Acceptability of and satisfaction with continuous glucose monitoring in pregnant women with type 2 diabetes mellitus: A service improvement project

Background

Type 2 diabetes mellitus (T2DM) accounts for 54% of women entering pregnancy with diabetes, doubling in the last two decades. ¹ Women with T2DM face additional healthcare inequalities and are frequently not prepared for pregnancy. In the UK, women with T2DM are more likely to be of non-White Ethnic groups and live in areas of social deprivation. ¹ The rate of neonatal deaths is also higher in T2DM mothers than in those with type 1 diabetes (T1DM). ² T2DM, poor glycated haemoglobin (Hba1c) control, and social deprivation are independent risk factors for perinatal death. Based on the findings of the CONCEPPT trial in women with T1DM in pregnancy, ³ NICE have recommended the use of continuous glucose monitoring (CGM). ⁴ The National audit report highlighted the need to evaluate whether Libre or intermittent scanning CGM (isCGM) use could improve levels of glucose control, including target HbA1c attainment, and ultimately affect neonatal outcomes in pregnant women with T2DM. ¹

Many women with T2DM are inexperienced in performing blood glucose monitoring when they become pregnant, with several of our patients reporting they found capillary fingerstick testing mentally and physically burdensome. ⁵ Compliance rates with finger stick monitoring can be variable, with many women not performing the recommended minimum amount of daily readings by NICE (four readings per day, including fasting and one-hour post-breakfast, lunch, and dinner). Poor compliance with monitoring makes it challenging to tailor treatment to individual needs, likely leading to undertreatment.

CGM could offer an alternative and improve compliance and engagement with diabetes care. The first step in evaluating the use of CGM in pregnant women with T2DM is to determine the acceptability of the technology. However, few studies have explored the role of isCGM for women with T2DM in pregnancy. It is not known whether this technology can improve compliance and engagement with glucose monitoring during pregnancy, or what the optimal time in range for women with T2DM might be in order to improve perinatal outcomes.

As part of a clinical service improvement project to improve glucose monitoring compliance in women with T2DM, we sought to determine the feasibility and acceptability of using isCGM in this population.

Methods

A service improvement project was conducted within the Oxford University Hospitals (OUH) NHS Foundation Trust, UK. The OUH provides maternity care across two sites (the John Radcliffe Hospital in Oxford and Horton Hospital in Banbury) for around 7500 pregnancies each year. Of these, about 50 are pregnant people with type 2 diabetes. The Trust guideline for the management of T2DM in pregnancy is in line with NICE recommendations for diabetes in pregnancy. ⁴ This project was aimed at women with diagnosed T2DM prior to pregnancy or meeting the criteria for T2DM during pregnancy. The inclusion and exclusion criteria are illustrated in Table 1.

OPEN IN VIEWER

Table 1.

Participant inclusion and exclusion criteria.

Inclusion criteria	Exclusion criteria
T2DM known pre-pregnancy or revealed in pregnancy with HbA1c ≥ 48 mmol/mol (6.5%)	Hypersensitivity to Freestyle Libre® device adhesive
18 years and above	Not consenting to CGM data share
Booked and planning to deliver at OUH NHS Foundation Trust	Insufficient English language comprehension
At least 20 weeks’ gestation
Demonstrated ability to use finger stick glucose monitoring

T2DM: type 2 diabetes mellitus; HbA1c: haemoglobin A1c; OUH: Oxford University Hospitals; CGM: continuous glucose monitoring.

The project lead (SB) approached all women meeting the inclusion criteria in the pregnancy and diabetes specialist clinic held at the John Radcliffe Hospital. Participation was voluntary and non-renumerated. If women agreed to participate, they were provided with information related to the Libre device system, as well as the sensors and reader. The majority of women who chose not to take part disliked the thought of having to wear a sensor all day. This was unrelated to ethnicity or whether or not they did not speak English. There was good participation from women who identified as being from different ethnic groups. Prior to enrolment, eligible women were encouraged to use the GDm-Health^TM app on their smartphones to record fasting and postprandial values. Training on how to use the Libre device was provided both face-to-face and virtually, with help from existing educational tools from the manufacturers (https://pro.freestyle.abbott/ie-en/home/scientific-resources-education/patient-education.html).

Participants were asked to perform all glucose testing by isCGM. Results were recorded in one of three ways: (1) by manually entering fasting and 1-hour post-prandial glucose readings into the GDM-Health App^TM, ⁶ (2) by recording the same values on a paper record, or (3) by sharing data through the Libre View App^TM where time in the recommended pregnancy range (3.5–7.8 mmol/mol) could be assessed. The results were remotely monitored and reviewed up to three times per week by a member of the diabetes clinical team. Participants attended antenatal multidisciplinary clinics every fortnight, with monthly ultrasound scans through the third trimester as per the existing Trust care pathway.

The primary outcomes for this project were feasibility and acceptability. Feasibility was determined by the number of participants who continued to use the device until delivery. Acceptability and satisfaction with the system were assessed through a structured questionnaire completed by participants during pregnancy or within four weeks of delivery: the validated Oxford Maternity Diabetes Treatment Satisfaction Questionnaire (OMDTSQ) was used as the questionnaire. ⁷ The OMDTSQ comprises 10 questions designed to assess general satisfaction with diabetes care, the acceptability and reliability of the technology, and the perceived relationship with the diabetes care team. Only nine questions were applied to the present study. Participants were asked to score their agreement with statements on a 7-point Likert-type scale, ranging from +6 (strongly agree) to −0 (strongly disagree). The OMDTSQ was plainly worded in English. At the end of the questions, women were encouraged to write free-text responses to provide additional feedback and/or suggestions. ⁷ The questionnaire was sent to participants electronically, with paper copies available in the clinic if preferred. At the end of the questionnaire, free text responses were captured about any aspects of their glucose monitoring or diabetes care. All questionnaires were anonymised and free text comments were analysed using thematic analysis. Information on individual characteristics and birth outcomes from the hospital electronic patient record (Cerner Millennium^TM).

Glucose control in the third trimester was determined by mean glucose in range on the Libreview App or on the GDM Health App. For those using Libreview, we used the T1DM CGM target of 70% in the pregnancy glucose range (3.5–7.8 mmol/L). For participants recording fasting and one-hour post-prandial interstitial glucose levels, we defined meeting target as not having more than three out-of-range readings per week.

Given the small sample size, descriptive baseline and delivery data are presented only, with no inferential statistics.

Results

Between March 2022 and April 2023, 48 pregnant women with T2DM were managed through our Trust. Of these, 23 women met the entry criteria and agreed to participate in the study, with 20 using the device until delivery (Figure 1). At follow-up, one participant dropped out of the project, and two stopped using isCGM and switched back to finger sticking for self-monitoring. A total of 22 participants answered the qualitative feedback survey between 28 and 36 weeks’ gestation. Whilst 14 women were in a second or subsequent pregnancy with type 2 diabetes, all participants were using the isCGM device for the first time.

OPEN IN VIEWER

Figure 1.

CONSORT flow diagram of patients eligible, recruited, numbers followed-up and included in analysis. Abbreviations: number of participants (n); intermittently scanned continuous glucose monitoring (isCGM).

Table 2 displays the baseline characteristics of participants and their pregnancy outcomes. Most women were overweight or obese, defined as having a body mass index (BMI) of 25–30 and over 30 kg/m², respectively. The BMI for the entire sample ranged from 24 to 45 kg/m² with only one participant with BMI less than 25 kg/m². Almost a third of participants developed a hypertensive complication of pregnancy, and most delivered by caesarean birth. There were no perinatal deaths.

OPEN IN VIEWER

Table 2.

Baseline characteristics of the 23 participants.

Participant characteristics	N (%)
Age range	23–42 years
BAME ethnicity, n (%)	10 (43%)
Parity, n
Primipara	2
Multipara	21
BMI, range (kg/m2)	25–45 kg/m2
Normal BMI (18.5–24.5 kg/m2)	4 (17%)
Class I and II obesity BMI (30–39.9 kg/m2)	13 (56%)
Class III obesity BMI (≥ 40 kg/m2)	6 (27%)
HbA1c, range % or mmol/mol	5.4%–12.2% or 35.5–109.8 mmol/mol
HbA1c < 6% or < 42.1 mmol/mol, n	12
HbA1c 6.1–6.5% or 43.2–47.5 mmol/mol, n	2
HbA1c > 6.5% or > 47.5 mmol/mol, n	9
Hypertension in pregnancy, n (%)	8 (34%)
Pre-existing	1
Pregnancy-induced	7
Mode of delivery, n (%)
Spontaneous vaginal delivery	0
Assisted vaginal delivery	2 (9%)
Elective caesarean section	10 (43%)
Emergency caesarean section	11 (48%)
Neonatal admission < 24 h, n (%)	12 (52%)

n: number of participants; BAME: Black, Asian, and Minority Ethnic; BMI: body mass index; GDM: gestational diabetes mellitus; T2DM: type 2 diabetes mellitus; HbA1c: haemoglobin A1c.

Satisfaction with Libre 2 CGM monitoring in pregnancy

Table 3 summarizes the qualitative feedback received through the questionnaires. Most participants were very satisfied with their care. Sixteen participants reported they were very satisfied with the convenience of CGM equipment for monitoring blood glucose. Additionally, 15 women found that the equipment was very reliable. There was less agreement over whether the type of monitoring fitted in well with daily activities throughout pregnancy, with around half of participants agreeing with this statement. Of the 14 participants who had experienced a previous pregnancy affected by T2DM, 11 rated their glucose monitoring experience as much better compared to that of previous pregnancies. Most participants expressed that they would be interested in using the CGM device for continuous blood glucose monitoring in future pregnancies, while 19 would recommend the device to friends and/or family also affected by T2DM irrespective of pregnancy status.

OPEN IN VIEWER

Table 3.

Summary of quantitative patient feedback.

Patient satisfaction	RATING/6 n (%) (total n = 22, except for Q10, where n = 14)
	0	1	2	3	4	5	6
Q1. Satisfaction with T2DM treatment	0 (0%)	0 (0%)	0 (0%)	0 (0%)	1 (4.7%)	1 (4.7%)	20 (95.7%)
Q2. Satisfaction with treatment being best for individual needs	0 (0%)	0 (0%)	0 (0%)	0 (0%)	0 (0%)	2 (9.5%)	20 (95.7%)
Q3. Team's understanding of diabetes control	0 (0%)	0 (0%)	0 (0%)	1 (4.7%)	0 (0%)	3 (14.3%)	18 (81.8%)
Q4. Relationship with team	0 (0%)	0 (0%)	0 (0%)	0 (0%)	2 (9.5%)	1 (4.7%)	19 (86.4%)
Q5. Convenience	1 (4.7%)	0 (0%)	0 (0%)	0 (0%)	2 (9.5%)	3 (14.3%)	16 (72.7%)
Q6. Reliability	1 (4.7%)	1 (4.7%)	0 (0%)	0 (0%)	3 (13.6%)	2 (9.5%)	15 (71.4%)
Q7. Applicability to lifestyle	1 (4.7%)	1 (4.7%)	0 (0%)	2 (9.5%)	2 (9.5%)	2 (9.5%)	14 (63.6%)
Q8. Omitted from results. Included for the purposes of control only
Q9. Satisfaction with team's understanding of individual diabetes	0 (0%)	0 (0%)	0 (0%)	0 (0%)	0 (0%)	2 (9.5%)	20 (95.7%)
Q10. Glucose monitoring compared to previous pregnancies with T2DM	1 (7.1%)	0 (0%)	0 (0%)	1 (7.1%)	0 (0%)	1 (7.1%)	12 (54.5%)

Question eight was omitted from the results and was included in the study for control purposes only. The total number of respondents to the survey was 22. n: number; Q: question; T2DM: type 2 diabetes mellitus.

Emergent themes of qualitative open feedback

Written comments supported the scores. Four themes were identified around positive experience with the technology (1) benefits of less finger pricking, (2) perceived easier control of hypoglycaemic events, (3) feeling that they had better care during pregnancy, and (4) recommending it for self, family, and friends. The three main areas where challenges were experienced were (1) poor alarm performance, (2) reliability around upper and lower glucose ranges, and (3) sensor stickiness to the skin. The themes are summarized in Table 4.

OPEN IN VIEWER

Table 4.

Themes of positive and negative experiences with supporting comments.

Theme (positive)
Benefits of reduced capillary glucose monitoring	‘Love not having to prick finger 8 times per day’. (Participant 18)‘No problem with device and comfortable’. (Participant 18)
Easier control of hypoglycaemic events	‘It has worked really well with Hypos’. (Participant 7)‘Whilst being pregnant I experienced lows during the night more frequently than before and the alarm woke me and enabled me to control it easier’. (Participant 2)
Better care during pregnancy	‘Couldn't ask for better care and monitoring’. ‘Highly recommend Libre 2 device’. (Participant 1)‘Helped with my personality disorder and helped with allaying anxiety compared to last pregnancy’. (Participant 20)
Recommending to friends and family	‘I found my blood sugars much more manageable with the Libre 2 device and would highly recommend the continued use for all type 1 and type 2 diabetics whether pregnant or not. I already have recommended the Libre 2 to my mother’. (Participant 2)
Theme (negative)
Poor Libre 2 alarm performance	‘Kept giving false hypos during night and day checked with finger prick and wasn't having a hypo’. (Participant)
Reliability around extreme blood glucose ranges	‘Many hypers have gone unnoticed. It has worked well with hypos. Overall happy with Libre’. (Participant 5)
Sensor stickiness	‘Doesn't stick on skin for very long and sometimes bruises quite bad’. (Participant 6)‘Fall of device was troublesome’. (Participant 19)‘I think the sensors would work better if they stuck more (for me it wasn't very good at staying in place), maybe need proper stickers for them’. (Participant 21)

Discussion

We demonstrate the feasibility and acceptability of isCGM among women with T2DM in the second and third trimesters of pregnancy in our population. We note that some women living with obesity reported adhesion problems with the sensors and concerns about device accuracy in the extremes of glucose. Overall, however, the response from participants was extremely positive towards using this technology. We believe this is a first step towards including isCGM as part of clinical care for women with T2DM in pregnancy.

The women participating in the study were representative of our clinic population, with 10 participants (43%) being of Black, Asian, or other minority ethnicity, and 83% being obese. Overall, an acceptable time in the glucose target range was achieved by the majority of women using isCGM, with 16 participants achieving a time in the range of more than 70% in the third trimester. This qualitative study has shown that women with T2DM found isCGM management using the Libreview 2 device system acceptable, suitable, and convenient. Based on the feedback provided, most women felt the technology was more convenient than finger sticking. They also felt well-supported and very satisfied with their care. There were no major barriers to uptake. However, reliability at extreme ranges of glucose and device sensor falling off was a noticed theme. This correlates with issues identified in other published research.^8–10 In this study, more than two-thirds of women continued to use the device for glucose monitoring, suggesting high acceptability.

CGM is an established method for comprehensive glucose data and is convenient for the patient. Our study suggests that CGM is acceptable to women with T2DM during pregnancy and feasible in the context of intensive monitoring required in the management of such high-risk pregnancies. However, several questions remain unanswered with regards to CGM use in pregnant patients with T2DM, namely (1) whether CGM is a better technique for women who have entered pregnancy mostly without the ability to even self-monitor their blood glucose, (2) whether CGM is an answer for women who struggle with escalating from no glucose monitoring to the intensive management required in pregnancy to prevent poor outcomes, and (3) whether CGM can improve maternal and fetal outcome in high-risk and unplanned pregnancies.

Conclusion

The result of this study encourages the progression in researching the role of CGM for the management of T2DM in pregnancy to improve women's experience and glucose monitoring compliance, which is strongly correlated with maternal and fetal outcomes. We propose that future research could expand this project as part of a multicentre study on the use of CGM for T2DM in pregnancy and also explore the correlation between the use of CGM in the management of T2DM in pregnancy and outcomes in a variety of populations. A detailed cost analysis, including the costs of equipment, staff, and infrastructures are essential in strengthening the candidature of CGM as an alternative way to self-monitoring blood glucose in high-risk pregnancies.

Supplemental Material

Supplemental Material