational messaging and behavior support for physical activity and food choices. HCPs also highlighted that self-management needs among low-income women may change from initiation of prenatal care to the postpartum period, and the context of mobile interventions should adapt to meet changing needs.
Data visualization and motivational support were also features strongly desired by HCPs. These findings are supported by previous qualitative research highlighting the need for mHealth functions and features for chronic disease management to be based in data-tracking to inform disease knowledge, visual cues to incite behavior change, and motivational messaging to encourage continued behavior regulation (36–38). Burford et al. (38) found that type 2 diabetes self-management skills were positively affected when personal data were entered and tracked within an mHealth app to produce daily visual reports for patient interpretation of health maintenance. Such findings suggest that future diabetes mHealth support tools could pair glucose tracking with education and responsive motivational features.
HCPs in our study also highlighted the importance of social connectivity in providing motivational support. Previous work has indicated that social engagement with others who are managing type 2 diabetes of GDM increased therapy adherence, diabetes knowledge, and self-management performance (39–41).
Although all HCPs agreed on the value integrating mobile technology into the clinical setting, some had reservations about operationalization and ease of use. HCPS believed that data-capturing tools such as mHealth apps could potentially streamline clinical activities and focus patient visits to target key areas of improvement, but that technology could not replace person-centered care. This theme is similar to research conducted by Nundy et al. (42), which discovered that HCPs believed receiving patient-generated health information via mHealth platforms had the ability to inform clinical visits by focusing on patients’ self-care needs, management barriers, and perspectives; however, when an mHealth tool was piloted in a clinical trial, HCPs were divided regarding the platform’s ability to inform clinical decision-making and patient knowledge. Our data suggest that these barriers exist in this setting as well, based on our findings of clinical integration and adoption themes; however, in this study, HCPs’ interest in mHealth tools was not limited to patient monitoring and tracking, but also included use of an educational and health behavior change tool in the promotion of diabetes self-management.
Strengths of our study include diverse HCP representation, allowing for inclusion of a broad range of professional experiences. The use of in-depth interviews also allowed for exploration of themes that may not be uncovered via quantitative research. Limitations of our study included that we intentionally did not offer specific diabetes management skills for HCP consideration, which may have limited conceptualization to management areas (i.e., diet and exercise) and not targeted behavior processes. Moreover, we only examined mobile apps in the delivery of diabetes education and support and did not investigate the use of other diabetes management technologies that may similarly interface with smartphones, such as continuous glucose monitoring devices. In addition, further research is required to determine whether the themes and subthemes generated by our HCP participants align with or diverge from those from other HCPs who care for similar urban populations.
Conclusion
Our findings suggest that HCPs embrace mHealth tools as an adjunct to obstetric care for women with diabetes. However, potential barriers to adoption and integration exist, including challenges with mHealth familiarity, credibility, clinical integration, and ongoing patient engagement. HCPs’ ability to recommend, apply, or adopt mHealth tools varied by profession and prior exposure. HCPs with limited understanding of mHealth utility, availability, and capacity to reach vulnerable populations present a barrier to patient uptake and usage of mHealth tools. As mHealth interfaces for pregnant women with diabetes become more common and complex, it is important to include HCP input in the development and design of digital health tools to ensure continued health promotion and enhance clinical acceptability and uptake. In addition to contributing to the literature on accessibility technologies and self-management education for women managing diabetes during pregnancy, this research presents a foundation to inform the design and implementation of mHealth interventions designed to achieve optimal pregnancy outcomes.
Article Information
Funding
L.M.Y. was supported by the National Institute of Child Health and Human Development (NICHD) grant K12 HD050121-11 at the time of the study. Additionally, this work was supported by NICHD grant 1R21 HD094271-01 and by the Friends of Prentice Women’s Health Grants Initiatives.
Duality of Interest
No potential conflicts of interest relevant to this article were reported.
Author Contributions
J.J. and K.L. collected and analyzed the data. J.J. wrote the initial manuscript draft. K.L., C.M.N., and L.M.Y. contributed to the analysis discussion and critically edited the manuscript. L.M.Y. wrote the applications for the grants that funded the study and designed the investigation. L.M.Y. is the guarantor of this work and, as such, had full access to all the data in the study and takes the responsibility for the integrity of the data and the accuracy of the data analysis.
