Comparison of a smartphone application blood glucose management system with standard clinic care in women with gestational diabetes mellitus

Introduction

Gestational diabetes mellitus (GDM), defined as any degree of glucose intolerance with onset or first recognition during pregnancy,¹ is an increasingly common condition complicating pregnancy. In Australia GDM affects about 12-14% of pregnant women,² almost tripling from 5% in 2005-2007.³ Although the prevalence of GDM may vary from 1.7% to 20% depending on diagnostic criteria and maternal characteristics,⁴ the number of women being diagnosed with GDM continues to increase at an annual rate of 1%.⁵ Contributing factors likely include an increase in the age and weight of pregnant women, the change in diagnostic criteria and also the changing ethnic makeup of Australian society.⁶ The incidence of GDM in the Fairfield Local Government Area has been listed in the top ten for NSW.⁷ After adopting the new World Health Organization Diagnostic Criteria,⁸ Fairfield Diabetes Clinic experienced a 21% increase in new GDM referrals and a 20% increase in total occasions of service. This resulted in significant challenges providing both maternity and diabetes services as women with GDM require close monitoring and support to manage their condition. In the current study setting, standard practice is for women with GDM to monitor and record BGLs four times a day with once or twice weekly reviews in the clinic, depending upon their glycaemic management. Many women are diagnosed with GDM early in their second trimester, resulting in them attending 10-14 diabetes-related clinic visits during their pregnancy. With limited resources to support the increasing number of women, a new model of care (MOC) that did not compromise maternal or neonatal outcomes, was acceptable to women and which could improve service delivery was required. To this end, a smartphone application with Bluetooth-enabled meter which allowed glucose values to be transmitted in real-time to a secure website with clinician review and feedback was implemented.

Mobile/Smartphones have portability, internet connectivity and increasing capacity to run complex applications that make them an ideal tool for health services to collect personal information (with informed consent), provide personalised health interventions, and potentially save time and cost when compared with standard healthcare^9&10. Research has shown mobile/smartphone technology to be a successful tool for the self-management of type 1 and type 2 diabetes.^11&12 They also hold promise as a beneficial tool to facilitating healthcare practice including assisting with information and time management, sharing health records, delivery of health information to healthcare professionals, real-time monitoring of patient vital signs and direct provision of care. ^{10, 11&12}

This paper reports comparative data on a range of maternal and neonatal outcomes and health service utilisation for women who used a smartphone app to support GDM care and those who received standard care (clinic visits).

Methods

Using a comparative study design, women who attended initial GDM education at Fairfield Diabetes Service between October 2016 and March 2017 comprised the standard care group, while women who attended the service between April and September 2017 and who agreed to use the smartphone app comprised the new MOC group. Women’s demographic, clinical and health service utilisation data were extracted from medical and clinic records and neonatal outcomes were extracted following delivery.

New MOC: women with GDM monitored their BGLs using the Accu Chek Guide meter which connected to a smartphone app (Accu Chek Connect) via wireless Bluetooth technology.¹⁴ A custom designed application on the smartphone receives the BGLs and then automatically uploaded data to a secure server that linked with the diabetes clinic, allowing relevant clinicians to review these data remotely on a secure website and respond quickly to out of range BGLs. A designated Diabetes Nurse Educator (DNE) was responsible for reviewing women’s BGLs on the secure website at least once weekly and women only needed to attend clinic appointments with the diabetes team every four weeks until 36 weeks gestation. If there were any concerns on the part of the staff or the women, they were advised to contact or make an appointment earlier with the diabetes team.

Data were analysed using descriptive and comparative statistics; Pearson’s Chi- Square test for categorical variables and Student t-tests for continuous variables. A p-value of <0.05 indicated statistical significance.

This study was conducted evaluating the new MOC as part of the routine clinical care service provided. Therefore, approval from the South Western Sydney Local Health District Research Ethics Committee was not required. The participants in new MOC provided informed consent by inviting the diabetes clinic to view their clinical data via email invitation.

Results

A total of 435 women were included in this study; 272 in the standard care group and 163 in new MOC group. A comparison of baseline characteristics is presented in Table 1. There were differences between the two groups, with more women in the standard care group requiring health care interpreters (43.2 vs 6.5%, p < 0.001) and having higher 2-hour plasma glucose levels following the 75g oral glucose tolerance test (8.1 ± 1.6 vs. 7.7 ± 1.6 mmol/L; p = 0.025). At baseline, women in the new MOC group had a higher body mass index (BMI) than women in the standard care group (27.3 ± 6.8 vs. 25.9 ± 6.2; p = 0.025).

Table 1: Comparison between the two groups: background characteristics

Women in the new MOC group attended fewer clinic appointments than women in the standard care group (6.4 ± 2.9 vs. 7.3 ± 3.6, p = 0.009). Even after excluding women requiring interpreters, those in the new MOC group still had fewer clinic appointments then the standard care group (6.3 ± 2.9 vs 7.1 ± 3.8; p = 0.038). However, the number of failure to attend (FTA) appointments was significantly higher in the new MOC group (1.2 ± 1.9 vs. 0.6 ± 1.3; p = 0.001) (Table 2).

Table 2: Attendance at diabetes service

No statistically significant differences were observed between groups for pregnancy or neonatal outcomes, including mode of delivery, admission to special care nursery, birthweight, neonatal hypoglycaemia and perinatal death (Table 3).

Table 3: Birth outcomes

Discussion

With the increasing burden of GDM leading to the implementation of innovative MOCs it is important to ensure maternal and neonatal outcomes are at least equivalent to those of women who receive standard care, which to date has been frequent face-to-face clinic visits. Our new MOC using a smartphone application provided clinicians with access to women’s real-time glucose values which in turn allowed them to respond quickly to out of range BGLs. This has the potential to deliver high quality healthcare that is also more convenient for women who are required to attend fewer clinic visits. This may also mean less time off work, and reduced transport and parking costs. Many women were diagnosed with GDM quite early in the second trimester, with standard care requiring them to attend clinic appointments every 1-2 weeks. This can be very challenging, especially for those who work or have to take care of young children at home.

Although our results do not demonstrate that smartphone technology was superior to standard care for women with diabetes in pregnancy in terms of birth outcomes, there was no evidence of harm, a finding supported in a systematic review of telemedicine technologies for pregnancy in diabetes.¹³ Our new MOC is a person-centred initiative promoting collaboration between women with GDM and healthcare professionals to improve health outcomes while improving user satisfaction. Overall, the feedback has been overwhelmingly positive, with comments including “I like the care factor—appreciate someone is there for you” and “Makes you more accountable knowing someone is watching.”

Our new MOC significantly reduced clinic visits without increased adverse GDM-related clinical outcomes. The FTA rates were higher in the new MOC group which will need to be monitored in the future. We believe some of this could be related to errors with the clinic bookings, as fortnightly appointments had been the usual appointment schedule for women with GDM and it may have taken some time for the clinic clerks to adjust to the longer follow-up interval. On the other hand, it was possible that these women were less vigilant with attending clinic appointments given the bulk of their management was not face-to-face. It may also be that the women felt more confident knowing their BGLs were being assessed by clinicians and therefore, may have had a lower threshold for missing some scheduled appointments.

In this study, we found no differences in pregnancy or neonatal outcomes between the two groups. This is somewhat reassuring that these women are not disadvantaged by this new MOC. While at the same time this new MOC could reduce the resource burden on the health care facility as well as improving clinic flow. It also gives women with GDM an alternative MOC that may potentially improve their satisfaction. Furthermore, this new MOC may be useful for remote centres where women have difficulty with clinic attendance.

There were several limitations to our study. First, data were collected retrospectively, which may have impacted on the quality of the data. There were few women from culturally and linguistically diverse (CALD) backgrounds who participated in the new MOC, and hence the outcomes for this group may be different. The new MOC was also only available for women who owned a smart phone which may have impacted findings. We did not assess the burden of BGL evaluation through the portal and phone on nursing resources and this may have impacted on cost effectiveness. We also did not formally assess client and clinician satisfaction with this new MOC, which would need to be ascertained in future studies. In addition to addressing the limitations identified in this study, future studies should measure additional outcomes including the percentage of BGLs within target, the number of missing readings and the tagging of other information like meals, exercise and diabetes medications.