Using diabetes technology in pregnancy

BEST PRACTICE GUIDE:

Using diabetes technology

in pregnancy

Image used with permission: Dexcom

BEST PRACTICE GUIDE: USING DIABETES TECHNOLOGY IN PREGNANCY Version 2.0, March 2020

CONTENTS

Contributors 3

Acknowledgements 3

Foreword 4

Objective 5

Introduction 5

Insulin pump therapy (CSII) before and during pregnancy 7

Evidence for insulin pump therapy (CSII) before and during pregnancy 7

Indications for insulin pump therapy (CSII) before and during pregnancy 7

Starting insulin pump therapy (CSII) in pregnancy 8

Using insulin pump therapy (CSII) in pregnancy 9

Management of unexplained hyperglycaemia for insulin pump (CSII) users in pregnancy 11

Figure 2: Rules for managing unexplained hyperglycaemia in pregnancy for insulin pump users 12

Sick day rules for insulin pump (CSII) users in pregnancy 13

Figure 3: Sick day rules in pregnancy for insulin pump users 14

Using glucose sensors before and during pregnancy 15

Evidence for RT-CGM before and during pregnancy 15

Indications for RT-CGM before and during pregnancy 16

Information provided by RT-CGM 16

RT-CGM set up in pregnancy 17

Using RT-CGM in pregnancy 19

Figure 4: Guidance for avoiding and treating hypoglycaemia 22

A note on hybrid closed loop systems in pregnancy 27

Evidence for flash glucose monitoring before and during pregnancy 30

Indications for flash glucose monitoring before and during pregnancy 30

Information provided by flash glucose monitoring 30

Flash glucose monitoring set-up in pregnancy 30

Using flash glucose monitoring in pregnancy 30

Using sensor data to improve glycaemic control in pregnancy: diet, activity and insulin considerations 32

Options for managing post-meal glucose excursions 32

Options for managing overnight glucose 35

An approach to the consultation and looking at downloads with pregnant women using diabetes technology 37

The first consultation in pregnancy 37

Issues to actively review periodically in pregnancy 38

Looking at the download 38

Using diabetes technology in particular circumstances in pregnancy 41

Using diabetes technology in hospital 41

Managing steroid induced hyperglycaemia using CSII 41

Using diabetes technology before and during birth 42

Figure 5: Protocol for managing glucose levels for women continuing on insulin pump therapy during

labour and birth or during casearean birth. 45

Using diabetes technology after birth 46

Management immediately after birth until discharge from hospital 46

Management in the postpartum period 47

Service considerations for supporting use of diabetes technology in pregnancy 50

Workforce requirements 50

Capacity and organisation 51

Pathways and programmes 55

System choice, informatics and data requirements 55

References 57

Appendix 1: Carbohydrate Choices 59

BEST PRACTICE GUIDE: USING DIABETES TECHNOLOGY IN PREGNANCY Version 2.0, March 2020

CONTRIBUTORS

Writing group

Caroline Byrne (Diabetes Specialist Nurse, Cambridge University Hospitals NHS Foundation Trust)

Jeannie Grisoni (Diabetes Dietitian, Cambridge University Hospitals NHS Foundation Trust)

Peter Hammond (joint lead) (Consultant Endocrinologist, Harrogate District Hospital and Leeds GIS; Head of

School of Medicine, Health Education England - Yorkshire and the Humber) pet[email protected]

Kate (Katharine) Hunt (joint lead) (Diabetes Consultant, King’s College Hospital NHS Foundation Trust; Visiting

Lecturer, King’s College London) kat[email protected]

Helen Murphy (Professor of Medicine Norwich Medical School; Professor of Women’s Health, Kings College

London; Honorary Consultant Physician, Cambridge University Hospitals NHS Foundation Trust)

Helen Rogers (Diabetes Nurse Consultant, King’s College Hospital NHS Foundation Trust)

Reviewing group

Valerie Ashton (patient representative)

Anna Brackenridge (Diabetes Consultant, Guy’s and St Thomas’ NHS Foundation Trust)

Pratik Choudhary (Senior Lecturer, King’s College London; Diabetes Consultant, King’s College Hospital NHS

Foundation Trust; and Chair, ABCD Diabetes Technology Network UK)

Holly Davies (patient representative, and JDRF, the type 1 diabetes charity)

Rachel Earp (Diabetes Specialist Midwife, East and North Hertfordshire NHS Trust)

Robert Lindsay (Reader in Diabetes & Endocrinology, University of Glasgow)

Lisa Long (Consultant Obstetrician, King’s College Hospital NHS Foundation Trust)

David McCance (Consultant Physician/Honorary Professor of Endocrinology, Belfast Health and Social Care Trust)

Jenny Myers (Consultant Obstetrician, St Mary’s Hospital, Manchester)

Julia Platts (Diabetes Consultant, University Hospital Llandough Cardiff, National Clinical Lead for Wales for

Diabetes)

Eleanor Scott (Professor of Medicine (Diabetes and Maternal Health), University of Leeds, and Leeds Teaching

Hospitals NHS Trust)

Emma Wilmot (Diabetes Consultant, University Hospitals of Derby and Burton NHS Foundation Trust; previous

Chair, ABCD Diabetes Technology Network UK)

Acknowledgements

We would also like to thank the many women with type 1 diabetes who have used diabetes technology in their

pregnancies and who continue to help us learn how best to use it.

ABCD DTN-UK would like to thank their 2019 sponsors. Without their support, the development and publication

of this document would not have been possible.

Đ Gold sponsors: Abbott Diabetes Care, Roche Diabetes Care, Ypsomed Ltd

Đ Silver sponsors: Advanced Therapeutics UK Ltd, Dexcom, Medtronic Ltd

Đ Bronze sponsors: Insulet International Ltd

BEST PRACTICE GUIDE: USING DIABETES TECHNOLOGY IN PREGNANCY Version 2.0, March 2020

FOREWORD

Antenatal care for women with diabetes has and continues to be a cornerstone of all diabetes

services around the NHS. The St Vincent’s declaration in 1989 was testament to the desire of all

relevant health care professionals to try and ensure outcomes for those with diabetes were at

par with those who didn’t.

Yet in spite of many years of efforts, work and development of antenatal services, neonatal

death, stillbirth, congenital anomaly, large and small for dates babies and neonatal unit

admission remain high by comparison with pregnancies in woman without diabetes and are of

considerable concern.

Some system wide changes which may help could be:

• Targeted and improved patient education and support around contraception and

pregnancy preparation with a focus on identifying those at highest risk.

• Greater empowerment of women to make routine diabetes self-management decisions.

• Increased awareness and training for all healthcare professionals.

• Development and implementation of new pathways for identification, referral

and treatment.

One of the roles of NHS England is to look at emerging research and work towards getting this

to the population at pace- to help translate the benefits shown in clinical trials to a wider patient

population. The CONCEPTT trial-led by Professor Helen Murphy and involving several members

of the working group for this guide -showed the benefit of using Continuous Glucose Monitoring

in pregnant women with type 1 diabetes- with impressive outcomes from a neonatal perspective.

As part of the drive to increase access to relevant technology and improve neonatal outcomes, the

NHS Long term Plan 2020 committed to all relevant patients having access to CGM.

As part of the implementation of this, NHS England have been working closely with the

Diabetes Technology Network to tackle the issue of education in an effort to minimise variation

in the delivery of this ambitious plan. It has been a delight to see this guide being produced

which hopefully will help in informing colleagues about the use of technology in the setting of

pregnancy in diabetes.

My thanks to all the authors and collaborators and especially to Drs Kate Hunt and Peter

Hammond who led in delivering this excellent piece of work- which hopefully will go a long way

to improving pregnancy outcomes in women with type 1 diabetes across the NHS.

Professor Partha Kar

National Specialty Advisor, Diabetes, NHS England

BEST PRACTICE GUIDE: USING DIABETES TECHNOLOGY IN PREGNANCY Version 2.0, March 2020

OBJECTIVE

This document aims to provide healthcare professionals with UK expert consensus on the best practice for

starting, managing and optimising insulin pump therapy (continuous subcutaneous insulin infusion (CSII)) and

glucose sensors (real-time continuous glucose monitoring (RT-CGM) and flash glucose monitoring) in pregnancy.

Introduction

Pre-pregnancy diabetes is associated with increased risk of a range of adverse pregnancy outcomes (NICE 2015).

Excellent glycaemic control throughout pregnancy is associated with reduced risk of these adverse outcomes

(NICE 2015), but this is very difficult to achieve. Current recommended targets are:

• Fingerstick capillary blood glucose (CBG) targets (NICE 2015):

› Fasting CBG <5.3 mmol/l

› 1 hour after meals CBG <7.8 mmol/l

› 2 hours after meals CBG <6.4 mmol/l

› And to maintain CBG >4 mmol/l

• Continuous glucose monitoring (CGM) targets (figure 1) (Battelino, Danne et al. 2019):

› Sensor glucose 3.5-7.8 mmol/l at least 70% of the time (>16 h 48 min per day)

› Sensor glucose >7.8 mmol/l less than 25% of the time (<6 h per day)

› Sensor glucose <3.5 mmol/l less than 4% of the time (<1 h per day)

» including sensor glucose <3.0 mmol/L less than 1% of the time (<15 min per day)

Diabetes technology may help improve glycaemic

control. Diabetes technology currently refers to insulin

pump therapy (continuous subcutaneous insulin

infusion (CSII)), glucose sensors (real-time continuous

glucose monitoring (RT-CGM) and flash glucose

monitoring) and the interaction between them. Closed-

loop insulin delivery is not included in this guide as this

is not routinely available for clinical use in pregnancy.

More detailed information for evidence outside

pregnancy is provided in the parallel DTN-UK Best

Practice Guides (https://abcd.care/dtn-uk-best-practice-

guides). There is evidence that RT-CGM versus usual

care in pregnancy improves neonatal outcomes (Feig,

Donovan et al. 2017).

NICE guidelines recommend that CSII and/or CGM can

be considered in pregnancy (CGM advice is currently

under review) (NICE 2015) (https://www.nice.org.

uk/guidance/ng3). The NHS Long Term Plan (www.

longtermplan.nhs.uk/) launched in January 2019 states:

‘The NHS will ensure that, in line with clinical guidelines, patients with type 1 diabetes benefit from life changing

flash glucose monitors from April 2019, ending the variation patients in some parts of the country are facing.

In addition, by 2020/21, all pregnant women with type 1 diabetes will be offered continuous glucose monitoring,

helping to improve neonatal outcomes’

Figure 1: Sensor glucose targets for pregnant women

with type 1 diabetes

TIME

<25%

>70%

<4%

>7.8 mmol/l

Target range:

3.5-7.8 mmol/l

<3.5 mmol/l

<3.0 mmol/l (<1%)

Sensor glucose target range

3.5-7.8 mmol/l

Each day aim for sensor glucose:

• In target range (3.5-7.8 mmol/I)

at least 70% of the time

(>16h 48 min per day)

• Above target (>7.8 mmol/I) less

than 25% of the time (<6 h per day)

• Below target (<3.5 mmol/I) less

than 4% of the time (<1 h per day)

› including <3.0 mmol/I less than

1% of the time (<15 min per day)

Adapted from Battelino et al 2019

BEST PRACTICE GUIDE: USING DIABETES TECHNOLOGY IN PREGNANCY Version 2.0, March 2020

For current diabetes technology to be used effectively and safely in pregnancy, it is important that both women and

their diabetes health professionals have a practical understanding of its strengths and limitations. Many of the

principles are the same as using diabetes technology outside of pregnancy. However, there are several challenges

particular to pregnancy, where the risk of hypoglycaemia in the mother has to be balanced against the risks of

hyperglycaemia to the fetus.

• Glycaemic targets are tighter than outside pregnancy.

• Time is critical as hyperglycaemia during early pregnancy is associated with increased risk for congenital

anomaly while hyperglycaemia in the second and third trimesters is associated with fetal growth acceleration,

preterm birth and neonatal complications.

• Postprandial glucose is targeted with the aim of minimising postprandial hyperglycaemia.

• Minimising postprandial hyperglycaemia carries a risk of delayed postprandial hypoglycaemia due to the tail of

the meal bolus, particularly before 20 weeks of gestation.

• Insulin absorption is increasingly delayed and with more day to day variability in late pregnancy (Goudie, Lunn

et al. 2014).

• Insulin requirements change markedly through pregnancy and after birth.

• Increased risk of hypoglycaemia, particularly in the late first and early second trimester and after birth.

Approximately 10% of pregnant women with type 1 diabetes have at least one hospital admission due to

hypoglycaemia (NPID).

• Increased risk of diabetic ketoacidosis (DKA), as pregnancy is a ketogenic state. 2.7% of pregnant women with

type 1 diabetes have at least one admission with DKA (NPID). DKA in pregnancy carries a risk of fetal death (16-

27% or higher) (Schneider, Umpierrez et al. 2003, Morrison, Movassaghian et al. 2017).

• Pregnant women with diabetes are usually highly motivated. However, the psychological impact and time

commitment required for women with diabetes to achieve and sustain their best possible glycaemic control

when preparing for pregnancy, throughout pregnancy and birth and then adjusting after birth should not

be underestimated.

A recurrent theme which limits access to diabetes technology is health professional time and training (White,

Goenka et al. 2014). This guide has been developed with the aim of sharing best practice from across the UK to

inform those who support women with diabetes using diabetes technology in pregnancy, or who would like to

provide this service. A multidisciplinary team of health professionals and women with diabetes provided input

into this guide. It is our hope that, by providing clear clinical and service pathways, this document will support

staff to deliver safe, effective and high-quality care. In parallel with this guide, the Diabetes Technology Network

is supporting the development of a series of online modules on use of glucose sensors in pregnancy, available at

https://abcd.care/dtn/education

BEST PRACTICE GUIDE: USING DIABETES TECHNOLOGY IN PREGNANCY Version 2.0, March 2020

INSULIN PUMP THERAPY (CSII) BEFORE AND DURING

PREGNANCY

Insulin pump therapy (continuous subcutaneous insulin infusion,

CSII) employs a battery operated, portable, programmable pump

to continuously deliver rapid-acting insulin via an infusion set

inserted subcutaneously (ABCD-DTN-UK 2018). The basal insulin

infusion rate can be varied and programmed in advance. Bolus

doses are given via the pump for meals or corrections and most

pumps incorporate bolus calculators.

Evidence for insulin pump therapy (CSII)

before and during pregnancy

Insulin pump therapy is more effective than multiple daily

injection (MDI) therapy in lowering HbA1c in non-pregnant

individuals and with a lower risk of hypoglycaemia (Pickup and

Sutton 2008, ABCD-DTN-UK 2018).

There are no randomised controlled trials (RCTs) comparing modern CSII and MDI therapies in pre-pregnancy or

established pregnancy (Rys, Ludwig-Slomczynska et al. 2018). The four RCTs were small and all conducted before

1990 (Rys, Ludwig-Slomczynska et al. 2018). Evidence is mainly derived from observational studies with the

associated limitations, including that glycaemic control may have influenced choice of CSII or MDI (Rys, Ludwig-

Slomczynska et al. 2018). Meta-analysis of observational studies reported lower first trimester HbA1c in women who

started CSII prior to pregnancy, compared to those on MDI (Rys, Ludwig-Slomczynska et al. 2018). Meta-analyses

of mostly observational studies also reported lower HbA1c in the second trimester in women using CSII, with no

significant difference in the third trimester HbA1c, and higher gestational weight gain (Rys, Ludwig-Slomczynska

et al. 2018). There is evidence that second and third trimester insulin dose adjustment is not sufficiently aggressive

to keep pace with rising insulin resistance among insulin pump users (Feig, Corcoy et al. 2018). The meta-analysis

reported no significant difference in the percentage of women who had severe hypoglycaemia (Rys, Ludwig-

Slomczynska et al. 2018).

The meta-analysis of mostly observational studies reported higher incidence of miscarriage (which may be related

to women on CSII booking for pregnancy care earlier, resulting in better recording of miscarriage) and large for

gestational age babies in women using CSII versus MDI, but no difference in any other maternal, fetal or neonatal

outcome (Rys, Ludwig-Slomczynska et al. 2018).

There is some evidence that use of insulin pump therapy in pregnancy is associated with less blood glucose

variability than MDI (Jotic, Milicic et al. 2020).

In the experience of the authors, pump therapy may be of benefit in pregnant women with severe hyperemesis

gravidarum, both in improving the ability to control blood glucose levels and alleviate some of the symptoms

of hyperemesis.

Indications for insulin pump therapy (CSII) before and during pregnancy

NICE has published clear guidance on indications for CSII for adults with diabetes, and these apply to women

planning pregnancy (NICE 2008). NICE does not make specific recommendations on indications for CSII in women

planning pregnancy and does not set different HbA1c thresholds, although does recommend further research in

this area (NICE 2015). A reasonable approach is to consider CSII in women with type 1 diabetes planning pregnancy

where HbA1c has remained at or above 53 mmol/mol (7.0%) on MDI therapy despite a high level of care, as described

by NICE (NICE 2008, ABCD-DTN-UK 2018).

NICE recommends that pregnant women with insulin-treated diabetes should be offered CSII during pregnancy

if adequate blood glucose control is not obtained by optimised MDI without significant disabling hypoglycaemia

(NICE 2015).

Image used with permission: Medtronic

BEST PRACTICE GUIDE: USING DIABETES TECHNOLOGY IN PREGNANCY Version 2.0, March 2020

Starting insulin pump therapy (CSII) in pregnancy

When initiating insulin pump therapy outside of pregnancy, including pre-pregnancy, it is usual to start with

a total daily dose (TDD) of insulin via the pump which amounts to 75% of the pre-pump TDD on subcutaneous

insulin injections (ABCD-DTN-UK 2018). The DTN Best Practice Guide on CSII details starting regimens including

basal rate options and bolus settings (ABCD-DTN-UK 2018).

In pregnancy, maintaining good glycaemic control during switch from MDI to pump is essential. There is no

evidence that glycaemic control deteriorates when women are converted from injections to pump therapy so when

the switch to pump therapy is indicated this should be done as soon as this can be safely organised. However, after

starting insulin pump therapy in pregnancy, daily contact is essential so that initial settings can be adjusted every

1-2 days until optimised.

In pregnancy, the recommended starting pump TDD is 85±15% of the injection TDD (table 1). Consider using a lower

starting pump TDD (e.g. 70% of the injection TDD) if the woman has problematic hypoglycaemia or if total daily

bolus is more than 60% of pre-pump TDD. Consider using a higher starting pump TDD (e.g. 95% of the injection

TDD) if HbA1c >64 mmol/mol (8.0%), bearing in mind that HbA1c may reflect time before the woman knew of her

pregnancy and her diabetes management may have changed.

Table 1: Calculations for insulin pump settings for starting in pregnancy

Starting pump total daily dose (TDD) calculation: pre-pump TDDx(0.85±0.15)

Consider lower starting pump TDD if problematic hypoglycaemia or if total daily bolus >60% of pre-pump TDD

Consider higher starting pump TDD if HbA1c ≥64 mmol/mol

Starting pump settings

Basal rate

Total basal=pump TDDx0.5

Insulin:carbohydrate ratio (I:C ratio)

(1 unit of insulin for X grams of

carbohydrate)

Insulin sensitivity factor (ISF) (1 unit of

insulin reduces glucose by Y mmol/l)

Glucose target

Insulin active time

Option 1: at basal rate at total basal/24

units per hour

Option 2: variable basal rate (see table 2)

Before 20 weeks gestation*

Breakfast: 300/pump TDD

Other meals: 400/pump TDD

(use existing I:C ratios if working)

Before 20 weeks gestation*

ISF 130/pumpTDD

(use existing ISF if working)

Target: 5 mmol/l

Insulin active time: 4 hours

Pump setting adjustment

Basal rates Insulin:carbohydrate ratio (I:C ratio) ISF

Adjust 1-2 hours before

Adjust by 10-20%

(Avoid formal basal rate testing

involving fasting in pregnancy)

Adjust by 20%

(see table 3)

Correction should bring glucose back to

target range within 2-3 hours without

causing hypoglycaemia

Adjust by 10-20% (see table 4)

*see text if starting pump after 20 weeks gestation

Having decided the starting pump TDD, 50% of this should be set as the basal infusion rate. There are two options

for the basal insulin profile: either a flat rate over 24 hours (pump TDD x 0.5/24 units per hour) or using a modified

basal rate profile with 4-6 blocks which reflects the diurnal variation (table 2).

Table 2: Modied basal rate prole (King’s College Hospital)

Time of day Basal rate = total pump basal/24 units per hour

Bedtime to 3 hours before waking 80-100%

3 hours before waking to waking up 100-120%

Waking to lunch 80-100%

Lunch to evening meal 80-100%

Evening meal to bedtime 100-120%

BEST PRACTICE GUIDE: USING DIABETES TECHNOLOGY IN PREGNANCY Version 2.0, March 2020

Women should continue to use their existing insulin:carbohydrate ratio (I:C ratio) and insulin sensitivity factor

(ISF) if these appear effective. Alternatively, if starting before 20 weeks of gestation, the starting bolus ratios can be

calculated using 300/pumpTDD for breakfast and 400/pumpTDD for other meals and the ISF using 130/pumpTDD.

These may need to be adjusted if starting insulin pump therapy after 20 weeks of gestation (see next section). We

recommend setting all glucose targets (high, low and single) at 5 mmol/l, as this means the target is independent of

the pump system used.

When starting insulin pumps in pregnancy it is particularly important that women understand that the pump

only contains rapid-acting insulin and that if the infusion set or pump fails ketosis/ketoacidosis can develop within

hours. They should know the rules for the management of unexplained hyperglycaemia (see below), have back up

insulin pens (rapid-acting and long-acting) and should be advised not to change the set in the evening.

Using insulin pump therapy (CSII) in pregnancy

Insulin requirements may reduce in the first trimester and then increase from 16-20 weeks of gestation and pump

settings will need to be adjusted to reflect this. Bolus insulin requirements increase to a considerably greater

extent than basal insulin during pregnancy, with a 3-4 fold change in insulin:carbohydrate ratios from early to late

pregnancy (greatest with breakfast), compared to a 1.25-1.5 fold increase in basal insulin requirements over the

same time period. This translates into a change in basal:bolus ratio from 50:50 in early pregnancy to approximately

between 35:65 and 25:75 in later pregnancy.

Pump settings should be reviewed frequently and adjusted using data from glucose profiles.

Basal rate adjustment

Basal rates should be adjusted at least 1-2 hours before the inflection point on a CGM trace. Generally, aim for

no more than six basal time blocks per 24 hours and avoid short (less than 1-2 hour) blocks unless there is a clear

requirement. Short time blocks are not usually needed, and multiple short time blocks makes it more difficult to

see patterns and make further adjustments. Basal rates should be adjusted by 10-20%. We don’t advise formal basal

rate testing with fasting in pregnancy because daytime basal rates are often lower than physiological requirements

due to large bolus doses and the likelihood of ketosis with consequent nausea is higher.

Bolus calculator setting adjustment

Bolus settings should be reviewed frequently and adjusted using data from glucose profiles. It may also be useful to

sense check the settings.

Basal:bolus ratio sense check

• Up to 20 weeks of gestation: 50:50 to 35:65.

• Beyond 20 weeks of gestation: 35:65 to 25:75.

Insulin:carbohydrate (I:C) ratio (1 unit of insulin for X grams of carbohydrate)

• Usually adjust by 20% (see table 3).

• To sense check.

› Up to 20 weeks of gestation: breakfast 300/TDD, other meals 400/TDD

› Beyond 20 weeks of gestation: breakfast 200/TDD, other meals 300/TDD

Insulin sensitivity factor (ISF) (1 unit of insulin reduces glucose by Y mmol/l)

• Correction should bring glucose back into target range within 2-3 hours, without causing hypoglycaemia

(outside pregnancy this is 4-5 hours) (ABCD-DTN-UK 2018).

• Usually adjust by 10-20% (see table 4) (Although it makes sense to adjust the ISF when a ratio is adjusted, in

pregnancy the ISF does not change strictly in proportion to the I:C ratio).

• To sense check.

› Up to 20 weeks of gestation 130/TDD

› Beyond 20 weeks of gestation consider 100/TDD, particularly if women are using RT-CGM with alerts

Targets: set target glucose levels (high, low or single) to 5.0 mmol/l. If the woman has problematic hypoglycaemia

consider setting glucose target at 5.5 mmol/l.

BEST PRACTICE GUIDE: USING DIABETES TECHNOLOGY IN PREGNANCY Version 2.0, March 2020

Table 3: Adjusting insulin:carbohydrate ratios in the bolus calculator.

Ratios are usually adjusted by about 20%.

In this table each step increases insulin dose by approximately 20% (or reduces insulin dose by approximately 17%)

1 unit of insulin for X grams

of carbohydrate

Units of insulin per 10g

carbohydrate portion

Increasing bolus

dose (steps

increase dose by

approximately 20%)

Decreasing bolus

dose (steps

decrease dose by

approximately 17%)

1.4 g 7.4 units

1.6 g 6.2 units

1.9 g 5.2 units

2.3 g 4.3 units

2.8 g 3.6 units

3.3 g 3.0 units

4.0 g 2.5 units

4.8 g 2.1 units

5.8 g 1.7 units

6.9 g 1.44 units

8.3 g 1.20 units

10 g 1 unit

12 g 0.83 units

14 g 0.69 units

17 g 0.58 units

21 g 0.48 units

25 g 0.40 units

Table 4: Adjusting insulin sensitivity factor (ISF) in the bolus calculator.

ISF is usually adjusted by 10-20%.

In this table each step increases corrective dose by approximately 10% (or reduces corrective dose by approximately 9%)

ISF: 1 unit of insulin reduces glucose by

Increasing

corrective dose

(steps increase dose

by approximately

10%)

Decreasing

corrective dose

(steps decrease dose

by approximately

9%)

1.0 mmol/l

1.1 mmol/l

1.2 mmol/l

1.3 mmol/l

1.4 mmol/l

1.5 mmol/l

1.7 mmol/l

1.9 mmol/l

2.1 mmol/l

2.3 mmol/l

2.5 mmol/l

2.7 mmol/l

3.0 mmol/l

3.3 mmol/l

3.6 mmol/l

4.0 mmol/l

4.4 mmol/l

4.8 mmol/l

5.3 mmol/l

5.8 mmol/l

BEST PRACTICE GUIDE: USING DIABETES TECHNOLOGY IN PREGNANCY Version 2.0, March 2020

Other considerations when using insulin pump therapy in pregnancy

• Management of unexplained hyperglycaemia and sick day rules are slightly different in pregnancy (see below).

• In case of set/pump failure, insulin pump users must carry a rapid-acting insulin pen with them at all times,

should have access to long-acting insulin pens and should document their pump settings regularly.

• Change infusion sets regularly, at least every 2-3 days, to minimise ketosis risk and optimise insulin absorption.

• Reinforce to perform set changes early in the day, not in the evening. Pump users should check glucose 2-4

hours after a set change to make sure the new set is working. Avoid going to sleep within 2 hours of a set

change (or set an alarm after 2 hours to wake up and check glucose).

• Site selection: cannula placement may be difficult as the abdominal skin tautens as pregnancy progresses and

sites may irritate more easily. Consider using alternative sites such as flanks, buttocks and thighs and different

cannulas, including metal.

• Reservoir volume: daily insulin requirements may exceed 100 units so consider using a pump with a 300 unit

reservoir if initiating pump therapy during pregnancy.

• Beyond 20 weeks of gestation, insulin absorption is delayed (Goudie, Lunn et al. 2014). Where possible, bolus

doses should be given:

› Up to 20 weeks of gestation: 15±10 min pre-meals

› Beyond 20 weeks of gestation: extend towards 45±15 min pre-meals

• Consider the use of a basal to bolus switch (or “Super Bolus”) (box 1) to cover postprandial peaks, particularly

breakfast, if increasing the bolus ratio results in late postprandial hypoglycaemia.

Box 1: Using a basal to bolus switch (‘Super Bolus’) – worked example

30 year old woman, 28 weeks gestation struggling with post-breakfast hyperglycaemia:

• Glucose peak 9–10 mmol/l (about one hour after breakfast).

• I:C ratio of 1:4 (1 unit for 4 g carbohydrate).

• Basal insulin infusion rate 08.00-13.00 = 1.3 units/hour.

She is already limiting her breakfast carbohydrate and giving her bolus 30 min before eating (see table 11). She tried

adjusting the breakfast I:C ratio from 1:4 g to 1:3.3 g. She found that, using the recommended pre-breakfast bolus at

08.00 am, the peak glucose (at about 1 hour) was <7.8 mmol/l but she will be hypoglycaemic around 10:00-11.00 am.

Option A: ‘Super Bolus’

The simplest version of a ‘Super Bolus’ is to calculate

basal insulin delivered over 2 (or 3) hours, add

this to the recommended meal bolus, and set a 0%

temporary basal rate for 2 (or 3) hours. e.g:

• Breakfast at ~ 08.30 contains 20 g carbohydrate.

• Recommended meal bolus 5 units. Basal insulin

delivered over 2 hours is 2.6 units.

• Adjust recommended bolus from 5 units to 7.6

units (given at 08.00 am)

• Set 0% temporary basal rate for 2 hours starting

at 08.00 am.

o Advantage: adjustment of meal bolus and setting

temporary basal are linked, so unlikely to do one

without the other

n Disadvantages: doing the calculation, several steps.

Option B:

Adjust I:C ratio settings. Plan to set a 0% temporary

basal rate for 2 (or 3) hours after giving meal bolus.

e.g:

• Change breakfast I:C ratio by at least 20% (from

1:4 g to 1:3.3 g)

• Breakfast at ~ 08.30 contains 20 g carbohydrate.

• Give recommended bolus (6.1 units).

• Set a 0 % temporary basal rate for 2 (or 3) hours

o Advantage: no calculation, single step

n Disadvantage: if the person forgets to set the

temporary basal rate they are highly likely to have

hypoglycaemia at 2-4 hours post bolus.

Management of unexplained hyperglycaemia for insulin pump (CSII) users in pregnancy

Set, or pump, failure can occur and if not detected and managed can potentially result in the development of ketosis/

ketoacidosis within hours (ABCD-DTN-UK 2018). All insulin pump users should be aware of the potential for set failure

and how to manage this. The guidance for management of unexplained hyperglycaemia is very similar to outside

pregnancy (ABCD-DTN-UK 2018). The differences are that the glucose threshold at which action is recommended is

lower (10 mmol/l rather than 13 mmol/l), checking for ketones is earlier and the ketone threshold for action is lower. This

is because pregnancy is a ketogenic state. Pregnant women with type 1 diabetes require blood ketone testing equipment.

Rules for managing unexplained hyperglycaemia for CSII users in pregnancy are given in figure 2.

BEST PRACTICE GUIDE: USING DIABETES TECHNOLOGY IN PREGNANCY Version 2.0, March 2020

Unexplained hyperglycaemia

Glucose >10 mmol/l (conrmed on ngerstick blood glucose)

Check blood ketones

Ketones ≤1 mmol/l

Correction bolus by pump

Check ngerstick glucose in 2 hours

Glucose unchanged or higher

Check blood ketones

Ketones ≤1 mmol/l

Use pen/syringe to give

correction injection

Change infusion set & reservoir

Check ngerstick glucose in 2 hours

Repeat the above steps ONCE (starting from the top).

If after doing this glucose still unchanged or higher:

• Seek medical help

• Convert to insulin regimen using rapid-acting & long-acting insulin

given by pen/syringe and remove pump.

Glucose improved

Return to usual regimen,

give further correction

through pump if required

Do not go to sleep with unexplained hyperglycaemia which has not resolved OR within 2

hours of a set change OR within 4 hours of a correction given by pen/syringe.

Ketones >1 mmol/l

Follow sick day rules for

pregnancy: ketones present

Use pen/syringe to give

correction injection

Change infusion set & reservoir

Figure 2: Rules for managing unexplained hyperglycaemia in pregnancy for insulin pump users

BEST PRACTICE GUIDE: USING DIABETES TECHNOLOGY IN PREGNANCY Version 2.0, March 2020

Insulin pump users should be encouraged to explore the reasons why the high glucose has occurred (ABCD-DTN-UK 2018).

Pump users must have back up insulin pens (rapid-acting and long-acting) and should document their pump

settings regularly. This is particularly important in pregnancy as pump settings are changed frequently. In the

event of pump failure, the emergency basal insulin would be the same as the total daily basal insulin on the pump

and the I:C ratio and ISF would be the same as on the pump.

Pump users should routinely check glucose 2-4 hours after a set change to make sure the new set is working. Avoid

going to sleep within 2 hours of a set change (or set an alarm after 2 hours to wake up and check glucose). If pump

users need to give a correction dose using a pen/syringe, and then continue using insulin pump therapy (usually

after changing the infusion set and reservoir), they should check glucose at 2 and 4 hours and wait at least 4 hours

before sleeping. This is because if the set/pump is not working, glucose may only start to rise once the effect of the

rapid-acting insulin given by pen/syringe is wearing off (3-4 hours).

Sick day rules for insulin pump (CSII) users in pregnancy

Pregnancy is a ketogenic state. Women with diabetes are at higher risk of developing diabetic ketoacidosis (DKA)

when they are pregnant. The NPID Audit reported 2.7% of pregnant women with type 1 diabetes have at least one

admission with diabetic ketoacidosis (NPID). Diabetic ketoacidosis in pregnancy carries a risk of fetal death (16-27% or

higher) (Schneider, Umpierrez et al. 2003, Morrison, Movassaghian et al. 2017). Women using CSII are at higher risk of

ketosis than those using MDI as any interruption in insulin delivery can result in hyperglycaemia and ketosis as the

insulin pump only contains rapid-acting insulin. Prompt detection of ketones is key to early treatment.

The sick day rules for pregnant women using CSII are given in figure 3. Fingerstick glucose checks (rather than

sensor glucose) should be used during illness. The general principles of sick day rules are similar to outside

pregnancy (ABCD-DTN-UK 2018). However, there are some differences. Pregnant women with type 1 diabetes

require blood ketone testing equipment, and should not be expected to manage using urine ketone testing. In

pregnancy blood ketone levels of above 1.0 mmol/l are considered significant requiring increased insulin and

monitoring, which is lower than the threshold of 1.5 mmol/l outside of pregnancy. In pregnancy the threshold for

attending hospital as an emergency is lower: straight away if ketones are above 3 mmol/l and if no improvement

after 2 hours if ketones 1.1-3 mmol/l. Women should continue to follow the sick day rules until they are assessed

and alternative treatment instituted if needed. We recommend early set change with bolus using pen/syringe. We

encourage women to maintain carbohydrate intake with bolus insulin (using their normal I:C ratio) as this will help

suppress ketogenesis.

Pregnant women with diabetes who are unwell with vomiting, abdominal pain or ketones MUST be assessed for

DKA even if glucose levels are normal. DKA can occur in pregnancy in a woman with known diabetes with a

normal blood glucose (JBDS-IP 2017). Guidance on management of diabetic ketoacidosis in pregnancy is given

by JBDS-IP (JBDS-IP 2017). Women who are not in DKA, but have ketone levels above 1.0 mmol/l which are not

quickly resolved using sick day rules, will require variable rate intravenous insulin (VRII) and appropriate fluid

replacement. Women may continue to use insulin pump therapy as well as VRII to allow basal rates to continue if

she is able to manage the pump herself.

Whilst on intravenous insulin infusion for ketosis or DKA, capillary blood glucose checks (not sensor glucose)

are required every hour. If pump therapy has been discontinued during the ketosis event, then once resolved an

overlap of 1 hour between restarting CSII and stopping intravenous insulin is required.

BEST PRACTICE GUIDE: USING DIABETES TECHNOLOGY IN PREGNANCY Version 2.0, March 2020

FEEL UNWELL?

Check FINGERSTICK blood glucose (BG) & ketones

NO KETONES

Ketones ≤1 mmol/l on blood test

KETONES PRESENT

Ketones >1 mmol/l on blood test

SEVERE ILLNESS

• Usual I:C ratio when eating

• Use corrective boluses via pump (every 2 hours if needed) if BG

is raised, even if not eating. When unwell you may nd you need

larger bolus doses to reduce blood glucose – override the bolus

advisor

• If glucose levels are above >8 mmol/l, increase basal rate by

10% by using a temporary basal

Sip sugar-free uids (at least 100 ml/hour)

Eat as usual if you are able (with bolus insulin).

If you don’t feel like eating, try to have some carbohydrate (e.g. 20g) with bolus insulin every 2-4 hours

Check ngerstick blood glucose (target 4-8 mmol/l) &

blood ketones (target <0.5 mmol/l) every 2-4 hours

Glucose improving

after 2 hours

YES NO

Glucose unchanged or

higher after 2 hours

Use pen/syringe to give

corrective bolus

Change infusion set & reservoir

(if not already done)

Return to usual regimen if:

• Blood ketones <0.5 mmol/l for at least 6 hours AND

• Blood glucose 4-8 mmol/l for at least 6 hours AND

• At least 4 hours since last injection given using pen/syringe AND

• You are eating & drinking and feeling better

Check ngerstick blood glucose (target 4-8 mmol/l) &

blood ketones (target <0.5 mmol/l) every 2 hours

Calculate Total Daily Dose (TDD) from previous day

Use pen/syringe to give rst bolus (see below)

Change infusion set & reservoir (if not already done)

Ketones improving

after every 2 hours

AND

if after 6 hours ketone ≤1mmol/l

& glucose ≤10 mmol/l

• Give 10% of TDD as bolus

insulin every 2 hours

• Plus use usual I:C ratio when

eating

• Increase basal rate by 30%

Prepare to go to hospital as

an emergency (A&E) if no

improvement after 2 hours

Blood ketones 1.1-3 mmol/l

Ketones NOT improved after 2

hours OR ketones >3 mmol/l

• Give 20% of TDD as bolus

insulin every 2 hours

• Plus use usual I:C ratio when

eating

• Increase basal rate by 50%

Go to hospital as an emergency

Blood ketones >3 mmol/l

Go to hospital as an emergency if:

• Blood ketones >3 mmol/l OR

• Blood ketones not improving after 2 hours OR

• Blood ketones still >1 mmol/l after 6 hours OR

• Blood glucose still >10 mmol/l after 6 hours OR

• You are having hypos

• You continue to vomit/unable to keep uids down

Continue to follow these rules until you are assessed & alternative

treatment started if needed

Figure 3: Sick day rules in pregnancy for insulin pump users

BEST PRACTICE GUIDE: USING DIABETES TECHNOLOGY IN PREGNANCY Version 2.0, March 2020

USING GLUCOSE SENSORS BEFORE AND DURING

PREGNANCY

Glucose sensors measure glucose in the interstitial fluid just below

the skin. Changes in interstitial glucose, and therefore sensor glucose,

lag about 5-10 minutes behind changes in blood glucose. It is generally

accepted that the sensor accuracy is not affected by the physiology of

pregnancy (Murphy 2019).

Real-time continuous glucose monitoring (RT-CGM) provides continuous

information about sensor glucose (updated every 5 minutes) with the

facility to set glucose alarms. Of the current RT-CGM systems, we have

extensive experience of supporting pregnant women using the Dexcom

G6 and the Medtronic Guardian Sensor 3. Dexcom G6 has a CE mark

specifically for use in pregnancy (meaning it complies with European Union

health, safety and environmental standards). Medtronic Guardian Sensor

3 has a CE mark that does not exclude use in pregnancy. The Dexcom

G6 sensor data is viewed using the Dexcom G6 App on a mobile phone,

the Dexcom receiver or used with the Tandem t:slim insulin pump. The

Medtronic Guardian Sensor 3 data is viewed using the Guardian Connect

App on a mobile phone or used with the MiniMed 640G or 670G insulin

pump systems. Note the MiniMed 670G Auto Mode feature (hybrid closed

loop system) is not suitable for use in pregnancy because the target is too

high for pregnancy. Other RT-CGM systems are available.

Flash glucose monitoring only provides a glucose reading when the user ‘scans’ the device and, currently, it does not

have alarms. Currently the FreeStyle Libre is the only flash glucose monitoring system available and has a CE mark

for use in pregnancy (meaning it complies with European Union health, safety and environmental standards).

Evidence for RT-CGM before and during pregnancy

The CONCEPTT trial provided evidence that, in women with type 1 diabetes, RT-CGM started in the first trimester

versus capillary glucose monitoring improves glycaemic control and improves neonatal outcomes, including

reduction in large for gestational age (number needed to treat to prevent 1 case (NNT) 6), neonatal hypoglycaemia

(NNT 8), neonatal intensive care admission (NNT 6) and length of hospital stay (Feig, Donovan et al. 2017).

In CONCEPTT, at 34 weeks of gestation the sensor glucose time in target range 3.5-7.8 mmol/l was 68%±13% in the

RT-CGM arm and 61%±15% in the usual care arm, an increase of 1.7 hours per day in target range. In CONCEPTT,

most women only achieved the recommended 70% time-in-target-range (figure 1, (Battelino, Danne et al. 2019)) in

the last 3-4 weeks of pregnancy. For optimal neonatal outcomes women should be supported to reach 70% time-in-

range as early as possible during pregnancy and then maintain this.

In CONCEPTT, the benefit of RT-CGM was seen both in women using pumps and women using multiple daily

injections suggesting that women do not need to be familiar with diabetes technology or be using pumps to benefit

from RT-CGM. Furthermore, the RT-CGM treatment effect was comparable in women with baseline HbA1c levels

above and below 58 mmol/mol (7.5%). Our experience is that women with low numeracy and/or literacy can also

benefit from RT-CGM with individualised advice and support.

CONCEPTT also included a group of women planning pregnancy (Feig, Donovan et al. 2017). There was no clear

evidence of benefit on glycaemic control at confirmed pregnancy, or after 24 weeks study period in women who did

not conceive, possibly due to the small sample size. However, as the authors noted, women may be uncomfortable

changing monitoring modality during early pregnancy and it may be logistically easier for services to plan RT-CGM

starts during pregnancy planning. The evidence for RT-CGM outside pregnancy will be covered in the ABCD-DTN-

UK Best Practice Guide for RT-CGM.

Image used with with permission: instagram.com/diabetesia

BEST PRACTICE GUIDE: USING DIABETES TECHNOLOGY IN PREGNANCY Version 2.0, March 2020

Indications for RT-CGM before and during pregnancy

Women preparing for pregnancy may be assessed for RT-CGM using the usual criteria for adults with diabetes.

For women who are pregnant, current NICE guidelines (NICE 2015), published before CONCEPTT was published,

are given in box 2. NICE are currently undertaking a partial update focused on CGM in women with type 1 diabetes

who are planning to become pregnant or already pregnant (https://www.nice.org.uk/guidance/ng3).

Box 2: NICE guidance on continuous glucose monitoring in pregnancy (NICE 2015) – (note this is under

review-see text)

• Do not offer continuous glucose monitoring routinely to pregnant women with diabetes

• Consider continuous glucose monitoring for pregnant women on insulin therapy:

› who have problematic severe hypoglycaemia (with or without impaired awareness of hypoglycaemia) or

› who have unstable blood glucose levels (to minimise variability) or

› to gain information about variability in blood glucose levels

The NHS Long Term Plan launched in January 2019 states

‘…by 2020/21 all pregnant women with type 1 diabetes will be offered continuous glucose monitoring, helping to

improve neonatal outcomes.’ (section 3.80, (www.longtermplan.nhs.uk/).

At the time of writing, NHS England are working to implement this plan, in Scotland availability of RT-CGM for all

pregnant women with type 1 diabetes is subject to a Scottish health technology assessment as well as assessment

in an updated SIGN guideline, and in Wales RT-CGM is recommended for pregnant women with type 1 diabetes

(Health_Technology_Wales 2019).

Information provided by RT-CGM

For people to use RT-CGM effectively and safely, the sensor settings should be appropriate, and both they and their

diabetes health professionals need to understand the information the RT-CGM is providing and how to use that

information. It is important to recognise that the extra information provided by RT-CGM can be overwhelming

and the psychological impact of seeing ‘out of range’ sensor glucose readings can be difficult to cope with and may

result in over-reaction. This is particularly important in pregnancy as women are understandably very keen to

keep sensor glucose in the target range.

Sensor glucose

RT-CGM measures glucose in the interstitial fluid just below the skin (whereas a fingerstick test measures capillary

blood glucose). The displayed sensor glucose value is updated every 5 minutes. Changes in interstitial glucose

usually lag 5-10 minutes behind changes in blood glucose. For example, if blood glucose has dropped steadily from

10mmol/l to 3mmol/l over the last hour, the fingerstick glucose will be around 3mmol/l but the sensor glucose will

still be around 4 mmol/l.

Even when blood and interstitial glucose are stable RT-CGM devices may give different results from fingerstick

CBG readings.

For these reasons there is often a difference between fingerstick CBG and sensor glucose. If there is a difference

the fingerstick CBG should be considered the more accurate.

Direction of change arrows

The arrows show how fast and in which direction sensor glucose has changed over the past 20 minutes (table 5).

Graphs

The graphs shown on the RT-CGM display show sensor glucose over the preceding 3 hours, 6 hours, 12 hours or 24 hours.

BEST PRACTICE GUIDE: USING DIABETES TECHNOLOGY IN PREGNANCY Version 2.0, March 2020

Table 5: What RT-CGM/ash glucose monitoring arrows mean

Dexcom G6

Medtronic

Guardian

Sensor 3

FreeStyle

Libre

Over past 20

min sensor

glucose has

been:

Approximate*

rate of change

If trend continues, approximately*:

How long to

change by 1

mmol/l?

How much will sensor

glucose change in 30

mins?







Rising very

quickly

>0.2 mmol/l/

min

< 5 min >5 mmol/l

 

Rising quickly

0.1-0.2 mmol/l/

min

Average 7 min 3-5 mmol/l

  

Rising

0.06-0.1

mmol/l/min

Average 15 min 2-3 mmol/l



No arrow



Stable or

changing slowly

< 0.06 mmol/l/

min

>20 min <2 mmol/l

  

Falling

0.06-0.1

mmol/l/min

Average 15 min 2-3 mmol/l

 



Falling quickly

0.1-0.2 mmol/l/

min

Average 7 min 3-5 mmol/l





Falling very

quickly

>0.2mmol/l/

min

< 5 min >5 mmol/l

*Approximate gures given for practical purposes and for harmonisation across devices.

RT-CGM set up in pregnancy

Issues to cover for women starting RT-CGM during pregnancy are listed in box 3 (page 28) and tips for established

users of RT-CGM who are newly pregnant in box 4 (page 29).

Calibration

Calibration requirements vary between RT-CGM systems and some do not need to be calibrated at all. At the time

of writing, Dexcom G6 does not require calibration, whereas Medtronic Guardian Sensor 3 does require calibration

(within the first 2 hours and then at least every 12 hours).

For systems requiring calibration, it is very important to get the calibration process right otherwise the RT-CGM

will be inaccurate. It is important to minimise the risk of inaccurate fingerstick CBG readings and calibrate when

there is the least likelihood of a difference between capillary blood glucose and interstitial glucose.

• To maximise likelihood of accurate capillary blood glucose reading.

› Clean hands

› Don’t use sites other than fingers

• Calibrate when glucose levels are stable.

› Before meal bolus or before bed (if >3 hours after eating)

› Sensor showing glucose stable (no up or down arrows)

› NOT after hypoglycaemia, exercise, or within 3 hours after a bolus (correction/meal)

• Calibrate when fingerstick & sensor glucose similar (within 3 mmol/l).

› This may sound counter-intuitive, but if the sensor and fingerstick readings are very different it is likely

that blood glucose is changing quite rapidly or there is an aberrant sensor or CBG reading and using this

will introduce error

• For systems needing calibration at least 12 hourly (e.g. Medtronic Guardian Sensor 3) if no calibration is

performed within 12 hours the sensor will alert and if the user does not calibrate it will stop providing a sensor

glucose reading. Therefore:

› For every fingerstick CBG performed think whether the above criteria are met and calibrate every time they are.

This is because it can be very difficult, particularly in pregnancy, to find enough times suitable for calibration

› Do a calibration in the evening. This is because otherwise the calibration alert will disturb sleep and if the

user then does not calibrate the sensor will not provide readings for the rest of the night

BEST PRACTICE GUIDE: USING DIABETES TECHNOLOGY IN PREGNANCY Version 2.0, March 2020

Alerts

RT-CGM systems have many possible glucose alerts (table 6). However, many of the negative comments from RT-

CGM users are about alerts/alarms. People may stop using the RT-CGM altogether because of alerts or may develop

‘alert fatigue’ where they ignore all the alerts including the more important ones. The general rule is only turn on

an alert if it is something the user can safely, and should, do something about at that time.

Table 6: RT-CGM glucose alerts and alarms

Alerts Set up Use/comment

Predictive high

glucose alert

A set number of minutes before

the sensor glucose is predicted

to reach the high glucose

threshold

O Rarely used: may trigger frequently (e.g.

after meals or hypo treatment) and generally

should not be acted on. Does not add much

to high glucose alert.

High glucose

alert

When the sensor glucose

reaches the high glucose

threshold

On To alert if sensor glucose is high at a level

which can and should be acted on.

Rate of rise alert

When the sensor glucose is

rising rapidly

O Rarely used: will trigger frequently (e.g. after

meals or hypo treatment) and generally

should not be acted on. Sometimes used as a

‘missed bolus’ alert.

Predictive low

glucose alert

A set number of minutes (usually

20-30 min) before the sensor

glucose is predicted to reach a

low glucose threshold

On To avoid hypoglycaemia.

Low glucose alert

When the sensor glucose

reaches a low glucose threshold

On or o To alert the user to hypoglycaemia or near-

hypoglycaemia. Some systems (e.g. Dexcom

G6) can have 2 separate low thresholds.

Rate of fall alert

When the sensor glucose is

falling rapidly

O Rarely used: rapidly falling glucose is only

important if approaching hypoglycaemia (which

will trigger predictive low glucose alert.)

High glucose alerts

• The useful high alert is the high glucose alert, which, provided the threshold is set at a suitable level, is

something the user can and should do something about. It alerts when the sensor glucose reaches the high

glucose threshold.

• High alert glucose thresholds are usually set at a level that may indicate system failure (e.g. set/pump failure or

missed bolus). High alert glucose threshold should be set at:

› Usual waking hours at 12-15 mmol/l

› Usual sleeping hours at 8-10 mmol/l

• Setting the thresholds lower than this will usually result in multiple alerts, with the risk of alert fatigue, and

may encourage inappropriate corrections with subsequent hypoglycaemia. It is important to explain this to

pregnant women otherwise many will be keen to set the threshold much lower.

• For new RT-CGM users the high alert thresholds should usually be set at the top of the range (i.e. waking hours

at 15 mmol/l, sleeping hours at 10 mmol/l). These can be brought down to the lower end of the range as they

become more familiar with the system and glycaemic control improves.

• Some more advanced RT-CGM users may choose to set the high alert glucose threshold lower than these

ranges. In such cases the risk of alert fatigue and/or inappropriate corrections should be discussed.

• The high repeat (Dexcom G6) or high snooze (Medtronic Guardian Sensor 3 used with MiniMed 640G or 670G

systems) should be set at 2 hours as this is the recommended minimum time between repeated corrective doses.

Low glucose alerts

RT-CGM systems have slightly different low glucose alert functionalities and need to be set up slightly differently.

Probably the most useful alert is the predictive low glucose alert as this is something the user can and should do

something about in order to avoid hypoglycaemia. It alerts a set number of minutes (usually 20-30 min) before the

sensor glucose is predicted to reach a low glucose threshold.

BEST PRACTICE GUIDE: USING DIABETES TECHNOLOGY IN PREGNANCY Version 2.0, March 2020

• Dexcom G6:

› ‘Urgent Low Soon Alert’: alerts 20 min before sensor glucose predicted to reach 3.1 mmol/l (the threshold of

3.1 mmol/l cannot be changed)

› ‘Urgent Low Alarm’: alarms when sensor glucose reaches 3.1 mmol/l (the threshold of 3.1 mmol/l cannot be

changed). This alarm cannot be turned off.

› ‘Low Glucose Alert’: alerts when the sensor glucose reaches the individualised low glucose threshold. The

purpose is to alert the user when sensor glucose is low but is not falling quickly enough to trigger the ‘Urgent

Low Soon Alert’. Users may wish to keep this alert on as the urgent low threshold of 3.1 mmol/l is quite

low. The threshold is usually set at 3.4-4.0 mmol/l depending on hypoglycaemia awareness and personal

preference. For example, if the woman has reduced hypoglycaemia unawareness the threshold is set at the

higher end of this range. For those who want to minimise alerts, a threshold at the lower end of the range

may be used. Some women will choose to set the threshold at the higher end of this range, accepting that this

will result in lots of alerts and some of the time the appropriate action will be to watch and wait.

• Medtronic Guardian Sensor 3 used with MiniMed 640G or 670G systems or the Guardian Connect App:

› ‘Alert Before Low’: alerts 30 min before sensor glucose predicted to reach the individualised low glucose

threshold. This is usually set at 3.4-3.6 mmol/l. If the woman has reduced hypoglycaemia awareness or

severe hypoglycaemia consider setting the low glucose alert threshold at 3.8-4 mmol/l.

› ‘Alert On Low’: alerts when the sensor glucose reaches the individualised low glucose threshold. This may be

kept off to limit the number of alerts, as the person should already be responding to the Alert Before Low

› The Guardian Connect App (but not the MiniMed 640G or 670G) has an ‘Urgent Low’ alarm set at 3.1 mmol/l

which cannot be turned off

The repeat (Dexcom G6) or low snooze (Medtronic Guardian Sensor 3) (time after an alert before another alert if

the alert condition still exists) should be set at 30 min. Treatment or avoidance of hypoglycaemia should usually

be complete within 30 min and the system will then warn of a further episode. Setting the repeat at a shorter time

may result in the alerts sounding when the woman is still treating the hypoglycaemia (which can be annoying)

or when the hypoglycaemia has been successfully treated/avoided but the sensor is lagging behind (which can

encourage overtreatment of hypoglycaemia).

Using RT-CGM in pregnancy

Issues to cover for women starting RT-CGM during pregnancy are listed in box 3 (page 28) and tips for established

users of RT-CGM who are newly pregnant in box 4 (page 29).

When to check ngerstick CBG

There may be a difference between fingerstick CBG and sensor glucose due to sensor lag or inaccuracy. If there is a

difference the fingerstick CBG should be considered the more accurate.

Situations to ALWAYS check ngerstick CBG

RT-CGM users should be advised to ALWAYS check fingerstick CBG (rather than using sensor glucose) in the

following situations (this is the same as outside pregnancy)

• To confirm hypoglycaemia AND monitor recovery from hypoglycaemia (Those with good hypoglycaemia

awareness may not need to repeat fingerstick to confirm recovery from hypoglycaemia if their symptoms have

improved. However, due to sensor lag, sensor glucose should not be used to guide the requirement for further

15g of quick acting carbohydrate).

• If symptoms do not match sensor glucose (e.g. if symptoms of hypoglycaemia and sensor glucose normal).

• If sensor reading seems unlikely in the circumstances.

• If sensor reading unreliable (e.g. no reading, or no arrow (in the Dexcom G6 system)).

• If required for calibration.

• During and after exercise.

• When following ‘sick day rules’ or ‘rules for management of unexplained hyperglycaemia’.

• When in hospital, RT-CGM SHOULD NOT be used to adjust variable rate intravenous insulin infusions (VRII).

Note the Medtronic Guardian Sensor 3 may give falsely high glucose readings after paracetamol. Dexcom G6 is not

affected by paracetamol (unlike earlier Dexcom systems).

BEST PRACTICE GUIDE: USING DIABETES TECHNOLOGY IN PREGNANCY Version 2.0, March 2020

Situations when it is ADVISED to check ngerstick CBG in pregnancy

In pregnancy, RT-CGM users should be advised to also check fingerstick CBG in the following situations.

• Before using a glucose value to calculate meal bolus (the glucose is used to calculate the correction part of the

meal bolus).

• Before taking action to avoid hypoglycaemia (e.g. when responding to the predictive low glucose alert).

• Before taking action to correct hyperglycaemia.

In pregnancy we advise this for all RT-CGM systems, including those licenced to be used instead of fingerstick

capillary blood glucose testing for treatment decisions (termed non-adjunctive systems). This advice is different

from outside pregnancy, because the glucose targets in pregnancy are very tight and time spent outside range is

particularly important in pregnancy so discrepancy between sensor glucose and fingerstick CBG is important. A

sensor that is giving readings that are slightly out will impact for the lifetime of that sensor, a duration that may

be important in pregnancy. However, experienced RT-CGM users may reasonably choose not to check fingerstick

glucose in these three situations if they are using a non-adjunctive system and are confident the sensor is working

well (e.g. smooth graph, no gaps, sensor glucose has been close to fingerstick glucose, and sensor glucose as

expected in the circumstances). However, we recommend testing fingerstick glucose an absolute minimum of once

per day, usually first thing in the morning, to monitor accuracy of that particular sensor.

Routine monitoring of fingerstick CBG at 1-2 hours post-meal is generally not required in women using RT-CGM in

pregnancy, provided none of the above apply.

Pregnant women using RT-CGM will generally require at least 4 CBG test strips per day.

When to look at RT-CGM readings

As a minimum, pregnant women should look at their RT-CGM display on getting up in the morning, pre-meals, 1

hour post-meals (for reflection in real-time), 2 hours post-meals (for reflection in real-time and possible corrective

action), pre-bed and if they wake in the night.

Responding to high sensor glucose

It is very important to remind RT-CGM users of the action profile of rapid-acting insulin and to emphasise that

a corrective dose of rapid-acting insulin will not reduce sensor glucose immediately, that a rising glucose may

continue to rise for 30-60 minutes after a correction, and to leave at least 2 hours between corrective doses. Giving

too large or multiple corrective doses will not reduce glucose more quickly but will cause hypos.

Hyperglycaemia after meals

Generally pregnant women should be advised NOT to correct postprandial hyperglycaemia because this will often

cause subsequent hypoglycaemia. Seeing postprandial hyperglycaemia should trigger ‘real-time’ reflection on the

cause and possible solutions for the future (see below). However, prolonged hyperglycaemia should be acted on. A

practical approach is if sensor glucose is high at 1 hour post-meal, check sensor glucose again at 2 hours post-meal.

If sensor glucose remains above target AND sensor glucose is either stable or increasing (upward arrow(s)) then it is

reasonable to give a corrective dose. CSII users should use the bolus advisor (which takes account of the insulin on

board). MDI users should use half the dose calculated using their individualised ISF or may use an App to calculate

the corrective bolus taking account of the insulin on board. Women should not correct if sensor glucose is falling

(downward arrows). Alternatively, if sensor glucose is high at 1 hour post-meal, exercise, such as a walk round the

block, can be used to reduce glucose if practical.

Hyperglycaemia after hypoglycaemia

Generally pregnant women using RT-CGM should be advised NOT to correct hyperglycaemia after hypoglycaemia

as this will often cause further hypoglycaemia. However, prolonged hyperglycaemia should be acted on. In

pregnancy, a practical approach is if sensor glucose is high 2 hours after treating hypoglycaemia and the sensor

glucose is either stable or increasing (upward arrow(s)) then it is reasonable to give a corrective dose (if CSII using

the bolus advisor OR if MDI using individualised ISF). Women should not correct if sensor glucose is falling

(downward arrows).

Hyperglycaemia at other times

Hyperglycaemia pre-meals or more than 3 hours after a meal bolus can be corrected as usual using individualised

ISF (see next section).

BEST PRACTICE GUIDE: USING DIABETES TECHNOLOGY IN PREGNANCY Version 2.0, March 2020

Unexplained marked hyperglycaemia (e.g. > 10 mmol/l) should trigger consideration of the cause such as system

failure (e.g. set failure for CSII users or missed insulin) or illness, and the appropriate response to this. It is essential

that CSII users know the rules for the management of unexplained hyperglycaemia (see section on management of

unexplained hyperglycaemia for CSII users in pregnancy and figure 2) which might be due to set/pump failure and

can potentially result in the development of ketosis/ketoacidosis within hours. It is also essential that pregnant

women with type 1 diabetes (whether CSII or MDI users) know the sick day rules for pregnancy as the risk of

ketosis is higher in pregnancy (see section on sick day rules for CSII users in pregnancy and figure 3).

Responding to predictive low glucose alerts and low glucose alerts.

The predictive low glucose alert is activated a set number of minutes before the sensor glucose is predicted (if the

sensor glucose keeps falling at the same rate) to reach the low glucose threshold. Due to the sensor lag and/or

the glucose threshold set, the capillary blood glucose may already be <3.5 mmol/l when the predictive low glucose

alert is triggered. The person should respond promptly to a predictive low glucose alert. The low glucose alert/

alarm is activated when the sensor glucose reaches a low glucose threshold. However, there may be a difference

between CBG and sensor glucose. Therefore, the woman should check fingerstick CBG immediately if they have

hypoglycaemia symptoms (even if sensor glucose normal), if the sensor is reading < 3.5 mmol/l, if the low glucose

alert is triggered or if the predictive low glucose alert is triggered. If there is a difference the fingerstick CBG

should be considered accurate.

Guidance for avoiding and treating hypoglycaemia is shown in figure 4. Hypoglycaemia avoidance (fingerstick

CBG ≥3.5 mmol/l) is managed by consuming 10g quick acting carbohydrate (table 7) and reviewing sensor glucose/

arrows. Hypoglycaemia (fingerstick CBG<3.5 mmol/l) is treated by consuming 15g quick acting carbohydrate

(table 7) and using fingerstick (NOT sensor) glucose to monitor recovery from hypoglycaemia. This is because the

sensor glucose lags 5-10 min behind blood glucose and so sensor glucose will continue to fall for 5-10 min even if

blood glucose is rising after effective hypoglycaemia treatment. Using sensor glucose to monitor recovery from

hypoglycaemia usually results in over treatment of hypoglycaemia. A practical tip is to advise RT-CGM users to

avoid looking at the RT-CGM display until fingerstick CBG shows recovery, as it is very difficult to ignore a falling

sensor glucose.

Table 7: Some quick acting (QA) carbohydrate suggestions for avoiding and treating hypoglycaemia

Hypoglycaemia avoidance.

5g quick acting

carbohydrate

Hypoglycaemia avoidance.

10g quick acting

carbohydrate

Hypoglycaemia treatment

15g quick acting

carbohydrate

Lift Shot (previously

Glucojuice)

1/3rd of a bottle 2/3rds of a bottle 1 bottle

Smooth orange juice 50 ml 100 ml 150 ml

Lucozade energy 60 ml 120 ml 180 ml

Lucozade sport 75 ml 150 ml 225 ml

Lift glucose tablets 1 2 4

Glucose tablets 2 3 5

Jelly babies 1 2 3

Hypoglycaemia (or predicted hypoglycaemia) 1.5 – 4 hours after meal bolus is common during pregnancy due to ‘tail’

of the large meal boluses. At this time most of the meal carbohydrate has been absorbed but there will be insulin

on board/active insulin from the meal bolus and there is a risk of further hypoglycaemia. After taking steps to

avoid hypoglycaemia using quick acting carbohydrate (figure 4), consuming a healthy carbohydrate snack to match

the insulin on board should help avoid further hypoglycaemia.

BEST PRACTICE GUIDE: USING DIABETES TECHNOLOGY IN PREGNANCY Version 2.0, March 2020

Fingerstick glucose ≥3.5 mmol/l &

consistent with sensor glucose

Fingerstick glucose <3.5 mmol/

(HYPOGLYCAEMIA)

Consume 10g QA carbohydrate

Consume 15g QA carbohydrate

After 10 min check sensor glucose

After 10 min check ngerstick glucose

Arrow(s) horizontal

or upwards

Fingerstick glucose

≥3.5 mmol/l

Fingerstick glucose

<3.5 mmol/l

Fingerstick glucose

same or higher

than previous

Arrow(s) downward

Fingerstick glucose

<3.5 mmol/l

Fingerstick glucose

≥3.5 mmol/l but lower

than previous

Check ngerstick glucose

Check ngerstick blood glucose

Consider timing of event and if further action needed

Figure 4: Guidance for avoiding and treating hypoglycaemia

Predictive low glucose alert or low glucose alert

(For those using ash glucose monitoring, follow this if sensor glucose is ≤6 mmol/l with  OR ≤5 mmol/l with  OR < 3.5 mmol/l)

<30 min after meal bolus.

Likely due to basal/

background insulin (or

pre-bolusing too early).

Meal carbohydrate

will be absorbed.

No further action No further action

No further action

1-4 hours after meal bolus.

Likely due to meal bolus.

Will have active

insulin on board

Healthy carbohydrate

snack e.g.

1-2 hours 20g

2-4 hours 10g

(or match to IOB-see text)

30-60 min after meal

bolus. Likely mismatch

between amount/timing

of insulin & carbohydrate.

Meal carbohydrate

being absorbed.

>4 hours after meal bolus.

Likely due to basal/

background insulin

HYPOGLYCAEMIA

AVOIDANCE

HYPOGLYCAEMIA

TREATMENT

BEST PRACTICE GUIDE: USING DIABETES TECHNOLOGY IN PREGNANCY Version 2.0, March 2020

• A practical approach is to have a healthy snack containing 20g carbohydrate if 1-2 hours after meal bolus or 10g

carbohydrate if 2-4 hours after meal bolus (table 8). This assumes the preceding meal was 40-60g. If the preceding

meal had a much higher carbohydrate content, and bolus insulin was matched to this, then insulin on board

will be higher and a larger amount of carbohydrate may be needed. If the preceding meal had a much lower

carbohydrate content, then insulin on board will be lower and a smaller amount of carbohydrate may be needed.

• Some women may wish to match the carbohydrate snack to the insulin on board/active insulin (the same

approach can be used after a corrective dose).

› Check or estimate ‘insulin on board/active insulin’

» CSII: check ‘insulin on board/active insulin’

» MDI: either use an App that estimates insulin on board/active insulin or estimate by assuming insulin on

board/active insulin reduces by 25% per hour (i.e. at 1 hour 75%, at 2 hours 50%, at 3 hours 25%, at 4 hours 0%).

› Consume healthy snack containing carbohydrate to match the insulin on board/active insulin MINUS 10g

(to take some account of that already consumed as QA carbohydrate in avoiding/treating hypoglycaemia)

» The easiest approach is to work out in advance the amount of carbohydrate per 1 unit insulin on board/

active insulin using the median of the mealtime ratios. Note as insulin resistance changes through

pregnancy the ratios and therefore the amount of carbohydrate required per unit of insulin on board/

active insulin will change:

Example 1: if ratios are breakfast 1:8g, lunch 1:12g, evening meal 1:9g the median ratio is 1:9g. If IOB 3

units, the woman would have a healthy snack containing (3x9)-10=17g carbohydrate

Example 2: if ratios are breakfast 1:4g, lunch 1:7g, evening meal 1:6g the median ratio is 1:6g. If IOB 3

units, the women would have a healthy snack containing (3x6)-10=8g carbohydrate

Example 3: if ratios are breakfast 1:4g, lunch 1:7g, evening meal 1:6g the median ratio is 1:6g. If IOB 1 unit,

the women would not need a snack ((1x6)-10=-4g)

» Some women will choose to work this out in real-time each time it occurs using the ratio in place at that

time of day

All these options are starting points, and may need adjusting according to individual response.

Predictive low glucose alert when there is no bolus insulin on board (i.e. > 4 hours after a meal or other bolus) is due

to basal insulin. CSII users may choose to suspend the basal for 1 hour rather than consume QA carbohydrate.

Table 8: Healthier snack choices containing carbohydrate

Carbohydrate content

Fruit

• Very small banana (weight in skin 75g)

• Small apple (weight 85g)

• Pear (weight 90g)

• Peach (weight 140g)

• Nectarine (weight 110g)

• Orange (weight 160g)

• 2 clementine/tangerine/satsuma (total weight 160g)

• 2 Kiwi (total weight 100g)

• 2 apricot/plum (total weight 110g)

• Blueberries (total weight 100g)

• Blackberries (total weight 180g)

• Strawberries (total weight 140g)

• Raspberries (total weight 200g)

• 10 grapes (total weight 70g)

• 12 cherries total weight 120g

Each about 10g

Oat cakes x 2 12 g

Ryvita x 2 12-13g

BEST PRACTICE GUIDE: USING DIABETES TECHNOLOGY IN PREGNANCY Version 2.0, March 2020

Carbohydrate content

Wholewheat / wholegrain cracker or crispbread x1 4-8g (see nutritional table on pack)

1 small slice wholemeal bread sandwich

(lled lean meat / tinned sh / cheese / salad / egg)

10g

1 medium slice wholemeal bread sandwich

(lled lean meat / tinned sh / cheese / salad / egg)

15g

1 small wholemeal roll (lled lean meat / tinned sh / cheese / salad / egg) 20g

Small pot Icelandic yogurt (150g) 11.5g

Small pot diet yogurt (120g) 12g

Small pot full fat yogurt (120g) 15g

Weight Watchers Chocolate mousse 16g

20g Popcorn (salted or natural) 9g

70% cocoa solids chocolate 30g weight e.g.

• 3 large squares eg Tesco

• 9 small squares eg Green & Blacks

• 2 extra dark Lindor chocolate balls

10g

Graze Peanut Butter & Chocolate Protein Oat bite x1 30g 12g

Nakd peanut delight bars x 1 35g 15g

Nature Valley Protein Peanut & Chocolate bars x 1 40g 10g

Using insulin suspend features in pregnancy

Some RT-CGM systems, when used with an appropriate insulin pump, have the facility for the sensor glucose

to trigger suspension and resumption of insulin delivery. Insulin suspension can be triggered when either the

sensor glucose is predicted to reach a low threshold if the sensor glucose keeps falling at the same rate (predictive

low glucose suspend) or when sensor glucose reaches a low threshold. The currently available systems are the

Medtronic Guardian Sensor 3 used in the MiniMed 640G or 670G systems (SmartGuard) and the Dexcom G6 with

the Tandem t:slim pump (Basal-IQ). The aim of such systems is to avoid, or limit the duration of, hypoglycaemia.

Insulin suspension must not be used to treat hypoglycaemia because it does not work quickly enough.

Insulin suspension may not be sufficient to avoid hypoglycaemia, particularly if there is insulin on board from a

bolus. In pregnancy, during waking hours most episodes of hypoglycaemia occur 1.5–4 hours after mealtime bolus

when there is insulin on board, due to the large pre-meal boluses required to limit the post-meal glucose peak. In

this situation, predictive suspend may not be sufficient to avoid hypoglycaemia, and the person needs to consume

quick acting carbohydrate. During sleeping hours episodes of hypoglycaemia are usually due to basal insulin and

the predictive suspends work well to avoid hypoglycaemia.

After insulin suspension, there is a risk of subsequent rebound hyperglycaemia which is particularly important

in pregnancy. Rebound hyperglycaemia is even more likely if the person consumes quick acting carbohydrate

AND the basal is suspended (sometimes called ‘double treatment’ of hypoglycaemia). To avoid ‘double treatment’,

if the person treats/avoids hypoglycaemia with quick acting carbohydrate, they should restart the basal as soon

as hypoglycaemia is treated/averted and if the predictive suspend feature is ON, the predictive alert (which will

usually prompt the person to consume quick acting carbohydrate) should usually be OFF and vice versa.

Using the MiniMed 640G or 670G systems insulin suspend features in pregnancy

MiniMed 640G or 670G systems insulin suspend features

The MiniMed 640G and 670G systems have both predictive low glucose suspend and low glucose suspend features.

Both can be programmed to be on or off at different times and the low threshold can be adjusted.

The predictive low glucose suspend feature (‘suspend before low’) is triggered when sensor glucose is predicted

to reach or fall below a level that is 1.1 mmol/l above the low glucose threshold within 30 min AND sensor glucose

is ≤3.9 mmol/l above the low glucose threshold. After a minimum 30 min suspend time, basal insulin delivery is

automatically restarted if sensor glucose is at least 1.1 mmol/l above the low limit AND sensor glucose is predicted

to be more than 2.2 mmol/l above the low threshold within 30 min. The maximum duration of insulin suspension is

BEST PRACTICE GUIDE: USING DIABETES TECHNOLOGY IN PREGNANCY Version 2.0, March 2020

2 hours. The user can manually resume basal insulin delivery at any time. After a suspend before low event, this

feature is unavailable for a period of time (30 min-4 hours depending on the user response). Note in the MiniMed

640G or 670G systems, if both alert before low and suspend before low are on, the alert is triggered when insulin

delivery is suspended (rather than 30 min before the low glucose threshold).

The low glucose suspend feature (‘suspend on low’) stops all insulin delivery when the sensor glucose reaches or falls

below the low glucose threshold. In earlier MiniMed systems (Paradigm Veo) insulin remained suspended for 2 hours

unless manually restarted. In the MiniMed 640G or 670G systems, basal insulin delivery is resumed automatically after

a suspend on low event using the same criteria as for resuming after suspend before low. After a suspend on low event,

the automatic suspend functionality is unavailable for a period of time (30 min-4 hours depending on the response).

In the MiniMed 640G or 670G systems, if either suspend before low or suspend on low feature is enabled the alert

on low is automatically turned on. If the pump is suspended and the user does not respond to the low glucose alert

within 10 min a siren sounds and an emergency message appears.

Use in pregnancy

If the low glucose suspend is triggered, the low glucose alert will sound automatically. The person should manage

as figure 4 (note insulin suspension should not be used to treat hypoglycaemia as it does not work quickly enough)

and then restart the basal as soon as hypoglycaemia has resolved.

As discussed above, in pregnancy insulin suspend features may not be enough to avoid hypoglycaemia during the

day and ‘double treatment’ of hypoglycaemia may result in hyperglycaemia.

One practical starting approach in pregnancy for the MiniMed 640G or 670G systems is:

• During usual waking hours.

› Low threshold set at 3.4-3.6 mmol/l

› Predictive low glucose alert ON (manage as figure 4)

› Predictive low glucose suspend OFF

› Low glucose suspend ON (as a safety net-should not be needed if the person responds to the predictive low

glucose alert appropriately.) If the low glucose suspend is triggered the person should manage as figure 4

and then restart the basal as soon as hypoglycaemia has resolved

• During usual sleeping hours.

› Low threshold set at 3.4-3.6 mmol/l

› Predictive low glucose alert OFF

› Predictive low glucose suspend ON

› Low glucose alarm ON automatically. If this is triggered, the person should manage as figure 4 and then

restart the basal as soon as hypoglycaemia has resolved

Using the Dexcom G6/Tandem t:slim insulin suspend features (Basal-IQ) in pregnancy

Basal-IQ system insulin suspend features

The Basal-IQ system has a combined predictive low glucose suspend and low glucose suspend feature. Unlike the MiniMed

system, both are either ON or OFF (cannot be programmed to be on or off at different times) and the low threshold is fixed.

The Basal-IQ suspends basal insulin when sensor glucose is predicted to reach or fall below 4.4 mmol/l or the actual

sensor glucose is ≤3.9 mmol/l. These values cannot be adjusted. If insulin is suspended, an active standard bolus

will complete but an extended bolus will be cancelled. Basal insulin delivery is resumed when either the sensor

value increases from the lowest point, if the prediction of going below 4.4 mmol/l is no longer valid or if insulin

delivery has been stopped for a cumulative 2 hours in the preceding 2.5 hours. These criteria are different from the

MiniMed systems and would be expected to restart insulin earlier.

Use in pregnancy

The Tandem t:slim pump became available in the UK in 2018, so we have limited experience of the Basal-IQ system

in pregnancy.

As discussed above, in pregnancy insulin suspend features may not be enough to avoid hypoglycaemia during the day

and ‘double treatment’ of hypoglycaemia may result in hyperglycaemia. The approaches suggested for the MiniMed

systems cannot be easily used with the Basal-IQ system because the Basal-IQ system cannot be programmed to be on

or off at different times and the low threshold is fixed. However, it is possible that the insulin restart criteria of the

Basal-IQ system will restart insulin earlier and therefore MAY be less likely to result in rebound hyperglycaemia.

BEST PRACTICE GUIDE: USING DIABETES TECHNOLOGY IN PREGNANCY Version 2.0, March 2020

A possible starting approach for those who chose to use the Basal-IQ system is:

• Basal-IQ system ON.

• Set the following alerts and manage as figure 4 if any are triggered. The basal will be restarted automatically

once glucose starts to rise.

• During usual waking hours.

› ‘Urgent Low Soon Alert’: ON (alerts 20 min before sensor glucose predicted to reach 3.1 mmol/l)

› ‘Urgent Low Alarm’: ON (cannot be turned off) (alarms when sensor glucose reaches 3.1 mmol/l)

› ‘Low Glucose Alert’: ON (at 3.4-4.0 mmol/l)

• During usual sleeping hours.

› ‘Urgent Low Soon Alert’: OFF

› ‘Urgent Low Alarm’: ON (cannot be turned off)

› ‘Low Glucose Alert’: OFF or ON (at 3.4-4.0 mmol/l)

Reviewing insulin suspend features in pregnancy

It is important to review how the insulin suspend functions are working for the person.

• If the suspends are happening at the same time of day or night, then the insulin settings should be adjusted.

• If hypoglycaemia is occurring despite alerts/suspends.

› review how the woman is responding to the alerts

› consider increasing the threshold, if applicable

• If there is marked rebound hyperglycaemia.

› check if the person is ‘double treating’

› if the predictive low glucose suspend is on:

» for those with reduced hypoglycaemia awareness or severe hypoglycaemia, consider the risks/benefits

before changing the threshold or turning off suspend features

» for the MiniMed 640G or 670G system,

- consider reducing the low threshold.

- consider turning the predictive low glucose suspend off (in which case increase the threshold back

up (to 3.4-3.6 mmol/l) and turn on the predictive low glucose alert, and the low glucose suspend and/

or low glucose alarm).

» for the Basal-IQ system:

- consider turning Basal-IQ off (in which case turn on the ‘Urgent Low Soon Alert’ and the ‘Low

Glucose Alert’ (at 3.4-4.0 mmol/l)).

- for those where the rebound hyperglycaemia is mainly during the day, consider manually turning

the Basal-IQ off when getting up every morning and back on when going to bed at night.

• What to do if sensor not used or not working. Consider reducing the overnight basal by 10%. Particularly if

impaired hypoglycaemia awareness, history of severe hypoglycaemia, sleeps alone, or 1 or more overnight

predictive low glucose/low glucose alerts/suspends per week.

Using sensor arrows

Using sensor arrows to modify the pre-meal bolus doses

There is some debate about whether and how to use the RT-CGM arrows to modify pre-meal insulin boluses. This may

be introduced once users have become familiar with the RT-CGM and some women may not use it at all. There are several

approaches to adjusting pre-meal boluses based on arrows. We usually use the ISF rule or the 10/20% rule (table 10).

• ISF (insulin sensitivity factor) based insulin rule: add or subtract a fixed amount of insulin from the calculated

dose based on the arrows. This is fairly straightforward. However, the amount of insulin will need to be

adjusted as insulin sensitivity changes through pregnancy.

• 10%/20% insulin rule: increase or decrease calculated bolus by 10% or 20% based on the arrows. This has the

advantage that it automatically takes account of insulin sensitivity changes through pregnancy. However, it

does involve more arithmetic. It does vary with varying meal carbohydrate content (which may or may not be

an advantage). In practice it generally gives similar results to the ISF rule, particularly in pregnancy where meal

carbohydrate content is usually 30-60g.

BEST PRACTICE GUIDE: USING DIABETES TECHNOLOGY IN PREGNANCY Version 2.0, March 2020

Table 10: Using arrows to modify pre-meal boluses

Dexcom G6

Medtronic

Guardian

Sensor 3

FreeStyle

Libre

Over past 20 min

sensor glucose has

been:

Option A

(ISF rule)

Option B

(10%/20% rule)

Calculate bolus (for food and any corrective) and:









Rising quickly

• Add 0.5 units (if TDD < 25)

• Add 1 unit (if TDD 25-60)

• Add 2 units (if TDD >60)

increase by 20%

  

Rising

• Add 0.2 units (if TDD < 25)

• Add 0.5 unit (if TDD 25-60)

• Add 1 unit (if TDD >60)

increase by 10%



No arrow



Stable or changing

slowly

No adjustment No adjustment

  

Falling

• Subtract 0.2 units (if TDD < 25)

• Subtract 0.5 unit (if TDD 25-60)

• Subtract 1 unit (if TDD >60)

reduce by 10%









Falling quickly

• Subtract 0.5 units (if TDD < 25)

• Subtract 1 unit (if TDD 25-60)

• Subtract 2 units (if TDD >60)

reduce by 20%

Using sensor arrows to guide healthy snacking

Eating 10-20g healthy carbohydrate snack at 60-90 min post-meal can “absorb” the insulin action tail of the meal-

time insulin to avoid a low glucose later. This can be fine-tuned using information from the sensor. For example,

if at 1-2 hours post-meal the sensor glucose is in target with up arrows the snack may not be needed (see table 12).

Healthy snacking is discussed in section ‘using sensor data to improve glycaemic control in pregnancy: diet, activity

and insulin considerations’.

Using sensor arrows to guide corrective insulin doses

The sensor arrows can be used to decide whether or not to give a corrective dose. For example, we usually advise

not to give a corrective dose if arrows show glucose is falling (table 12).

A note on hybrid closed loop systems in pregnancy

Three hybrid closed loop systems are either currently licenced for use in the UK or likely to become available in the

near future.

The MiniMed 670G system has a hybrid closed loop feature (Auto Mode) and is available in the UK at the time

of writing. This is not suitable for use in pregnancy because the glucose target of 6.7 mmol/L is set too high for

pregnancy and cannot be changed.

The Control-IQ system (using the Dexcom G6 sensor and Tandem t:slim pump) is not available in the UK at the

time of writing. The target glucose for the Control-IQ system is set at 6.7 mmol/L, which is too high for pregnancy.

However, it has a “sleep mode” feature which uses a lower glucose target of 6.1 mmol/L. During pregnancy, the “sleep

mode” feature could be applied 24 hours a day (set from midnight to 23.59h) although the target of 6.1 mmol/l may

still be too high for pregnancy.

The Cambridge Artificial Pancreas system (Cam APS FX) (using the Cambridge algorithm, Dexcom G6 sensor and

Dana pump) received a CE mark for use in pregnancy in 2020. It is currently used in research clinic settings (http://

www.uea.ac.uk/aidapt/home) and may be available in selected NHS clinics (https://camdiab.com). In addition to

automated basal insulin delivery, it also allows the user to increase or decreased basal insulin delivery (using the

‘boost’ or ‘ease-off’ functions). The glucose target can be set by the user, with current recommendations to aim for

5.5 mmol/l in early pregnancy, lowering to 5.0 mmol/L after 16-20 weeks.

BEST PRACTICE GUIDE: USING DIABETES TECHNOLOGY IN PREGNANCY Version 2.0, March 2020

Box 3: Starting RT-CGM or ash glucose monitoring during pregnancy

Sensor start visit: key issues to cover

Many of the key issues are directed at avoiding over-reaction to sensor glucose readings.

• Prepare for the psychological impact of seeing ‘out of range’ sensor glucose readings. This is particularly

important for women starting glucose sensors in pregnancy.

• When sensor use will be stopped (usually at 3 months after birth, unless the woman fulfils criteria for

RT-CGM or flash glucose monitoring outside of pregnancy).

• Sensor glucose may be different from CBG (due to sensor lag and sensor accuracy). If different the

fingerstick CBG should be considered accurate.

• Tips for calibration, if the system requires it

• When to check fingerstick glucose

• When to check sensor glucose

• What the arrows mean

• Set up low glucose alerts but usually keep all other alarms off.

› Dexcom G6: Urgent Low Soon, Urgent Low, Low Glucose alerts;

› Medtronic Guardian Sensor 3: Alert Before Low

• Avoiding & treating hypoglycaemia (figure 4).

› RT-CGM: responding to alerts; flash glucose monitoring: if sensor glucose is either ≤6 mmol/l with 

OR ≤5 mmol/l with , or < 3.5 mmol/l

› Do not use sensor glucose to monitor recovery from hypoglycaemia

› Hypoglycaemia at 1.5-4 hours post-meal is common in pregnancy. Emphasise the need for a healthy

carbohydrate snack after avoiding/treating hypoglycaemia at this time.

• Avoid correcting post-meal hyperglycaemia & hyperglycaemia after hypoglycaemia. If sensor glucose

still above target range at 2 hours with sensor glucose either stable or increasing (upward arrow(s)) then

it is reasonable to give a corrective dose.

• Set up on the appropriate system for viewing and sharing sensor data

Issues to cover at subsequent visits

• Setting up the high glucose alerts

• Using arrows to guide post-meal healthy snacking

• Using arrows to modify pre-meal boluses (some women, with their diabetes health care professionals,

may decide not to use this)

• Setting up insulin suspend functions (if applicable) (for those with hypoglycaemia unawareness/severe

hypos this may need to be set up at sensor start)

• Support and encourage ‘reflection in real-time’. Users will often have started to do this themselves.

• Interpreting RT-GCM downloads

• Target range is 3.5-7.8 mmol/l. Every 5% increase in time in target range improves outcomes. Aiming for

sensor glucose:

› In target range (3.5-7.8 mmol/l) at least 70% of the time (>16h 48 min per day)

› Above target (>7.8 mmol/l) less than 25% of the time (<6 h per day)

› Below target (<3.5 mmol/l) less than 4% of the time (<1 h per day)

» including < 3.0 mmol/l less than 1% of the time (<15 min per day)

• What to do if sensor not used or not working for RT-CGM (with alerts) users.

BEST PRACTICE GUIDE: USING DIABETES TECHNOLOGY IN PREGNANCY Version 2.0, March 2020

Box 4: Tips for established users of RT-CGM or ash glucose monitoring who are newly pregnant

Key issues that should be emphasised at the first pregnancy visit (and in pre-pregnancy clinic) include

things that are different in pregnancy and problems that are more likely to occur in pregnancy due to the

tight glucose targets.

• Sensor should be worn as much as possible (aiming for at least 90-100% of the time)

• Target range is 3.5-7.8 mmol/l. Every 5% increase in time in target range improves outcomes. Aiming for

sensor glucose:

› In target range (3.5-7.8 mmol/l) at least 70% of the time (>16h 48 min per day)

› Above target (>7.8 mmol/l) less than 25% of the time (<6 h per day)

› Below target (<3.5 mmol/l) less than 4% of the time (<1 h per day)

» including < 3.0 mmol/l less than 1% of the time (<15 min per day)

• There will be many ‘out of range’ sensor glucose readings and it is important to avoid over-reacting to these.

• When to check fingerstick glucose. In addition to the ‘always’ list, which is the same as outside

pregnancy, in pregnancy advise women to check fingerstick glucose pre-meals, before taking action to

avoid hypoglycaemia, or correct hyperglycaemia.

• When to check sensor glucose

• Discourage setting too many alerts/alarms. Particularly discourage setting the high glucose alert too low.

• Avoiding and treating hypoglycaemia (figure 4).

› RT-CGM: responding to alerts; flash glucose monitoring: if sensor glucose is either ≤6 mmol/l with 

OR ≤5 mmol/l with , or < 3.5 mmol/l)

› Do not use sensor glucose to monitor recovery from hypoglycaemia

› Hypoglycaemia at 1.5-4 hours post-meal is common in pregnancy. Emphasise the need for a healthy

carbohydrate snack after avoiding/treating hypoglycaemia at this time.

• If the RT-CGM system supports insulin suspend functions, review strategies for avoiding rebound

hyperglycaemia

• Avoid correcting post-meal hyperglycaemia & hyperglycaemia after hypoglycaemia. If sensor glucose still

above target range at 2 hours with sensor glucose either stable or increasing (upward arrow(s)) then it is

reasonable to give a corrective dose.

• Using arrows to guide post-meal healthy snacking

• Using arrows to modify pre-meal boluses (some women, with their diabetes health care professionals,

may decide not to use this). For those already using the ‘ISF’ rule, advise this will need to be adjusted as

insulin sensitivity changes through pregnancy

• What to do if sensor not used or not working for RT-CGM (with alerts) users. Consider reducing the overnight

basal by 10%. Particularly if impaired hypoglycaemia awareness, history of severe hypoglycaemia, sleeps

alone, or 1 or more overnight predictive low glucose/low glucose alerts/suspends per week

• Make sure set up on the appropriate system for viewing and sharing sensor data

BEST PRACTICE GUIDE: USING DIABETES TECHNOLOGY IN PREGNANCY Version 2.0, March 2020

Evidence for ash glucose monitoring before and during pregnancy

Flash glucose monitoring, like RT-CGM, measures interstitial glucose. It generates a glucose value every 15

minutes and stores this data. However, unlike RT-CGM, it only displays the glucose data when the user scans

the device and, currently, it does not have alerts/alarms. The FreeStyle Libre is the only flash glucose monitoring

system currently available.

FreeStyle Libre readings have been compared against fingerstick

capillary blood glucose measurements in pregnancy (Scott, Bilous

et al. 2018) and the FreeStyle Libre device has a CE mark for use in

pregnancy. However, there are no randomised controlled trials of

flash glucose monitoring versus fingerstick monitoring or versus

RT-CGM in pregnancy or outside pregnancy. An observational

study of women with type 1 diabetes using RT-CGM or flash

glucose monitoring in pregnancy showed that women using flash

glucose monitoring spent more time with sensor glucose below

3.5 mmol/l at all gestational ages (Kristensen, Ögge et al. 2019). It

is not clear whether this relates to baseline differences between

women using RT-CGM or flash glucose monitoring, differences in

diabetes self-management behaviours (for example women using

RT-CGM might take more action to avoid hypoglycaemia), or to

differences in sensor accuracy at lower glucose concentrations

(Murphy 2019). However, it reinforces the advice that women

using glucose sensors should check fingerstick glucose before

taking action to treat or avoid hypoglycaemia (figure 4).

Indications for ash glucose monitoring before and during pregnancy

Women preparing for pregnancy may be assessed for flash glucose monitoring using the usual criteria for adults

with diabetes (Health_Technology_Wales 2018, Healthcare_Improvement_Scotland 2018, NHS_England 2019).

Flash glucose monitoring is not covered in the current NICE guidelines (NICE 2015). However, at the time of

writing NHS England will reimburse CCGs for the cost of flash glucose sensors for pregnant women with type 1

diabetes for 12 months in total inclusive of postpartum period (NHS_England 2019). In Northern Ireland, Scotland

and Wales, pregnant women with type 1 diabetes would meet criteria for flash glucose monitoring due to the

frequency of glucose checking required (Health_Technology_Wales 2018, Healthcare_Improvement_Scotland 2018).

Information provided by ash glucose monitoring

Like RT-CGM, flash glucose monitoring displays sensor glucose, direction of change arrows (table 5) and graphs.

Unlike RT-CGM, current information is only provided when the user scans the device.

Flash glucose monitoring set-up in pregnancy

The FreeStyle Libre does not require calibration and, at the time of writing, does not have alerts/alarms. As it does

not have alerts, the glucose target range on the Libre reader/App should be set at 3.9-7.8 mmol/l (the Libre reader/

App will not allow the lower limit to be set below 3.9 mmol/l). (Note the target range for reports can be adjusted to

3.5-7.8 mmol/l when viewing downloads in LibreView-see section ‘An approach to the consultation and looking at

downloads…’).

Using ash glucose monitoring in pregnancy

Most of the advice for using flash glucose monitoring in pregnancy is the same as for RT-CGM. The key difference

is that flash glucose monitoring does not have alerts for predicted or actual hypoglycaemia or for hyperglycaemia.

This means it is important to scan the device regularly and understand what combination of sensor glucose and

arrows might suggest risk of hypoglycaemia.

Image used with with permission: Abbott

BEST PRACTICE GUIDE: USING DIABETES TECHNOLOGY IN PREGNANCY Version 2.0, March 2020

When to check ngerstick CBG

Women should be advised to test fingerstick CBG as for RT-CGM.

When to look at ash glucose monitoring readings

As for RT-CGM, pregnant women should look at flash glucose readings (‘scan’) on getting up in the morning, pre-

meals, 1 hour post-meals (for reflection in real-time), 2 hours post-meals (for reflection in real-time and possible

corrective action), pre-bed and if they wake in the night. It is particularly important that this is done in women

using flash glucose monitoring as there are no alerts/alarms. This means that glucose may be out of range for

a prolonged period without being noticed, particularly overnight. Women might consider setting an alarm or

planning to wake up once in the night to check sensor glucose and take action if out of target range.

Using ash glucose monitoring information to try to avoid hypoglycaemia

Flash glucose monitoring does not have predictive low glucose alert or a low glucose alert. A practical approach

is to assume that if sensor glucose is either ≤6 mmol/l with  OR ≤5 mmol/l with , hypoglycaemia is likely within

the next 15-30 min. In these situations, or if sensor glucose is <3.5 mmol/l, the person should manage this as for

response to predictive low glucose alert/low glucose alert (see figure 4).

The following are managed in the same way as for RT-GCM: hypoglycaemia detection, treatment and monitoring

recovery; responding to high sensor glucose; using arrows to modify pre-meal bolus doses; and reflection and

downloads.

BEST PRACTICE GUIDE: USING DIABETES TECHNOLOGY IN PREGNANCY Version 2.0, March 2020

USING SENSOR DATA TO IMPROVE GLYCAEMIC

CONTROL IN PREGNANCY: DIET, ACTIVITY AND

INSULIN CONSIDERATIONS

One of the ways RT-CGM works is to allow the user to see the effect of different foods, activity and insulin (dose and

timing) on their glucose levels both in real-time and on downloads. Users should be encouraged to reflect in real-time

to modify future behaviour. For example:

• my peak post-meal sensor glucose was at target, I think that is because…, next time I will do the same thing!

• my peak post-meal sensor glucose was above target, I think that is because…, next time I will…(think food,

activity, insulin).

Pregnant women using RT-CGM should be encouraged to look at their RT-CGM downloads at least twice per week.

The following guidance is based on experience gained from supporting women with diabetes optimising their glucose

levels during pregnancy. It is important to acknowledge that individual women’s experience will vary and advice

should be tailored. Some women will need to make more dietary changes than others and attention should be given

to ensure their diet meets the additional nutritional requirements for pregnancy and food safety recommendations

(women can be directed to NHS Choices for further information).

There are no specific UK recommendations for weight gain in pregnancy. International guidelines recommend

women with pre-pregnancy BMI 18.5-24.9 kg/m

gain 11-16 kg over their pregnancy, with those underweight

recommended to gain more weight and those overweight or obese recommended to gain less weight (IOM 2009) For

women who are obese, there is no evidence that a stable weight or minimal weight loss (<5kg) is associated with

adverse pregnancy outcomes. However, weight loss resulting in the presence of ketones should be avoided. The

UK guidance discourages “eating for two” and advises that an additional 200 kcals/day are required only in the last

trimester in pregnancy. Women should avoid excessive weight gain as this will lead to greater insulin resistance and

make achieving glucose levels within pregnancy targets more challenging.

Sensor data is useful to identify glucose excursions and facilitate conversations. There is something powerful about

seeing the data that helps focus the mind with curiosity rather than judgement and makes it easier to identify

habitual behaviours that might not ordinarily come to light during a consultation. It can be helpful to consider

potential changes for diet, activity and insulin.

Options for managing post-meal glucose excursions

In pregnancy the post-meal glucose targets are:

• peak (which usually occurs about 1 hour) after meals <7.8 mmol/l.

• 2 hours after meals <6.4 mmol/l.

Postprandial glucose tends to swing up faster and higher during pregnancy. The amount of bolus insulin required to

limit this peak to below 7.8 mmol/l may cause a glucose below target at 1.5-4 hours because of the duration of action

of rapid-acting insulin. Broadly speaking, the peak (1 hour) postprandial glucose is predominantly influenced by

meal carbohydrate content, type of food (glycaemic index), activity and the timing of insulin in relation to the meal.

The 2 hour postprandial glucose is predominantly influenced by the type of food (high fat or protein meals result in

prolonged hyperglycaemia), activity and amount of insulin in relation to carbohydrate.

The following guidance provides a range of strategies to limit the post-meal rise and reduce the risk of a later low

glucose and is summarised in table 11. Different women are likely to find different options practically useful.

BEST PRACTICE GUIDE: USING DIABETES TECHNOLOGY IN PREGNANCY Version 2.0, March 2020

Table 11: Options for managing post-meal glucose excursions

Peak=above target

AND

2-4hr=within target

Peak=above target

AND

2-4hr=below target

Peak=within target

AND

2-4hr=below target

Peak (1hr) =above

target

AND

2-4 hr=above target

Adjust diet

Carbohydrate quantity at each

meal (restrict to 15-60g/meal)

   

Carbohydrate type (switch to

lower GI)

   

Increase lean protein or healthy fats

   

Healthy snack containing

10-20g carbohydrate at 60-90 min

post-meal

   

Split the meal

   

Adjust activity

Activity (15 min) at < 1 hour

post-meal.

* *  

Adjust insulin

Pre-bolus: increase time between

bolus & meal by 5-10 min.

  * 

Insulin:carbohydrate ratio: adjust

by 20% to give more insulin (see

table 3)



(with other options)



(with other options)

 

‘Basal to bolus switch’

(‘superbolus’) (for CSII users)

(see box 1)

  * 

*May need to reduce insulin:CHO ratio if choose this option.

Dietary considerations for managing post-meal glucose excursions

• Total carbohydrate intake 150-180g/day achieves the best glucose levels for most women. Those who eat

more than 200g carbohydrate per day tend to see more glucose out of target. In pregnancy, we recommend a

minimum of 120g carbohydrate per day and advise pregnant women not to restrict their carbohydrate intake

below this. This is important as people with diabetes may follow lower carbohydrate diets for glycaemic

control outside of pregnancy. Women who eat less than 120g carbohydrate per day tend to see more ketone

production and their diets tend to be less nutritionally balanced.

• Limit the carbohydrate load at each occasion by spreading carbohydrate intake over 3 meals with between-meal

healthy snacks. The following carbohydrate loads work well for most:

› Breakfast 15-20g

› Mid-morning healthy snack 10-20g

› Lunch 40-60g

› Mid-afternoon healthy snack 10-20g

› Evening meal 40-60g

› (post evening meal carbohydrate-containing snack is usually not needed as evening meal tends to be higher

in protein/fat and activity levels lower)

• Accurate carbohydrate counting. People with type 1 diabetes should have access to a type 1 diabetes education

programme (ABCD-DTN-UK 2017b). However, pregnant women may not have completed such a programme,

or may need a refresher, and should be offered additional education on carbohydrate counting as needed.

Miscalculating carbohydrate by 10g can make a 2-3 mmol/l difference in glucose. Although it is not possible to

be 100% accurate, women should be encouraged to weigh foods where possible and use the tools available such

as carbohydrate counting lists, Apps (e.g. my fitness pal / Carbs & Cals) and food labels. In pregnancy, women

may choose to avoid foods where estimating carbohydrate content is more difficult e.g. avoiding mixed sources

such as risotto, pasta bakes and limiting meals out.

BEST PRACTICE GUIDE: USING DIABETES TECHNOLOGY IN PREGNANCY Version 2.0, March 2020

• Carbohydrate type: Choosing carbohydrates that are slowly digested and absorbed (low glycaemic Index - GI)

can help limit postprandial peaks and level out glucoses between meals (see appendix 1). Women should be

counselled on how to reduce the post-meal rise by actions such as: freezing bread; cooling rice / pasta / potato

and then reheating or eating cold; eating more vegetables, including pulse vegetables with carbohydrates.

• Breakfast suggestions to optimise glucose levels: breakfast time is when insulin resistance is greatest and

therefore carbohydrate is least tolerated. Women should be given examples of good low GI breakfast that

are ≤20g carbohydrate such as 1 slice of wholegrain or rye bread/toast with non-carbohydrate topping (egg

/ mushrooms / tomato / avocado / cream cheese) or non-sweetened plain yogurt (e.g. Greek style or natural

yoghurt) with a portion of fruit topped with nuts / seeds.

• Keeping a food diary. Seeing how glucose rises after meals may be unsettling for many women. Women should

be supported to see this information as useful, helping them to identify which meals / carbohydrate choices are

working well and which may be best avoided or need additional management strategies. Keeping a food diary

helps to use post-meal sensor data reflectively.

• Protein: Increasing protein quantities at meals can help aid satiety, meet the additional 6g daily requirement for

pregnancy, help level out glucose levels between meals and avoid the post-meal dip (delayed conversion protein

to glucose).

• Fat: There may be benefit to adding healthy fat (e.g. eggs, olive oil, nuts, cream cheese) with meals to limit

immediate post-meal glucose rise and extend glucose tail. This should be used with caution for women who are

overweight or obese.

• Eating carbohydrate-containing healthy snacks between meals. Eating 10-20g carbohydrate at 60-90 min post-

meal can “absorb” the insulin action tail of the meal-time insulin to avoid a low glucose later. For example, if

there is a pattern of peak glucose at target with subsequent predicted or actual hypoglycaemia, introducing a

10-20g carbohydrate healthy snack at 60-90 min post-meal can avoid hypoglycaemia. This can be fine-tuned

using information from the sensor. For example, if at 1-2 hours post-meal the sensor glucose is in target with up

arrows the snack may not be needed (see table 12).

• Dietary data from CONCEPTT showed that 44% of daily carbohydrate intake came from non-recommended

sources (e.g. confectionary, biscuits and pastries) (Neoh, Grisoni et al. 2019). Women should be counselled on

making healthy snack choices (table 8). For women who do not require additional carbohydrate between meals,

lists of carbohydrate free healthy snacks should be made available (table 9).

• Splitting the meal: If peak (1 hour) above target and 2 hour predicted/actual hypoglycaemia consider ‘splitting

the meal’. For example, if the meal is a sandwich (40g carbohydrate) and a piece of fruit (15g carbohydrate), bolus

for 55g carbohydrate, eat the sandwich and delay eating the fruit for about 60-90 min).

• Hydration: staying well hydrated will ensure optimal insulin circulation and aid glucose management. There are

no specific UK recommendations for pregnancy but 1.2 – 1.6 L per day is likely to meet requirements for most.

Table 9: Healthier snack choices with minimal carbohydrate

The following snack suggestions will have little or no eect on blood glucose levels so are good choices if you are

feeling hungry between meals.

Vegetable stick for example, cucumber, carrot, celery

Dips: hummus, salsa, cream cheese, cottage cheese, guacamole, peanut butter

Cherry tomatoes

Cubes cheese with pickled onion or gherkin

Natural nuts / seeds

Boiled egg

Olives

Sun blush / sun dried tomatoes

Mini stued peppers

Lettuce roll-ups – ll with chopped egg, tuna and mayonnaise, cream cheese, cottage cheese, grated cheese, cooked meat

(left overs from roast or bolognaise sauce), well cooked prawns

Sugar free jelly

Frozen sugar free squash ice lollies

Vegetable cup a soup

Lean ham rolled with cream cheese

BEST PRACTICE GUIDE: USING DIABETES TECHNOLOGY IN PREGNANCY Version 2.0, March 2020

Activity considerations for managing post-meal glucose excursions

• Activity post-meal is a great tool for limiting the post-meal glucose rise and can make up to a 2 mmol difference.

Women should try to be active for 15 minutes after meals and be encouraged to consider their day to make

space and time to enable this to happen. The type of activity can be walking, cycling, being active around the

house, stairs. Women who have pelvic pain may consider seated activity.

Insulin considerations for managing post-meal glucose excursions

• Pre-bolusing. Outside pregnancy if rapid-acting insulin is injected just before a meal, the average rise in glucose

can be 8-10 mmol/l whereas if rapid-acting insulin is injected 15-20 min pre-meal the average rise in glucose is 3-5

mmol/l. In late pregnancy time to peak insulin is delayed (Goudie, Lunn et al. 2014). Where possible, bolus doses

should be given:

› Up to 20 weeks of gestation: 15±10 min pre-meals

› Beyond 20 weeks of gestation: extend towards 45±15 min pre-meals

• If the sensor glucose is rising and/or above target pre-meal, some women may find it helpful to give the bolus

(for the meal and any correction) and then wait until the sensor glucose graph flattens out or glucose starts to

fall before eating

• Increasing the I:C ratio will reduce glucose levels at 1 and 2 hours post-meals (table 3)

• Basal to bolus switch (“Super Bolus”) (box 1) is an option for some women using insulin pumps.

• For CSII users, insulin should usually be given as a normal bolus, as insulin is needed upfront to limit the

post-meal rise. Occasionally, women may find that particular meals (high fat (above 40g) and/or high protein

(above 25g)) result in glucose in target at 1 hour but above target at 2 hours and beyond. One option is to avoid

such meals in pregnancy. An alternative is to increase the total bolus and use a dual wave bolus. A reasonable

starting point in pregnancy is to increase the bolus calculated from the I:C ratio by 30% and use a dual wave

(combination) bolus with 70% given initially and the remainder over 3 (2-6) hours depending on the individual’s

experience (ABCD-DTN-UK 2018).

• Morning Sickness: if women are vomiting they may feel nervous about giving their insulin early. Women should

be advised that they will still see a rise in glucose after a meal even when they have vomited. A useful strategy is

to give third to half their bolus insulin before eating and the rest when they are confident they will keep it down.

Options for managing overnight glucose

Dietary considerations for managing overnight glucose

• Eat evening meal before 19.30 hrs (or at least 3 hours before going to bed).

• Apply above guidance to ensure post evening meal glucose in target.

• Any snacks taken in the evening should be carbohydrate free (table 9).

Insulin considerations for managing overnight glucose

• Basal rates should be adjusted at least 1-2 hours before the inflection point on a CGM trace. Generally, aim for

no more than six basal time blocks per 24 hours and avoid short (less than 1-2 hour) blocks unless there is a clear

requirement. Short time blocks are not usually needed, and multiple short time blocks makes it more difficult

to see patterns and make further adjustments.

• Basal rates should be adjusted by 10-20%.

BEST PRACTICE GUIDE: USING DIABETES TECHNOLOGY IN PREGNANCY

Table 12: Using sensor glucose, sensor arrows and time in relation to meal to decide on action.

Sensor arrow Sensor glucose

Dexcom

Medtronic

Guardian

Sensor 3

Libre

Pre-meal (target 4-5.3 mM) 1h post-meal (target 4-7.8mM) 2-4 h post-meal (target 4-6.4 mM) No active insulin (target 4-5.3 mM)

Predicted

low / low

In target

Above

target

Predicted

low / low

In target

Above

target

Predicted

low / low

In target

Above

target

Predicted

low / low

In target

Above

target









Calculate bolus** & add 1 unit

(if TDD 25-60)***

Reect, check

again 1h

Reect, DON’T

correct,

check again

Reect check

again 1h

Reect,

consider

corrective

dose

Check again

Calculate

corrective

& add 1 unit

(if TDD 25-

60)***

  

Calculate bolus** & add 0.5

unit (if TDD 25-60)***

Reect, check

again 1h

Reect, DON’T

correct,

check again

Reect check

again 1h

Reect,

consider

corrective

dose

Check again

Calculate

corrective &

add 0.5 unit

(if TDD 25-

60)***



No arrow



Treat/ avoid

hypo* then

bolus

Calculate bolus**, no

adjustment

Treat/ avoid

hypo*, then

20g CHO HS

Reect, check

again 1h

Reect, DON’T

correct,

check again

Treat/ avoid

hypo*, then

10g CHO HS

Reect

Reect,

consider

corrective

dose

Treat/ avoid

hypo*

No action

Calculate

corrective

  

Treat/ avoid

hypo* then

bolus

Provided NOT predicted low,

calculate bolus** & subtract

0.5 unit (if TDD 25-60)***

Treat/ avoid

hypo*, then

20g CHO HS

Provided NOT

predicted low,

15-20 g CHO

Reect, check

again 0.5-1h

Treat/ avoid

hypo*, then

10g CHO HS

Provided NOT

predicted low,

consider 10-

15g CHO HS

Reect, check

again 0.5-1h

Treat/ avoid

hypo*

Provided NOT

predicted low,

consider 10g

CHO HS

Check again

0.5-1h









Treat/ avoid

hypo* then

bolus

Provided NOT predicted low,

calculate bolus** & subtract 1

unit (if TDD 25-60)***

Treat/ avoid

hypo*, then

20g CHO HS

Provided NOT

predicted low,

15-20g CHO

Reect, check

again 15-30

min

Treat/ avoid

hypo*, then

10g CHO HS

Provided NOT

predicted low,

consider 10-

15g CHO HS

Reect, check

again 15-30

min

Treat/ avoid

hypo*

Treat/ avoid

hypo*

Reect, check

again 15-30

min

* For hypoglycaemia treatment/avoidance see gure 4.

** Calculate bolus for food and any corrective.

*** If TDD < 25 units OR > 60 units the amount of insulin is dierent, see table 10;

CHO=carbohydrate, HS=healthy snack, TDD=total daily dose

BEST PRACTICE GUIDE: USING DIABETES TECHNOLOGY IN PREGNANCY Version 2.0, March 2020

AN APPROACH TO THE CONSULTATION AND LOOKING

AT DOWNLOADS WITH PREGNANT WOMEN USING

DIABETES TECHNOLOGY

Before consultation refer to Language Matters (NHS_England 2018). This guide provides practical examples of language

that will encourage positive interactions with people living with diabetes and subsequently positive outcomes.

The rst consultation in pregnancy

• Consider congratulating the woman on her pregnancy.

• Ask how she is finding her pregnancy, without reference to diabetes.

• Acknowledge that being pregnant can be a hard experience for many women who don’t have diabetes. As a

person who has diabetes it can be harder still.

• Explore what support the women has which may help her managing her diabetes in pregnancy. For example,

can her partner, family or friend support her. Does she have access to online groups for pregnant women with

diabetes, or face to face groups?

• Signpost to further information about pregnancy in women with type 1 diabetes (e.g. JDRF pregnancy toolkit).

• Contact details for the diabetes pregnancy team (routine and emergency).

• Share the outcomes for pregnant women with diabetes compared to the general population. The majority

of pregnant women with type 1 diabetes leave hospital after birth with a healthy baby. However, the risks

of adverse pregnancy outcomes are higher than in the general maternity population. Some women will

wish to know the statistics (NPID). Sharing this data with women right at the beginning of their pregnancy

demystifies the risks that they face. These risks can be reduced, although not completely eliminated, by

optimising glucose levels.

• Review expected care during pregnancy (frequency of appointments, scans etc) (NICE 2015).

• Review ‘non-glycaemic’ risk factors (blood pressure, smoking, medications), presence of diabetes complications

(retinopathy, nephropathy, neuropathy etc) and other interventions (offer folic acid 5mg if not already taking,

and consider aspirin to reduce the risk of pre-eclampsia) (NICE 2015). Further discussion of these issues is

outside the remit of this guide.

• Review hypoglycaemia awareness, any episodes of severe hypoglycaemia, hypoglycaemia treatment. Ensure

has glucagon at home and in date and partner/family/friends know how to use it. Review DVLA rules.

• Ensure has a blood ketone meter with in-date strips. Review sick day rules for pregnancy and that should use

fingerstick glucose when following sick day rules.

• Discuss the glucose targets in pregnancy that are associated with improved neonatal outcomes (reduced risk

of large for gestational age, neonatal hypoglycaemia and NICU admission). For women already using glucose

sensors, discuss aiming for at least 70% of time in range 3.5-7.8 mmol/l. Keep in mind that for many women these

targets seem unachievable. Emphasise that every 5% increase in time in range 3.5-7.8 mmol/l improves outcomes.

• Discuss expected changes in insulin requirements in pregnancy. At the first visit discuss particularly that

insulin requirements are likely to reduce in the first trimester.

• Document pre-pregnancy insulin doses/settings.

• For CSII users:

› Rules for the management of unexplained hyperglycaemia in pregnancy. Should use fingerstick glucose

when managing unexplained hyperglycaemia

› Must carry a rapid-acting insulin pen with them at all times, in case of set/pump failure. Should have

access to long-acting insulin pens

› Document their pump settings regularly (in case of pump failure)

› Reinforce to perform set changes early in the day, not in the evening

• Offer RT-CGM or flash glucose monitoring according to guidelines. Discuss when sensor use will be stopped

(usually at 3 months postpartum, unless the woman fulfils criteria for RT-CGM or flash glucose monitoring

outside of pregnancy).

• For established sensor users, see tips in box 4.

BEST PRACTICE GUIDE: USING DIABETES TECHNOLOGY IN PREGNANCY Version 2.0, March 2020

Issues to actively review periodically in pregnancy

There are some issues that should be actively reviewed periodically, perhaps at 12, 20 and 28 weeks of gestation.

• Ascertain if there is additional support that she needs, for example from psychological services. People with

diabetes are more than twice as likely to be diagnosed with depression (NHS_England 2014).

• Review expected care during pregnancy (appointments, scans etc) (NICE 2015)

• Review ‘non-glycaemic’ risk factors and diabetes complications (NICE 2015). Further discussion of these issues

is outside the remit of this guide.

• Discuss expected changes in insulin requirements in the upcoming weeks.

• Review hypoglycaemia awareness, severe hypoglycaemia, hypoglycaemia treatment, glucagon, DVLA rules.

• Ensure has a blood ketone meter with in-date strips. Review sick day rules for pregnancy and that should use

fingerstick glucose when following sick day rules.

• Nutritional value of snacks. Encourage healthy snacks (see section ‘Using sensor data to improve glycaemic

control in pregnancy:…’).

• For CSII users:

› Rules for the management of unexplained hyperglycaemia in pregnancy. Should use fingerstick glucose

when managing unexplained hyperglycaemia

› Must carry a rapid-acting insulin pen with them at all times, in case of set/pump failure. Should have

access to long-acting insulin pens

› Document their pump settings (in case of pump failure)

› Reinforce to perform set changes early in the day, not in the evening

• For sensor users

› What to do if sensor not used or not working for RT-CGM (with alerts) users. Consider reducing the

overnight basal by 10%. Particularly if impaired hypoglycaemia awareness, history of severe hypoglycaemia,

sleeps alone, or 1 or more overnight predictive low glucose/low glucose alerts/suspends per week

› When in hospital, RT-CGM or flash glucose monitoring SHOULD NOT be used to adjust variable rate

intravenous insulin (VRII)

Looking at the download

Data from glucose sensors have changed how we consider glycaemic control: particularly looking at the peak post-

meal glucose (rather than certain time points) and the time in, above and below target range. However, there is a

lot of information and it may be useful to have a structured approach.

• Make sure to point out all the positives. Both pregnant woman and health professionals tend to focus on what

is ‘wrong’ so it can be improved and may forget to acknowledge positives.

• Ask the woman what they think needs to be addressed: they live with their diabetes every day.

• Ask the woman when changes were last made and what they were. They may have made changes (to diet,

activity, insulin pump settings or MDI doses) and it is helpful to know this when looking at the download.

• Support real-time reflection and reflection on downloads and changes made.

• Make small achievable goals and reassure women that every 5% extra time-in-range (3.5-7.8 mmol/l) is

associated with improved neonatal outcomes.

Review of settings and structure on the download

It is helpful to sense check the settings and structure of carbohydrate intake, insulin delivery and RT-CGM because

it can help to generate ideas about what might need to be adjusted. These ideas can be checked with the woman

and using the daily graphs and changes made if needed.

Carbohydrate intake

• Total carbohydrate: recommended 150-180 g/day, minimum 120g/day. Remember the total carbohydrate

reported only includes what is entered: a considerable amount of carbohydrate may be consumed as snacks

without bolus.

• Carbohydrate per meal: in pregnancy usually recommended breakfast 15-20g, lunch and evening meal 40-60g.

BEST PRACTICE GUIDE: USING DIABETES TECHNOLOGY IN PREGNANCY Version 2.0, March 2020

Insulin delivery

• Set change every 2-3 days.

• Over-riding the bolus calculator. The reason for over-riding should be sought. Remember some of this may be

over-riding the bolus advice to account for arrows.

• Basal:bolus split:

› Up to 20 weeks of gestation 50:50 to 35:65

› Beyond 20 weeks of gestation 35:65 to 25:75

• Insulin:carbohydrate ratios (1 unit of insulin for X grams of carbohydrate):

› Up to 20 weeks of gestation: breakfast 300/TDD, other meals 400/TDD

› Beyond 20 weeks of gestation: breakfast 200/TDD, other meals 300/TDD

• Bolus calculator target: 5 mmol/l (5.5 mmol/l if problematic hypoglycaemia).

• Insulin sensitivity factor (ISF) (1 unit of insulin reduces glucose by Y mmol/l):

› Up to 20 weeks of gestation 130/TDD

› Beyond 20 weeks of gestation consider 100/TDD, particularly if women are using RT-CGM (with alerts)

› If more insulin resistant at certain points in the day, is ISF in keeping with this?

RT-CGM / ash glucose monitoring

• Check the % of sensor use. Expect at least 90% sensor use with modern systems.

• For Libre users check number of scans aiming for at least 8 per day.

• Time in target glucose range.

Current recommended CGM targets (Battelino, Danne et al. 2019) are sensor glucose:

› In target range (3.5-7.8 mmol/l) at least 70% of the time (>16h 48 min per day)

› Above target range (>7.8 mmol/l) less than 25% of the time (<6 h per day)

› Below target range (<3.5 mmol/l) less than 4% of the time (<1 h per day)

» including < 3.0 mmol/l less than 1% of the time (<15 min per day).

The aim is to achieve these targets as early as possible in pregnancy and then maintain them. To display time in

this range on the download

» Dexcom G6: in the Clarity system for each person go to ‘Interactive Report’, ‘Settings’ tab and adjust the

thresholds to 3.5-7.8 mmol/l. The information for the adjusted time in range is given in the ‘AGP’ tab

» Medtronic Guardian Sensor 3. The MiniMed 640G system does not provide information on time in

range for RT-CGM. For those using MiniMed 670G, in the CareLink system go to ‘My Profile’, ‘My

Report Settings’, ‘Glucose Settings’ and adjust the thresholds to 3.5-7.8 mmol/l. The information for the

adjusted time in range is given in ‘Assessment and Progress’ report. This will change the report settings

throughout

» Libre: in LibreView for each person go to ‘Glucose History’, ‘Glucose Reports’, ‘Report Settings’ and

adjust the report settings to 3.5-7.8 mmol/l. For Libre users focus on the time above range aiming for

<25% time >7.8mmol/L

› Compare with previous Time in Range. If Time in Range is decreasing reasons for this should be sought.

› Emphasize the amount of time spent in range and remember that every 5% increase is important (many

women will focus on the out of target readings and will get disheartened)

• For Libre users focus on the mean glucose and time above range aiming for <25% time >7.8mmol/L, again

remembering that every 5% less time above range is associated with improved neonatal outcomes.

• Target mean glucose 6.0 mmol/l. Compare with previous mean glucose.

• Check the low alert settings (depends on the system).

• Check the insulin suspend system settings (depends on the system).

• High alert settings.

› Alert on high

» During usual waking hours 12-15 mmol/l,

» During usual sleeping hours 8-10 mmol/l

› Generally no other high alerts.

• Number of alerts (for RT-CGM).

› No more than 6-8 glucose alerts per 24 hours. More than this risks ‘alert fatigue’ and suggests something

needs to be changed (which might be the insulin delivery ‘settings’, pre-emptive action to avoid the alert,

the response to the alert or the alert settings)

BEST PRACTICE GUIDE: USING DIABETES TECHNOLOGY IN PREGNANCY Version 2.0, March 2020

Review the daily graphs and agree changes

• Generally look at the download since changes were last made which may be only a few days ago. In pregnancy,

there is little point in looking at the summary data over the preceding month because insulin requirements

may have changed and it is likely that multiple changes have been made.

• Check potential issues generated by review of settings and structure of carbohydrate intake, insulin delivery

and RT-CGM/flash glucose monitoring. If confirmed on review of the trace, suggest changes. Note: if not

confirmed on review of the trace changes may not be required.

• Focus on the overnight (sleeping hours) trace.

› Are there too many alarms/suspends?

› Is the glucose in target going to bed? If not:

» Eat evening meal before 19.30 hrs (or at least 3 hours before going to bed)

» Apply above guidance to ensure post evening meal glucose in target

» Any snacks taken in the evening should ideally be carbohydrate free

› Is the trace in target overnight? The overnight targets are 4-5.3 mmol/l. Remember a trace that is flat

at 6-7.8 mmol/l will look good on the daily graphs at first sight, but is a prolonged period above target.

However, given variation in insulin requirements night-to-night it is very difficult to achieve glucose 4-5.3

mmol/l every night with current diabetes technology

» Adjust basal rate at least 1-2 hours before the graph inflection. Generally, aim for no more than six basal

time blocks per 24 hours and avoid short (less than 1-2 hour) blocks unless there is a clear requirement.

» Basal rates should be adjusted by 10-20%

» Reflect on total basal before and after changes. This is particularly important for women using CSII

where it is easy to make very small adjustments

• Focus on postprandial glucose profile.

› Targets, if achievable without causing problematic hypoglycaemia (NICE 2015), are:

» Peak (which is usually about 1 hour) after meals <7.8 mmol/l

» 2 hours after meals <6.4 mmol/l

› Strategies for managing postprandial glucose excursions are given table 11 and section ‘Using sensor data to

improve glycaemic control in pregnancy: diet, activity and insulin considerations.’

› Are there too many post-meal predictive low glucose alerts? Aim no more than 1-2 per day. If occurring

after the same meal consider:

» Introducing a healthy carbohydrate snack 30 min before usual time of predictive low glucose alert

» Basal to bolus switch (“Super Bolus”) (box 1)

• Is the action in response to the predictive low glucose alert working (i.e. are hypos being avoided without

causing rebound hyperglycaemia)?

• Check for overtreatment of hypoglycaemia. This may be too much carbohydrate, using long acting rather than

quick acting carbohydrate, ‘double treatment’ (quick acting carbohydrate AND basal suspend), using the sensor

to monitor recovery from hypos.

• Are corrections for hyperglycaemia being used appropriately (look specifically for use of post-meal corrections,

post hypoglycaemia corrections, multiple corrections and correcting when glucose is falling).

• Look at the basal requirement during the day

› In pregnancy, this is often lower than true basal requirements (because of the high bolus doses). For this

reason basal rate testing (missing a meal / or zero carbohydrate meal) is generally not used in pregnancy

Furthermore, pregnancy is a ketogenic state

› Adjust basal rate at least 1-2 hours before the graph inflection. Generally, aim for no more than six basal

time blocks per 24 hours and avoid short (less than 1-2 hour) blocks unless there is a clear requirement.

› Basal rates should be adjusted by 10-20%.

BEST PRACTICE GUIDE: USING DIABETES TECHNOLOGY IN PREGNANCY Version 2.0, March 2020

USING DIABETES TECHNOLOGY IN PARTICULAR

CIRCUMSTANCES IN PREGNANCY

Using diabetes technology in hospital

Women using insulin pump therapy (CSII) are usually safest remaining on CSII if admitted to hospital, unless

incapacitated (ABCD-DTN-UK 2017a). Women using RT-CGM or flash glucose monitoring can continue to use it

as they would outside hospital provided they are confident the sensor is working well (e.g. smooth graph, no gaps,

sensor glucose has been close to fingerstick glucose, and sensor glucose as expected in the circumstances).

RT-CGM or flash glucose monitoring SHOULD NOT be used to adjust intravenous variable rate insulin infusions.

Managing steroid induced hyperglycaemia using CSII

Corticosteroids may be given to promote fetal lung maturation. All women between 24-34 weeks of gestation

who are at risk of pre-term birth within the next 7 days should be given antenatal corticosteroids (FIGO 2019).

In women between 34 and 36 weeks of gestation at risk of pre-term birth, antenatal corticosteroids should be

discussed. In those women who are having a planned pre-labour caesarean section up to 38+6 weeks of gestation,

corticosteroids should be considered. In this situation 1-4 doses of steroid (betamethasone or dexamethasone)

are usually administered. This steroid treatment can significantly elevate blood glucose levels and women are

usually admitted for 24-48 hours (depending on the steroid regimen and impact on glucose levels) for intensive

monitoring and intravenous insulin treatment. However, in most cases women on insulin pump therapy can use

this to effectively manage the elevation in blood glucose. Such women should still be admitted, because they

may need to be started on intravenous insulin therapy if glucose remains above target despite best efforts using

insulin pump. However, they can be considered for early discharge potentially reducing overnight stays.

The response to steroids is unpredictable with increases in insulin requirements potentially ranging from 20

to 100%. The effect on glucose levels usually persists for about 24 hours after the last steroid dose. The key to

successfully managing blood glucose levels over this period is frequent monitoring: at least hourly during the day,

and two hourly overnight. This is greatly facilitated by use of RT-CGM or flash glucose monitoring.

We recommend:

• The obstetric team should inform the diabetes team before (or as soon as possible after) steroids are started.

The diabetes team will:

› discuss with the woman whether she wishes to try to manage steroid induced hyperglycaemia using her

own pump

› provide guidance (outlined below)

› prescribe a VRII (without glucose) to be added if glucose remains above target despite best efforts

using insulin pump. The VRII should be individualised, taking into account the woman’s usual insulin

requirements and that she is continuing on CSII

• If the woman chooses to manage the steroid induced hyperglycaemia using her own pump, she will be

responsible for the management of her glucose testing and pump.

• Check glucose hourly (2 hourly overnight). Target 3.5-7.8 mmol/l. Glucose can be monitored using the sensor,

but women should check a fingerstick glucose at least every 4 hours.

• At 4-6 hours after first steroid dose, start a 50% temporary basal rate increase (150% of usual basal insulin

infusion rate). This may need to be adjusted.

• At 4-6 hours after first steroid dose, increase mealtime bolus doses by 50%. This is best achieved by using the

bolus calculator as normal (which calculates the insulin dose for the meal and any correction required), then

adding 50% to the recommended dose and overriding.

• Additional corrective doses should be given using the woman’s usual insulin sensitivity factor (we do not

routinely change ISF) and target (usually 5 mmol/l). Corrective doses may be given every hour if needed.

The woman may use the bolus calculator to work out the corrective dose, but may need to override the

recommended dose (as this will take into account the insulin on board).

BEST PRACTICE GUIDE: USING DIABETES TECHNOLOGY IN PREGNANCY Version 2.0, March 2020

• VRII (without glucose) should be added:

› If glucose levels are not within target after two correction boluses (assessed 1 hour after 2nd correction)

› If blood ketones ≥0.5 mmol/l (blood ketones should be checked if glucose > 10 mmol/l)

• If VRII (without glucose) is required:

› CSII should be continued, including boluses for food, but not corrective doses

› The VRII should be individualised, taking into account the woman’s usual insulin requirements and that

she is continuing on CSII. The VRII should ideally be written in advance.

› Fingerstick glucose (NOT sensor glucose) should be used to adjust VRII. Midwives are responsible for

ensuring fingerstick glucose is checked hourly and for adjusting VRII. The woman continues to be

responsible for managing her CSII.

• The woman’s standard insulin regimen can usually be reinstated 24 hours after the last steroid dose, but

occasionally the increased doses need continuing for up to 72 hours if blood glucose levels remain elevated.

Using diabetes technology before and during birth

Women should usually continue their recommended glucose monitoring and insulin delivery regimen until in

established labour or fasting prior to caesarean section.

Women on CSII may be converted to VRII plus glucose for labour or in preparation for a caesarean birth (traditional

management). Women who choose, may continue to use their CSII during labour, or fasting prior to a caesarean,

and birth , provided their glucose levels are within an acceptable range (see figure 5) and the woman / birth partner

are able and willing to manage their CSII.

The safety of using insulin pump therapy through delivery is well-established with evidence that continuing

insulin pump therapy is at least as good in maintaining glycaemic control as switching to intravenous insulin

(Drever, Tomlinson et al. 2016). Furthermore, the ability to switch to a pre-programmed basal rate after birth

means that women continuing on their pump are less likely to experience the marked variability in glucose levels

often seen postpartum in those who have been on an intravenous insulin infusion. Pump manufacturers do not

recommend the use of insulin pumps in the operating theatre but in our experience there is no problem with

continuing insulin pump therapy through a caesarean section. However, if there are concerns, the long insulin

action time in late gestation means that the insulin pump can be disconnected immediately before knife-to-skin

contact, provided the pump is reconnected IMMEDIATELY postpartum and is disconnected for an ABSOLUTE

MAXIMUM of 60 min, thus avoiding the need for intravenous VRII.

Continuous glucose monitoring (CGM) has been shown to be accurate during delivery (Cordua, Secher et al. 2013).

Women continuing CSII (or MDI) during labour or leading up to a caesarean birth, who are using either RT-CGM

or flash glucose monitoring can continue to use this during delivery ensuring that a glucose level is recorded at

least hourly. However, if the sensor glucose is out of target range 4-7 mmol/l a capillary blood glucose level should

be checked before action is taken. In addition, capillary blood glucose should be checked at least 4 hourly. If a

VRII plus glucose is started capillary blood glucose should be checked hourly (RT-CGM or flash glucose monitoring

SHOULD NOT be used to adjust intravenous variable rate insulin infusions.)

Planning for management of glycaemic control during labour or prior to a caesarean birth

We recommend a discussion at a 34-36 week clinic appointment to plan for birth. This should include

• Prescribe individualised VRII (plus glucose).

• For CSII users, decision about whether the woman will plan to continue on CSII through labour, or prior to

caesarean, and birth or convert to VRII (plus glucose).

• For women planned to continue on CSII:

› educate the woman and her birth partner on using CSII through labour or prior to caesarean birth and

provide written information

› check birth partner is able to set temporary basal rates and give correctives through the pump

› explain the responsibilities of the woman, her birth partner and staff. Women should understand that

most staff on labour ward will not be familiar with insulin pumps.

› document the insulin sensitivity (correction) factor to be used

• Determine postpartum insulin pump settings /doses (see box 5) and document in the woman’s care plan.

For CSII users, the postpartum basal rate can be programmed into the pump in advance as an additional

BEST PRACTICE GUIDE: USING DIABETES TECHNOLOGY IN PREGNANCY Version 2.0, March 2020

basal profile. For most pumps, all other settings must be programmed after birth. In the future, partial or

fully automated insulin delivery systems (closed-loop systems) may be approved for use in pregnancy and

consideration should be given regarding appropriate glucose targets for the postpartum period.

• For sensor users, explain:

› Can continue to use sensor as normal prior to established labour / fasting for caesarean section and after

birth. For women continuing CSII (or MDI) through birth, they can continue to use sensor ensuring that a

glucose level is recorded at least hourly. However, if the sensor glucose is out of target range 4-7 mmol/l a

capillary blood glucose level should be checked before action is taken. In addition, capillary blood glucose

should be checked at least 4 hourly

› If a VRII (plus glucose) is started capillary blood glucose should be checked hourly (RT-CGM or flash glucose

monitoring SHOULD NOT be used to adjust intravenous variable rate insulin infusions.)

• A checklist of what supplies need to be brought into hospital (table 13).

• Provide with hospital menu carbohydrate content.

• Diabetes team contact details and diabetes follow up arrangements for after birth.

Table 13: Checklist for labour ward bag for women using diabetes technology

Hypoglycaemia treatment of your choice

Carbohydrate and non-carbohydrate snacks

Glucose meter and strips

Hospital menu carbohydrate content

Insulin pump users

Spare sets of batteries x 2

Reservoirs / cartridges x 2

Vial of rapid-acting insulin x 1

Infusion sets (including lines) x 5 and inserter device (if using)

Insulin syringes x 10

Vial of long-acting insulin

Information about using insulin pump through birth (if applicable)

Information about post-birth pump settings

Sensor users

Spare sensor and inserter device

Transmitter charger (if applicable)

Reader/receiver/phone charger

BEST PRACTICE GUIDE: USING DIABETES TECHNOLOGY IN PREGNANCY Version 2.0, March 2020

Managing glucose levels using CSII through labour

A protocol for managing glycaemic control through labour and birth using CSII is shown in figure 5. This is based

on a protocol developed by Peter Hammond, which several UK centres have been using since at least 2013 (A

simplified version is presented in the DTN-UK guideline for managing CSII in hospitalised patients (ABCD-DTN-

UK 2017a)).

• The pump cannula and/or sensor should be sited well clear of the potential surgical field in case of emergency

or planned caesarean section. The pump is best sited just below the rib cage posteriorly.

• An individualised VRII (plus glucose) should be prescribed in advance in case it is required.

• Women should have a cannula inserted.

• Glucose should be checked and recorded hourly. This can be sensor glucose. However, if the sensor glucose

is out of target range 4-7 mmol/l a capillary blood glucose level should be checked before action is taken. In

addition, capillary blood glucose should be checked at least 4 hourly.

• The woman should continue her usual basal infusion rates and give correction doses through the pump using

individualised ISF and target 5 mmol/l.

• Target glucose range 4-7 mmol/l.

• If glucose remains above 7 mmol/l despite giving correction doses following the protocol, an individualised VRII

(plus glucose) should be started and CSII stopped. Remove CSII and tubing and place in a suitable container

(there is no need to turn off CSII nor to remove the subcutaneous cannula).

• If the woman or birth partner is unable or unwilling to manage CSII an individualised VRII (plus glucose)

should be started and CSII stopped.

• Immediately after birth, basal rates should be reduced to the planned postpartum basal rates, or a 50%

temporary basal rate can be used. As soon as possible after birth (and certainly before the first bolus) the

woman MUST change the bolus calculator settings to her postpartum settings.

• If the VRII plus glucose is used, insulin rates should be halved at birth. CSII can be restarted once the woman

is able to self-manage the pump. All settings should be changed to the planned postpartum settings. The VRII

should continue for 60 minutes after restarting CSII.

Managing glucose levels using CSII through planned caesarean birth

Women who are fasting overnight prior to a planned caesarean birth are advised to check their glucose

at 3am and on waking in the morning and take corrective action if glucose out of target range 4-7 mmol/l.

Hypoglycaemia should be treated with oral quick acting carbohydrate, and women should inform the

anaesthetist if this has been necessary. After an episode of hypoglycaemia during fasting, consider reducing the

basal rate using a temporary basal rate setting. On the morning of the planned caesarean, the protocol in figure 5

should be followed from waking.

Responsibilities of sta, the woman and her birth partner for women continuing CSII through birth

The key to successful use of insulin pump therapy during labour, or prior to caesarean, and birth is to have a clear

protocol which all staff on the labour ward are aware of, including not only the obstetric staff but other staff who

may be involved, such as anaesthetists. The birth partner should be closely involved in planning for what is going

to happen at the time of birth so that they are able to manage the pump, if needed.

While the woman remains on CSII, the woman and her birth partner are responsible for checking and documenting

glucose hourly, giving correction via CSII, adjusting basal rates and pump settings as required including after birth.

The midwife is responsible for ensuring the woman / birth partner remains able and willing to manage the CSII,

that glucose is checked and documented hourly, and that if glucose is persistently (see below) above 7 mmol/l, VRII

plus IV glucose is started and the CSII stopped. If the woman is on VRII plus glucose, the midwife is responsible for

checking capillary blood glucose (NOT sensor glucose) hourly and adjusting VRII rates as prescribed.

BEST PRACTICE GUIDE: USING DIABETES TECHNOLOGY IN PREGNANCY Version 2.0, March 2020

Figure 5: Protocol for managing glucose levels for women continuing on insulin

pump therapy during labour and birth or during casearean birth.

Start once in established labour, or on waking on the morning of a planned caesarean, or when made nil by

mouth for an emergency caesarean

Glucose <4 mmol/l

(conrmed on BG)

Treat hypoglycaemia

as per hospital

protocol

Following one

unexplained

hypoglycaemia,

reduce basal rate

by 25-50% using

a temporary basal

rate setting

If further episode

of hypoglycaemia,

reduce basal rate

by another 25%

Glucose

4-7 mmol/l

Glucose 4-7 mmol/l

resume usual plan

Continue at the

reduced basal rate

until birth. Check

glucose hourly.

Continue usual

basal rate. Check

glucose hourly

YES NO

Immediately after birth change to planned

postpartum basal rate or use a 50%

temporary basal AND as soon as possible

after birth (BEFORE rst bolus) change bolus

calculator settings to postpartum settings.

Start individualised VRII (plus glucose) & stop CSII (remove

CSII & tubing & place in suitable container).

Midwife is responsible for hourly FINGERSTICK blood glucose &

adjusting VRII. Immediately after birth half VRII rates.

Patient/partner able and willing to continue on CSII during birth.

Continue CSII on current settings.

Insert IV cannula in case IV insulin/uids required

Check glucose hourly. Target 4-7 mmol/l.

Sensor glucose can be used. If sensor glucose outside target range check

ngerstick blood glucose (BG) before taking action. If sensor glucose remains

within target range ngerstick BG should be checked at least 4 hourly.

Glucose

>7 mmol/l

(conrmed on BG)

Give 3rd

correction dose

Glucose 7.1-10

mmol/l (conrmed

on BG)

Give 2nd correction dose Give 2nd correction dose

Glucose 7.1-10 mmol/l (conrmed on BG)

Glucose >7 mmol/l (conrmed on BG)

Give 1st correction dose through CSII (calculated by aiming for

glucose of 5 mmol/l using individualised ISF)

Glucose >10 mmol/l (conrmed on BG)

Glucose >10 mmol/l

(conrmed on BG)

Glucose >7 mmol/l

(conrmed on BG)

Check glucose

after 1/2 hour

Check glucose after 1 hour

Check glucose after 1/2 hour

Start individualised VRII

(plus glucose) & stop CSII

(remove CSII & tubing & place

in suitable container).

BEST PRACTICE GUIDE: USING DIABETES TECHNOLOGY IN PREGNANCY Version 2.0, March 2020

USING DIABETES TECHNOLOGY AFTER BIRTH

Management immediately after birth until discharge from hospital

Insulin requirements drop immediately after birth and women are at increased risk of hypoglycaemia in the

first hours-to-days after birth. The emphasis should be on avoidance of hypoglycaemia. Several factors may

contribute including:

• Continued action of pregnancy insulin doses (due to large insulin doses and longer duration of insulin action in

late pregnancy).

• Continuing to use pregnancy insulin doses/pump settings, which may be due to not changing the settings or

not believing that the much lower recommended doses are enough.

• Continuing to aim for tight glycaemic targets of pregnancy.

• Reduced hypo awareness.

• Nausea and vomiting.

• Breastfeeding.

We recommend postpartum insulin pump settings be agreed (see box 5) and documented in the woman’s individual

care plan at 34-36 weeks. (The postpartum basal rate can be programmed into the pump in advance as an additional

basal profile. For most pumps, all other settings must be programmed after birth.).

Box 5: Postpartum insulin pump settings

(Note for the first week or so postpartum glucose target range should be 6-10 mmol/l (to avoid hypoglycaemia)

and post-meal glucose excursions up to 12-15 mmol/l are expected and acceptable.)

There are four key changes:

Insulin:carbohydrate ratio (I:C ratio)

• Pre-pregnancy settings (consider adjusting to give less insulin if tight glycaemic control pre-pregnancy)

• OR between 1:10g and 1:15g

Insulin sensitivity factor (ISF)

• Pre-pregnancy settings (consider adjusting to give less insulin if tight glycaemic control pre-pregnancy)

• OR 1 unit to reduce glucose by between 3 and 4 mmol/l

Target for bolus calculations

• Maintain single target at 5 mmol/l (consider increasing (e.g. to 6.5 mmol/l) if impaired awareness of

hypoglycaemia)

Basal rate:

The basal infusion rate should be reset according to one of the following:

• pre-pregnancy basal rate profile (consider reducing if tight glycaemic control pre-pregnancy), reduced by

20% if breastfeeding

• 50% of the basal rate profile in late pregnancy (any time after 34 weeks of gestation), reduced by a

further 20% if breastfeeding

• A total daily basal of 0.25 units/kg based on the woman’s pre-pregnancy weight or 0.2 units/kg if breastfeeding

BEST PRACTICE GUIDE: USING DIABETES TECHNOLOGY IN PREGNANCY Version 2.0, March 2020

• For women who continued to use CSII through birth, basal rates should be reduced immediately after birth to

the planned postpartum basal rates, or a 50% temporary basal rate can be used. As soon as possible after birth

(and certainly before the first bolus) the woman MUST change the pump settings to her postpartum settings.

• If the VRII plus glucose is used for birth, insulin rates should be halved at birth. CSII can be restarted once the

woman is able to self-manage the pump. The woman should change all settings to the planned postpartum

settings, re-site the pump (perform a set change) and restart the CSII. The VRII should continue for 60 minutes

after restarting CSII.

It is commonly advised to omit the bolus insulin for the first light meal after birth.

Women should be advised that for the first week or so postpartum glucose target range should be 6-10 mmol/l

(JBDS-IP 2017) (to avoid hypoglycaemia) and post-meal glucose excursions up to 12-15 mmol/l are expected and

acceptable. This can be difficult to get used to after the tight glycaemic targets in pregnancy. Women should be

advised to check glucose routinely (by whichever method they are using) when they wake in the morning, pre-

meals, (2 hours post-meals for those using sensors), pre bed and when getting up in the night. Additional glucose

checks may be required around breastfeeding.

Women should be advised that the recommended bolus doses via the bolus calculator will be much lower than

they are used to in pregnancy, and to ‘believe’ the bolus calculator. The diabetes team should offer to review pump

settings daily until discharge from hospital.

Management in the postpartum period

For the first few weeks at least, life with a new baby can be

unpredictable, with less routine, less sleep and demand feeding.

There continues to be a risk of hypoglycaemia. Women will

require support in transitioning back to outside of pregnancy

targets and practices and managing glycaemic control around

breastfeeding, if applicable. Doses of insulin will be lower and

this can take some getting used to. Diabetes technology can be

extremely useful in this time.

Postpartum glucose monitoring and targets are discussed in the

previous section. After the first couple of weeks postpartum, the

recommended monitoring and targets are the same as outside of

pregnancy.

Hypoglycaemia remains a risk as through the whole of pregnancy.

Women should be reminded

• Always have hypoglycaemia treatments to hand.

• Do fingerstick CBG to confirm hypoglycaemia.

• Re-check after 10-15 minutes if continue to feel symptomatic or if impaired awareness of hypoglycaemia

• Consider additional 10-15g long acting carbohydrate depending on time of day.

• Reduced hypoglycaemia awareness can be more common in pregnancy. Encouraging women to avoid

hypoglycaemia will help restore hypo awareness.

Using insulin pump therapy (CSII): dierences after pregnancy

• Meal bolus timing. Outside of pregnancy, meal bolus doses should usually be administered 15-20 minutes before

eating as this is associated with a lower postprandial glucose excursion (ABCD-DTN-UK 2018). However, we

recommend advising parents/carers of young babies to bolus immediately prior to eating to reduce the risk of

getting distracted between bolus and eating.

• Insulin pump settings (see box 5).

• Generally, wait at least 48 hours before adjusting settings (unless hypoglycaemia).

• Use temporary basal rates until patterns are more established.

• As in pregnancy, unexplained hyperglycaemia can be a sign of set failure and risk of ketosis. For management

outside of pregnancy see (ABCD-DTN-UK 2018).

Image used with permission

BEST PRACTICE GUIDE: USING DIABETES TECHNOLOGY IN PREGNANCY Version 2.0, March 2020

Using glucose sensors: dierences after pregnancy.

• Different target glucose range outside pregnancy (3.9-10 mmol/l) (Battelino, Danne et al. 2019)

• Less focus on tight post-meal targets. For the first week or so postpartum, post-meal glucose excursions up to

12-15 mmol/l are expected and acceptable.

• When to look at RT-CGM readings / scan flash glucose monitor is different outside pregnancy. Women should

be advised to check glucose routinely when they wake in the morning, pre-meals, 2 hours post-meals for those

using sensors, pre bed and when getting up in the night (at least 8 times per 24 hours).

• When to check fingerstick glucose. For systems with a non-adjunctive licence (e.g. Dexcom G6 and FreeStyle

Libre), outside of pregnancy it is not necessary to check fingerstick glucose to calculate pre-meal or other

bolus dose or to routinely monitor the accuracy of that particular sensor. The ‘ALWAYS’ list is the same as in

pregnancy.

• Low alerts can be even more important postpartum for hypo avoidance. Generally, keep low alerts/predictive

low alerts on. Consider raising the low glucose alert thresholds, particularly if impaired awareness of

hypoglycaemia. Consider changing predictive low glucose suspend settings (if using) to ON all the time (rather

than only during usual sleeping hours)

• High glucose alert thresholds should be increased to avoid alarm fatigue. They should be set at a level that may

indicate system failure (e.g. pump failure). For example:

› Usual waking hours at 15-20 mmol/l

› Usual sleeping hours at 10-13 mmol/l

Breastfeeding/expressing

Breastfeeding and expressing breast milk can cause hypoglycaemia in some women, particularly in the first few

days and when milk supplies increase (Achong, McIntyre et al. 2016)

• Postpartum CSII settings should reflect this (box 5). Pump settings will need ongoing adjustment.

• Always keep hypoglycaemia treatment within reach when breastfeeding.

• Women using glucose sensors should be advised to check their sensor glucose before breastfeeding. If sensor

glucose below 6 mmol/l advise to consume 10-15 g carbohydrate before feeding. Some may find it helpful

to check their sensor glucose every 30-60 minutes for 3 hours after the start of a breastfeeding episode to

understand their individual pattern.

• As recommended for all women who are breastfeeding/expressing, encourage healthy eating with increased

carbohydrate (JBDS-IP 2017). When breastfeeding is fully established, up to 450 extra calories per day may be

needed (JBDS-IP 2017). This can be consumed as healthy snacks containing 10-15g carbohydrate with each feed

without insulin cover which will additionally help prevent hypoglycaemia.

• A glass of milk is a good option for a 10-15g carbohydrate snack to help hydration which is also important whilst

breastfeeding.

• Reduced temporary basal rates can be considered for longer feeds but generally are too slow to take effect

and the nature of breastfeeding can be smaller more frequent feeds where carbohydrate snacks can be much

faster acting.

• When women stop breastfeeding/expressing, pump settings, particularly basal rate settings, may need to

be increased.

Postpartum follow up

Following regular appointments and healthcare professional contact during pregnancy it can be difficult for

women to suddenly have reduced contact, particularly if they are new to technology in pregnancy. Plans should be

put in place at the 34-36 week appointment for postpartum arrangements.

• Contact details for diabetes support immediately post discharge.

• 6-12 week postpartum diabetes follow up appointment.

› Review ongoing use of diabetes technology (see next section)

› Review insulin pump settings and/or RT-CGM to ensure in line with usual guidance outside of pregnancy

› For women who are breastfeeding, discuss that when they stop breastfeeding/expressing, insulin pump

settings, particularly basal rate settings, may need to be increased

› Discuss contraception and future pregnancy planning

› Refer back to usual diabetes care team

BEST PRACTICE GUIDE: USING DIABETES TECHNOLOGY IN PREGNANCY Version 2.0, March 2020

Stopping diabetes technology postpartum

• Some women may have started diabetes technology specifically for pregnancy and provision of ongoing

diabetes technology will need to be reviewed postpartum against the indications outside of pregnancy. This

should be made clear to women when they start diabetes technology.

• We recommend that diabetes technology started specifically for pregnancy should continue to be offered for 3

months postpartum.

• For those stopping diabetes technology, support should be provided to allow safe transfer. Use of technology

may deskill women in managing their diabetes without diabetes technology. They may perceive their life

without the technology much more negatively (Lawton, Blackburn et al. 2019). These factors should be taken

into account when stopping technology to minimise its impact.

BEST PRACTICE GUIDE: USING DIABETES TECHNOLOGY IN PREGNANCY Version 2.0, March 2020

SERVICE CONSIDERATIONS FOR SUPPORTING USE OF

DIABETES TECHNOLOGY IN PREGNANCY

This section explores the service requirements for supporting women with diabetes using diabetes technology

in pregnancy. It is aligned with the ABCD DTN-UK Best Practice Guide: CSII – A guide to service requirements

(ABCD-DTN-UK 2017b). The requirements for services supporting RT-CGM/flash glucose monitoring alone are

slightly different from services supporting both insulin pump therapy AND RT-CGM/flash glucose monitoring.

This is because the consequences of inappropriately managed CSII are likely to be more serious.

Workforce requirements

The core multidisciplinary team

We recommend the core multidisciplinary team (MDT) providing support for women using diabetes technology

(pumps and/or sensors) in pregnancy should include:

• Consultant Diabetologist, trained in the relevant diabetes technology and its use in pregnancy.

• Diabetes Specialist Nurse (DSN), trained in the relevant diabetes technology and its use in pregnancy.

• Diabetes Specialist Dietitian, trained in the relevant diabetes technology and its use in pregnancy.

• Obstetrician with awareness of the relevant diabetes technology as it relates to their practice.

• Midwives with awareness of the relevant diabetes technology as it relates to their practice.

• Access to clinical psychology services with interest and experience of diabetes related issues.

NICE pump guidance (NICE 2008) stipulates that the core multidisciplinary team (MDT) providing a pump service

(outside of pregnancy) should include pump trained Consultant Diabetologist, DSN, Diabetes Specialist Dietitian

and access to clinical psychology and must be in place (and trained) prior to initiation of a pump service. These

staffing requirements apply to services supporting insulin pump use in pregnancy. We recommend that for services

supporting women using insulin pumps in pregnancy, the diabetes healthcare professionals should be trained in

using insulin pumps specifically in pregnancy and the core team should also include a consultant obstetrician and

midwives with awareness of insulin pumps as it relates to their practice. There is also a requirement for a wider

framework of pump-trained or ‘pump-aware’ diabetes team members, outside the core diabetes and pregnancy

MDT, who have training in the management of pump specific problems such as set failures or pump failures and

in support for the use of insulin pumps for inpatients on labour ward or antenatal wards. Ideally the pump service

should provide a 24 hour, 7 day emergency advice service to deal with clinical problems out of hours, which may

require working with other local centres (ABCD-DTN-UK 2017b). This is particularly important for women using

pumps in pregnancy.

For services supporting RT-CGM/flash glucose monitoring alone we recommend the core MDT includes a

consultant diabetologist, DSN and diabetes dietitian trained in the use of RT-CGM/flash glucose monitoring in

pregnancy and an obstetrician and midwives with awareness of RT-CGM/flash glucose monitoring as it relates

to their practice. There is no requirement for a wider framework of RT-CGM/flash glucose monitoring trained

diabetes team members outside the core diabetes and pregnancy MDT or out of hours support specifically related

to the sensors.

Diabetes technicians or health care assistants (HCAs) as part of the MDT

A diabetes technician or HCA is a valuable member of the team, particularly for larger services, as they free up

clinician time.

Their role may include:

• Downloading pumps and sensors before the consultation, and transferring the report to the electronic patient

record, so that it is available for person with diabetes and clinician to view during the consultation.

• Teaching the person with diabetes the ‘mechanics’ of using a pump or sensor (e.g. how to insert a pump cannula

or sensor; how to navigate the menus and functions, how to link to Apps).

• Teaching the person with diabetes how to download their data at home, share their data with their clinic and

order supplies.

BEST PRACTICE GUIDE: USING DIABETES TECHNOLOGY IN PREGNANCY Version 2.0, March 2020

Workforce competencies

Consultant diabetologists, DSNs and Diabetes Dietitians should be trained in the use of diabetes technology in

pregnancy. For insulin pump therapy, we recommend the team gain initial experience outside the pregnancy

service where possible, as the team needs to be experienced particularly in dealing with pump emergencies (such as

set or pump failures) as there are potentially serious consequences for both mother and fetus. Staff competencies

need to be continually updated. Routine in-house diabetes technology updates are recommended to ensure the

whole team remains up to date.

Obstetricians and midwives in the diabetes pregnancy service, and obstetricians, midwives and anaesthetists

caring for women with type 1 diabetes on labour ward, should have an awareness of diabetes technology as it

relates to their practice. This is covered in more detail in ABCD-DTN-UK: CSII therapy in hospitalised patients

(ABCD-DTN-UK 2017a).

Key points include:

• Diabetes technology does not ‘manage’ diabetes. Women using current diabetes technology spend more time

thinking about and managing their diabetes than women using MDI and fingerstick glucose monitoring and

require diabetes self-management skills. Women using diabetes technology can still have hypoglycaemia or

hyperglycaemia.

• Insulin pump therapy (also called continuous subcutaneous insulin infusion (CSII)): people on CSII do not take

any long-acting insulin so if there is any interruption to insulin delivery then hyperglycaemia and ketoacidosis

can develop very quickly.

• Glucose sensors (real-time continuous glucose monitoring (RT-CGM) and flash glucose monitoring): there may

be a difference between fingerstick CBG and sensor glucose and sensor glucose lags behind fingerstick CBG.

Fingerstick CBG (with clean hands) should be considered accurate. Therefore:

› DO NOT use sensor glucose to check for hypoglycaemia

› DO NOT use sensor glucose to monitor recovery from hypoglycaemia

› DO NOT use sensor glucose to adjust variable rate intravenous insulin infusion

• Diabetes technology should only be adjusted by its user (who has received extensive training), her birth

partner (provided he/she has been trained by the diabetes team) or a trained member of the diabetes team. If a

pregnant woman is in hospital and she, or her birth partner, is unable to manage:

› the insulin pump (CSII): Start variable rate intravenous insulin infusion (VRII) IMMEDIATELY. Remove

CSII and tubing and place in a suitable container (there is no need to turn off the CSII nor to remove the

subcutaneous cannula).

› the sensor (RT-CGM or flash glucose monitoring): the sensor can be left in place. Institute fingerstick

capillary glucose monitoring.

Please inform the diabetes team.

Diabetes technology is expensive, so please put all removed technology in a safe place and document.

• Staff responsibilities for women using CSII through birth (see separate section).

Capacity and organisation

Capacity requirements

Estimate the numbers of women eligible for CSII, RT-CGM and flash glucose monitoring in your clinic for

forthcoming years. Estimate the numbers: already established on diabetes technology prior to planning pregnancy

or becoming pregnant; starting diabetes technology in pre-pregnancy; and starting diabetes technology in

pregnancy. Current CSII use varies between centres, but in UK centres with higher pump use, 50% of women

with type 1 diabetes are using CSII before pregnancy and 60-80% by the end of pregnancy. As of April 2019, all

pregnant women with type 1 diabetes are eligible for flash glucose monitoring and by 2020/21 all should be offered

continuous glucose monitoring.

BEST PRACTICE GUIDE: USING DIABETES TECHNOLOGY IN PREGNANCY Version 2.0, March 2020

Starting diabetes technology before and during pregnancy-service considerations

Diabetes teams should have skills and knowledge to undertake CSII and sensor initiation independently. Initiation

is usually led by a specialist nurse or a dietitian within the team.

Outside of pregnancy, diabetes technology starts are usually done in groups (ABCD-DTN-UK 2018). Women who

are planning pregnancy who fulfil criteria for diabetes technology can usually join these groups, although they

may need to be prioritised. Pregnant women can also join these groups, however are likely to require fast track

individual diabetes technology starts. For women starting sensors for pregnancy, current recommendations

envisage starting RT-CGM and/or flash glucose monitoring once women are pregnant. It is important that this is

considered in job-planning.

Both pump and sensor starts are twofold. Build in the following time/skill staffing requirements for your service

1. Teaching the person with diabetes the ‘mechanics’ of using the device (e.g. how to insert a pump cannula or

sensor; how to navigate the menus and functions, how to link to Apps).

2. Training and supporting the woman in day to day use of the technology, to make sense of the data and to make

the necessary changes herself.

Group pump start is discussed in the DTN Best Practice Guide: CSII (ABCD-DTN-UK 2018)

For one to one pump therapy initiation during pregnancy:

• Day 1 – insulin start. Allow 2 hours for the face to face visit.

• Daily contact by phone/email for the first week. Allow 1-2 hours throughout the week for this.

• Day 3 – review cannula change and complete education. Allow 1 hour. This can be done virtually for some women.

• Twice weekly (or more) contact by phone/email for the second week. Allow 30-60 min throughout the week

for this.

For one to one sensor initiation during pregnancy:

• Day 1: allow 1 - 2 hours.

• Week 2: allow 1 hour.

Follow up for women using diabetes technology before and during pregnancy

Nature of contact.

Contacts may be:

• Face to face

• Remote consultations, usually with the diabetes educators via telephone, webcam or email

Follow up for women with type 1 diabetes planning pregnancy

Women with diabetes who are planning to become pregnant should be offered pre-pregnancy care and advice

before discontinuing contraception (NICE 2015). This may be best provided in a dedicated pre-pregnancy service.

Pre-pregnancy care is discussed in detail in NICE guideline (NICE 2015). Women with type 1 diabetes planning

pregnancy should be encouraged to attend type 1 diabetes structured education. Those who are using glucose

sensors should aim for 70% in glucose range 3.9-10 mmol/l. They should be advised of the availability of diabetes

technology in pregnancy. Frequency of contact should be individualised, but likely to be monthly virtual contact

with 3 monthly face to face appointments.

Follow up for women with type 1 diabetes who are pregnant

NICE recommend that women with diabetes have contact with the joint diabetes and antenatal clinic for

assessment of blood glucose levels every 1-2 weeks throughout pregnancy (NICE 2015). Women using diabetes

technology should continue to be seen in the joint diabetes and antenatal clinic where they have access to specialist

obstetric and midwifery care as well as diabetes technology expertise. It is our experience that women may need

more frequent contact (whether or not they a using diabetes technology) particularly when insulin requirements

are changing rapidly, when the target glucose levels are not achieved or when a new treatment has been started.

BEST PRACTICE GUIDE: USING DIABETES TECHNOLOGY IN PREGNANCY Version 2.0, March 2020

For pregnant women already established on diabetes technology we recommend the clinic bases capacity requirements

on an average of weekly scheduled contact alternating face-to-face and remote consultations throughout pregnancy

plus one unscheduled contact (usually remote) per month. Some women will need more frequent and some less

frequent contact. Women starting diabetes technology in pregnancy will need more frequent contact (see below).

Face-to-face clinic appointment duration

It is important to allow adequate appointment time. (ABCD-DTN-UK 2017b). The consensus from across the

working group is:

• 30-45 minutes for certain visits, e.g.

› first visit in pregnancy (45 min)

› detailed review (to include proactive review of hypoglycaemia, sick day rules, and specific issues for CSII

and sensor users (see section ‘Issues to actively review periodically in pregnancy')) perhaps at 12, 20 and 28

weeks (30 min)

› 34-36 week visit (30-40 min) to include planning for diabetes management for birth and after birth.

• 20-30 minutes for routine follow up

Note these appointment times do not include time for downloading or obstetric/midwife review. Women with

additional medical problems or diabetes complications may require longer appointments. The appointment length

also depends on the model used for the MDT clinic detailed below. Some models (e.g. models 3 and 4 below) will

require extra capacity for ad-hoc reviews by more than one MDT member to deliver the desired efficiencies. The

clinic template must also take into account time needed for post-clinic MDT meetings (essential requirement for

models 2-5 below).

Face-to-face clinic set-up and structure

There are three main clinic types for face to face consultant led clinics for women using diabetes technology

in pregnancy

1. Women using diabetes technology seen in a general joint diabetes antenatal clinic. Such clinics include women

with pre-existing diabetes (e.g. type 1 and type 2 diabetes) and women with gestational diabetes. Care must be

taken that women using diabetes technology are seen by staff with the appropriate competencies and there is

time for longer appointments.

2. Women using diabetes technology seen in a separate joint diabetes antenatal clinic specifically for women with

pre-existing diabetes.

3. Women using diabetes technology in pregnancy seen in a diabetes clinic separate from antenatal clinic. Note

this does not meet NICE requirements for joint diabetes and antenatal clinic care, however, may be useful

particularly in early pregnancy when there is less requirement for obstetric / midwifery input.

Within this there are five models of providing the diabetes aspects of diabetes in pregnancy care (discussed in

detail in ABCD-DTN-UK Service Delivery Guide (ABCD-DTN-UK 2017b)). In brief

• Model 1: all women seen simultaneously in a joint MDT appointment with diabetes doctor, DSN and dietitian.

This is resource intensive, but may be appropriate for teams starting a new diabetes technology in pregnancy

service with small patient numbers or services where pregnant women with pre-existing diabetes are seen in a

separate clinic to those with gestational diabetes.

• Model 2: all women seen by each member of the diabetes MDT individually and sequentially at each visit in a

1-stop shop fashion. Very resource intensive.

• Model 3: women seen by one or more diabetes MDT team members at each appointment matched according to

the needs of the woman and the requirements at that stage of pregnancy. Requires: either pre-clinic or in-clinic

triage; all members of the diabetes pregnancy MDT to be able to function as diabetes educators and see diabetes

technology users independently while recognising when a woman may benefit from specific physician, nurse or

dietitian input; and the capacity to arrange reviews by additional MDT members within the clinic.

• Model 4: mixture of MDT and single clinician appointments matched according to the needs of the woman and

the requirements at that stage of pregnancy. Requirements as for model 3.

• Model 5: group diabetes educator sessions in addition to individual scheduled appointments (delivered

according to model 1-4). Group sessions in pregnancy may not be practical due to relatively small numbers at

different stages of pregnancy with different requirements. However, group sessions may be useful for starting

diabetes technology (see below).

BEST PRACTICE GUIDE: USING DIABETES TECHNOLOGY IN PREGNANCY Version 2.0, March 2020

Models 2-5 require either a post clinic multidisciplinary meeting (MDM) or the ability to discuss patients within clinic.

We have found models 3 or 4 with the option for model 2 at the first visit in pregnancy to be the most workable.

Remote consultations

Remote consultation by telephone, email or webcam are useful for supporting change between face-to-face

clinic visits. Consultations may be scheduled or ad hoc. The advantages and requirements are outlined in ABCD

DTN-UK Best Practice Guide: Service Requirements (ABCD-DTN-UK 2017b). Remote consultations require the

person with diabetes to download remotely. Remote consultations typically take 15 min however additional time

for administration (e.g. time for set-up, capturing activity, documentation etc) and review of downloads will be

required.

Clinical MDT meetings

Clinical MDT meetings are important for discussing cases and for fostering a consistent approach to diabetes

technology by the whole team. It is usually not practical to include obstetricians and midwives in a clinical MDT

meeting focussed on diabetes technology and glycaemic control and a separate clinical MDT meeting may be required.

The format, remit and frequency of a diabetes clinical MDT meeting will depend on the size of the service, clinic

model (for example model 1 may not require a separate MDM) and experience of the team. Cases may be discussed

in a specific diabetes technology and pregnancy MDM, a pregnancy and diabetes MDM (including all women

with T1DM or pre-existing diabetes whether or not using diabetes technology) or as part of a general T1DM MDM

(including non-pregnant people). The forum for discussing women planning pregnancy must also be considered.

Criteria for MDT discussion may be all pregnant (± planning pregnancy) women using, or being considered for,

diabetes technology at every MDT meeting or specific criteria such as:

• Considering starting diabetes technology.

• Considering changing diabetes technology (e.g. flash glucose monitoring to RT-CGM).

• Not reaching targets.

• Complex, challenging and difficult cases, problems or situations.

• Problematic hypoglycaemia (impaired awareness of hypoglycaemia or severe hypoglycaemia)

• Routine discussion of all pregnant women using diabetes technology at particular pre-defined times in

pregnancy.

• For those who started diabetes technology in pregnancy, plan for ongoing provision or withdrawal of diabetes

technology after pregnancy.

Clinical MDT meetings might be needed weekly, or less frequently depending on all the above. It is particularly

important to avoid delay in appropriate change in therapy in pregnancy. The MDT meetings do not replace the

need for in-clinic discussions and actions and some decisions may need to be made outside the MDT meetings,

particularly if less frequent.

Service development MDT meetings

Service development meetings specific to the diabetes technology in pregnancy service are important. These

should include

• Audit, planning and reflection (local audits and National Pregnancy in Diabetes Audit (NPID)).

• Service development.

• Safety: individual patient safety concerns, device concerns, mortality and morbidity.

• Staff development: sharing best practice, agreeing clinic policy (e.g. to decide the levels to use for sensor

alerts, use of sensor suspensions during the day, compared with overnight), guideline development, education,

training needs of the team, mentorship and quality assurance.

It is important to consider the appropriate involvement of both the wider diabetes and pregnancy team

(anaesthetists, obstetricians and midwives) and wider diabetes team.

BEST PRACTICE GUIDE: USING DIABETES TECHNOLOGY IN PREGNANCY Version 2.0, March 2020

Job planning

Job planning requirements are outlined in ABCD DTN-UK Best Practice Guide: Service Requirements (ABCD-DTN-

UK 2017b) and includes time for:

• Outpatient clinics and associated administration.

• Additional or ad-hoc face to face contacts.

• Virtual clinics and remote consultations and associated administration.

• Reviewing and analysing pump and sensor downloads.

• Preparing, supporting and running patient group education sessions.

• Clinical and service MDT meetings.

• Ongoing staff training.

In addition, for services supporting women using diabetes technology in pregnancy time is required for:

• Fast track individual diabetes technology starts.

• Support for women using diabetes technology when in hospital (e.g. for steroids or birth).

Pathways and programmes

Access to type 1 diabetes education programmes

Access to type 1 diabetes education programmes is an essential requirement for any service looking after people

with type 1 diabetes (ABCD-DTN-UK 2017b). Outside of pregnancy, people with type 1 diabetes being considered

for diabetes technology will usually have completed structured education (ABCD-DTN-UK 2017b) and women

with type 1 diabetes planning pregnancy should be encouraged to complete structured type 1 diabetes education if

they have not already done so (NICE 2015). Pregnant women may fulfil criteria for diabetes technology but have

not completed structured education. It is not reasonable to defer starting diabetes technology until the person

has completed structured education and one-to-one input will be required. Such women should be encouraged to

complete T1DM structured education either in pregnancy or after pregnancy, particularly if ongoing use of diabetes

technology is being considered.

Out of hours support pathway

As outlined in ABCD DTN-UK Best Practice Guide: Service Requirements (ABCD-DTN-UK 2017b), all pump users

should be educated on how to deal with clinical diabetes emergencies or technical problems out of hours with

written algorithms, and emergency contact details. Ideally the pump service should provide a 24 hour, 7 day

emergency advice service to deal with clinical problems out of hours, which may require working with other local

centres. This is particularly important for women using pumps in pregnancy.

Women using RT-CGM or flash glucose monitoring alone do not require out of hours support specifically related to

the diabetes technology.

Inpatient and emergency department support

The team should ensure that pathways are in place in the emergency department, antenatal and labour ward to

manage pregnant women using diabetes technology, particularly pumps, presenting as emergencies out of hours.

Processes for the referral to the diabetes pregnancy team should be in place.

System choice, informatics and data requirements

Pump & sensor choice:

• It is important that the team are comfortable and skilled in the products they are using and, while this may on

occasion lead to reduced choice of pumps available within a service, safety of the insulin pump user must be

paramount. An insulin pump service should only offer a range of pumps which they feel their team are able to

safely support. (ABCD-DTN-UK 2018). This also applies to sensors.

• Large insulin requirements in the second and third trimesters make pumps with cartridges/reservoirs of less

than 200 units more time consuming for the person with diabetes and the consumables more costly.

• You must be able to review pump and sensor data. Consider how your service is going to manage seeing

the downloads.

BEST PRACTICE GUIDE: USING DIABETES TECHNOLOGY IN PREGNANCY Version 2.0, March 2020

Data download

Access to software and IT infrastructure to download and display data from pumps, meters, RT-CGM, and flash

glucose monitoring devices in clinic is essential. This may already be in place in the diabetes clinic, but will need

additional work if the Diabetes Pregnancy clinic is held outside the usual location of the Diabetes Clinic. Consider

the work that needs to be done with your ICT team to get permission for software to be installed onto clinic

computers. Consider the training the team requires to review data presented in different ways. Consider the ease

for which people using the technology can download their own data so that it can be seen by them and shared with

you. Those using diabetes technology should be advised to download before their appointments.

The existing downloading facilities needed by services depend on the diabetes technology used:

• CareLink for Medtronic pumps, Enlite and Guardian Sensor 3 sensors. Data is stored in the ‘cloud’. Users can

download at home.

• Clarity: for Dexcom RT-CGM. For those using the Dexcom mobile phone app, data are uploaded automatically.

Those using the Dexcom receiver can download at home. Clinics have their own identity for people to ‘share’

their data to a specific diabetes department. Clinics can add a separate antenatal clinic so that the pregnancy

time-in-range glucose targets can be applied.

• Glooko/Diasend: for Insulet & Roche pumps and Dexcom (RT-CGM) receivers (but not App). Data is stored in

the ‘cloud’. Hardware and cables are needed to download in clinic. Users can download at home.

• LibreView: for Libre. For those using the LibreLink mobile phone app, data are uploaded automatically. Those

using the Libre reader, users can download at home. Clinics have their own identity for people to ‘share’ their

data to a specific diabetes department. Clinics can add a separate antenatal clinic so that the pregnancy time-

in-range glucose targets can be applied.

Clinic letters and communication

A particular challenge in pregnancy is timely documentation and communication of changes made at the frequent

contacts (both face-to-face and remote) in a way that is accessible to the woman and all the relevant health care

professionals both at face-to-face and remote reviews without generating large volumes of repeated data. The

approach is likely to be different in different services.

It may be appropriate to generate structured clinic letters at selected appointments (e.g. first visit, 12 weeks, 20

weeks, 28 weeks, 34 weeks) including a comprehensive up to date summary of all details about the pump, settings

and clinical issues so that a new clinician unfamiliar with the patient could continue their care. Other visits may

record key issues and changes made.

National Pregnancy in Diabetes Audit (NPID)

All centres in England and Wales are required to submit data to the National Pregnancy in Diabetes Audit

(NPID). This will include data on use of insulin pump therapy and glucose monitoring (RT-CGM or flash glucose

monitoring). This allows centres to benchmark their data and outcomes.

BEST PRACTICE GUIDE: USING DIABETES TECHNOLOGY IN PREGNANCY Version 2.0, March 2020

REFERENCES

ABCD-DTN-UK (2017a). Guidelines for managing continuous

subcutaneous insulin infusion (CSII, or 'insulin pump') therapy in

hospitalised patients. abcd.care/dtn-uk-best-practice-guides.

ABCD-DTN-UK (2017b). Best Practice Guide: Continuous subcutaneous

insulin infusion (CSII) - A guide to service requirements. abcd.care/

dtn-uk-best-practice-guides.

ABCD-DTN-UK (2018). Best Practice Guide: Continous subcutaneous

insulin infusion (CSII). A clinical guide for adult diabetes services. .

abcd.care/dtn-uk-best-practice-guides.

Achong, N., H. D. McIntyre, L. Callaway and E. L. Duncan (2016).

"Glycaemic behaviour during breastfeeding in women with Type 1

diabetes." Diabet Med 33(7): 947-955.

Battelino, T., T. Danne, R. M. Bergenstal, S. A. Amiel, R. Beck, T. Biester, E.

Bosi, B. A. Buckingham, W. T. Cefalu, K. L. Close, C. Cobelli, E. Dassau, J. H.

DeVries, K. C. Donaghue, K. Dovc, F. J. Doyle, 3rd, S. Garg, G. Grunberger,

S. Heller, L. Heinemann, I. B. Hirsch, R. Hovorka, W. Jia, O. Kordonouri,

B. Kovatchev, A. Kowalski, L. Laffel, B. Levine, A. Mayorov, C. Mathieu,

H. R. Murphy, R. Nimri, K. Norgaard, C. G. Parkin, E. Renard, D. Rodbard,

B. Saboo, D. Schatz, K. Stoner, T. Urakami, S. A. Weinzimer and M. Phillip

(2019). "Clinical Targets for Continuous Glucose Monitoring Data

Interpretation: Recommendations From the International Consensus

on Time in Range." Diabetes Care 42(8): 1593-1603.

Cordua, S., A. L. Secher, L. Ringholm, P. Damm and E. R. Mathiesen

(2013). "Real-time continuous glucose monitoring during labour

and delivery in women with Type 1 diabetes - observations from a

randomized controlled trial." Diabet Med 30(11): 1374-1381.

Drever, E., G. Tomlinson, A. D. Bai and D. S. Feig (2016). "Insulin pump

use compared with intravenous insulin during labour and delivery: the

INSPIRED observational cohort study." Diabetic Medicine 33(9): 1253-1259.

Feig, D. S., R. Corcoy, L. E. Donovan, K. E. Murphy, J. F. R. Barrett, J.

J. Sanchez, T. Wysocki, K. Ruedy, C. Kollman, G. Tomlinson, H. R.

Murphy and C. C. Group (2018). "Pumps or Multiple Daily Injections in

Pregnancy Involving Type 1 Diabetes: A Prespecified Analysis of the

CONCEPTT Randomized Trial." Diabetes Care 41(12): 2471-2479.

Feig, D. S., L. E. Donovan, R. Corcoy, K. E. Murphy, S. A. Amiel, K. F.

Hunt, E. Asztalos, J. F. R. Barrett, J. J. Sanchez, A. de Leiva, M. Hod,

L. Jovanovic, E. Keely, R. McManus, E. K. Hutton, C. L. Meek, Z. A.

Stewart, T. Wysocki, R. O'Brien, K. Ruedy, C. Kollman, G. Tomlinson

and H. R. Murphy (2017). "Continuous glucose monitoring in pregnant

women with type 1 diabetes (CONCEPTT): a multicentre international

randomised controlled trial." Lancet 390(10110): 2347-2359.

FIGO (2019). "FIGO Working Group on Good Clinical Practice in Maternal-

Fetal Medicine. Good clinical practice advice: Antenatal corticosteroids

for fetal lung maturation." Int J Gynaecol Obstet 144(3): 352-355.

Goudie, R. J. B., D. Lunn, R. Hovorka and H. R. Murphy (2014).

"Pharmacokinetics of Insulin Aspart in Pregnant Women With Type 1

Diabetes: Every Day Is Different." Diabetes Care 37(6): e121-e122.

Health_Technology_Wales (2018). "The clinical and cost-effectiveness

of FreeStyle Libre flash glucose monitoring for the management of

type 1 and type 2 diabetes. EAR004. www.healthtechnology.wales."

Health_Technology_Wales (2019). "Continous glucose monitoring in

pregnancy. GUI012 (10.19) www.healthtechnology.wales."

Healthcare_Improvement_Scotland (2018). "Advice Statement 009-18.

What is the clinical and cost effectiveness of FreeStyle Libre flash

glucose monitoring for patients with diabetes mellitus treated with

intensive insulin therapy? www.healthcareimprovementscotland.org".

IOM (2009). Weight Gain During Pregnancy: Reexamining the

Guidelines. http://www.nationalacademies.org/hmd/Reports/2009/

Weight-Gain-During-Pregnancy-Reexamining-the-Guidelines.aspx.

JBDS-IP (2017). Management of glycaemic control in pregnant women

with diabetes on obstetric wards and delivery units. . http://www.

diabetologists-abcd.org.uk/JBDS/JBDS_Pregnancy_final_18082017.pdf.

Jotic, A., T. Milicic, K. Lalic, L. Lukic, M. Macesic, J. Stanarcic

Gajovic, M. Stoiljkovic, M. Gojnic Dugalic, V. Jeremic and N. M. Lalic

(2020). "Evaluation of Glycaemic Control, Glucose Variability and

Hypoglycaemia on Long-Term Continuous Subcutaneous Infusion

vs. Multiple Daily Injections: Observational Study in Pregnancies

With Pre-Existing Type 1 Diabetes." Diabetes Therapy https://doi.

org/10.1007/s13300-020-00780-7.

Kristensen, K., L. E. Ögge, V. Sengpiel, K. Kjölhede, A. Dotevall, A.

Elfvin, F. K. Knop, N. Wiberg, A. Katsarou, N. Shaat, L. Kristensen

and K. Berntorp (2019). "Continuous glucose monitoring in pregnant

women with type 1 diabetes: an observational cohort study of 186

pregnancies." Diabetologia 62(7): 1143-1153.

Lawton, J., M. Blackburn, D. Rankin, C. Werner, C. Farrington, R.

Hovorka and N. Hallowell (2019). "Broadening the Debate About

Post-trial Access to Medical Interventions: A Qualitative Study of

Participant Experiences at the End of a Trial Investigating a Medical

Device to Support Type 1 Diabetes Self-Management." AJOB Empir

Bioeth 10(2): 100-112.

Morrison, F. J. R., M. Movassaghian, E. W. Seely, A. Curran, M. Shubina,

E. Morton-Eggleston, C. A. Zera, J. L. Ecker, F. M. Brown and A. Turchin

(2017). "Fetal Outcomes After Diabetic Ketoacidosis During Pregnancy."

Diabetes Care 40(7):e77-379.

Murphy, H. R. (2019). "Continuous glucose monitoring targets in type

1 diabetes pregnancy: every 5% time in range matters." Diabetologia

62(7): 1123-1128.

Neoh, S. L., J. A. Grisoni, D. S. Feig, H. R. Murphy and t. C. C. Group

(2019). "Dietary intakes of women with Type 1 diabetes before and

during pregnancy: a pre-specified secondary subgroup analysis among

CONCEPTT participants." Diabetic Medicine 10.1111/dme.13937(0).

NHS_England (2014). Action for Diabetes. January 2014. . https://www.

england.nhs.uk/rightcare/wp-content/uploads/sites/40/2016/08/act-

for-diabetes-31-01.pdf.

BEST PRACTICE GUIDE: USING DIABETES TECHNOLOGY IN PREGNANCY Version 2.0, March 2020

NHS_England (2018). Language Matters: Language and diabetes.

https://www.england.nhs.uk/publication/language-matters-language-

and-diabetes/.

NHS_England (2019). Flash Glucose Monitoring: National

Arrangements for Funding of Relevant Diabetes Patients. 7 March

2019., https://www.england.nhs.uk/wp-content/uploads/2019/03/flash-

glucose-monitoring-national-arrangements-funding-v1.1.pdf.

NICE (2008). Continous subcutaneous insulin infusion for the

treatment of diabetes mellitus. Technology appraisal guidance. TA151.

NICE (2015). Diabetes in pregnancy: management from preconception

to the postpartum period. NICE guideline. NG3.

NPID. "National Pregnancy in Diabetes Audit." from https://digital.

nhs.uk/data-and-information/clinical-audits-and-registries/national-

pregnancy-in-diabetes-audit

Pickup, J. C. and A. J. Sutton (2008). "Severe hypoglycaemia and

glycaemic control in Type 1 diabetes: meta-analysis of multiple daily

insulin injections compared with continuous subcutaneous insulin

infusion." Diabet Med 25(7): 765-774.

Rys, P. M., A. H. Ludwig-Slomczynska, K. Cyganek and M. T. Malecki (2018).

"Continuous subcutaneous insulin infusion vs multiple daily injections in

pregnant women with type 1 diabetes mellitus: a systematic review and

meta-analysis of randomised controlled trials and observational studies."

European Journal of Endocrinology 178(5): 545.

Schneider, M. B., G. E. Umpierrez, R. D. Ramsey, W. C. Mabie and K. A.

Bennett (2003). "Pregnancy Complicated by Diabetic Ketoacidosis."

Maternal and fetal outcomes 26(3): 958-959.

Scott, E. M., R. W. Bilous and A. Kautzky-Willer (2018). "Accuracy, User

Acceptability, and Safety Evaluation for the FreeStyle Libre Flash

Glucose Monitoring System When Used by Pregnant Women with

Diabetes." Diabetes Technol Ther 20(3): 180-188.

White, H. D., N. Goenka, N. J. Furlong, S. Saunders, G. Morrison, P.

Langridge, P. Paul, A. Ghatak and P. J. Weston (2014). "The UK service

level audit of insulin pump therapy in adults." Diabetic Medicine

31(4):412-418.

BEST PRACTICE GUIDE: USING DIABETES TECHNOLOGY IN PREGNANCY Version 2.0, March 2020

APPENDIX 1: CARBOHYDRATE CHOICES

This table lists foods that many pregnant women with diabetes have found result in glucose levels above target

post-meal. The ‘try instead’ list suggests some alternatives that can work well. Many foods in the ‘try instead’

list will still need to be eaten in limited quantities to avoid glucose levels above target post-meal. Individual

responses may differ

n REFINED CARBOHYDRATES TO AVOID

(High glycaemic index (GI))

o TRY INSTEAD

(Low glycaemic index (GI))

All white breads:

loaf, rolls, pitta, naan, non-traditional baguette, croissant,

chapattis, Panini, wraps.

High bre breads:

Rye bread and sourdough bread have the lowest GI.

Whole-wheat, stoneground, granary and multi-grain

varieties of breads have lower GIs.

Chapattis made with whole meal our

Freezing bread rst can help lower the GI

White our based foods:

Cakes, biscuits, cream crackers, water biscuits, Ritz, Tuc,

Yorkshire pudding, dumplings, pizza, pastry (pies, pasties,

quiche, sausage rolls, spring rolls).

Breaded & battered foods e.g. sh ngers, battered sh

Oatcakes

Whole-wheat crackers & crisp-bread

e.g. Ryvita, Cracker wheat.

Wheatmeal Digestives, Hobnobs, Hovis biscuits (one or two)

Low bre & sugar coated breakfast cereals:

Cornakes, Rice Krispies, Special K, Sugar Pus, Cocoa

Pops, sweetened muesli.

High bre cereals:

Jumbo oats

Most women don’t tolerate any cereal in pregnancy. You may

tolerate small amounts of some high bre cereals earlier in

pregnancy (up to 20 weeks): All Bran, Bran Buds, Shredded

Wheat

See breakfast guidance

Rice, pasta, grains:

No types need to be avoided.

The best rice is basmati. Brown rice & whole-wheat pasta

may give benet.

Cooling rice, pasta and potato after cooking and then eating

cold or re-heating will lower the GI

Couscous, bulgur wheat, semolina, tapioca, quinoa

Processed potato products:

Oven chips, French Fries, Smiley faces, waes, Croquettes,

frozen roast potatoes, instant potato, ready meals with

instant potato topping

Home cooked potatoes:

Boiled is best

Lightly mashed (non-instant)

Small baked potato,

Sweet potato, yam, cassava

Processed savoury snacks:

Hula Hoops, Quavers, Pringles, Monster Munch, French Fries,

Skips, baked crisps

Sliced potato crisps (e.g.Walker’s or Kettle crisps)

Ryvita snacks

Vegetable crisps

Salted or natural popcorn

Cold drinks:

Fruit juices, smoothies, full sugar squash and zzy drinks

Lucozade

Water.

Sugar free squash, sugar free carbonated drinks.

DASH water. Soda water.

Sugar:

Sugar, glucose, maltose, dextrose, honey, treacle and syrup

Articial sweeteners if a variety are used and in small

quantities

Splenda, Sweetex, Hermesetas, Nutrasweet, Candarel, Stevia

Preserves/spreads:

Jam, marmalade, honey, lemon curd, maple syrup,

chocolate spread

Marmite, Vegemite, nut butters such as peanut butter

BEST PRACTICE GUIDE: USING DIABETES TECHNOLOGY IN PREGNANCY Version 2.0, March 2020

n REFINED CARBOHYDRATES TO AVOID

(High glycaemic index (GI))

o TRY INSTEAD

(Low glycaemic index (GI))

Sweets / desserts:

Melon, mango, pineapple(some people may tolerate small

portions),

Dried fruit

Sweets, chocolates, mints

Sweet puddings

Tinned fruit in syrup

Fresh fruit, frozen fruit, tinned fruit in natural juice (juice

drained o)

Sugar free Jelly

Yogurt: natural, Greek-style, Icelandic style (high protein

such as Skyr), fruit yogurt (under 15g total carbohydrate per

portion or pot)

70% cocoa solids chocolate

Full fat ice cream (no added biscuits/caramel ripple/etc.)

Condensed, evaporated milk Crème Fraiche, cream

Ready meals/stir in sauces/take away:

Some ready meals & sauces contain signicant amounts

of sugar for example sweet & sour sauces, jar or packet

Chinese sauces, Chinese takeaway, tomato soup, baked

beans, tinned spaghetti

Reduced sugar baked beans (drain o as much sauce as

possible)

Bed-time & Malted drinks such as Ovaltine, Horlicks, drinking

chocolate.

Cadbury’s Highlight, Ovaltine Options, cocoa powder.

www.dtn-uk.care

@DTN_UK