Real time continuous glucose monitoring (rtCGM) offers hundreds of real-time interstitial glucose readings per day and allows the wearer to enable alerts for hypo- and hyperglycaemia. This allows the patient to observe daily patterns in glucose levels to improve long term glycaemic control and to make treatment decisions to avoid hypo- and hyperglycaemia. Through this mechanism rtCGM has been shown to be effective in reducing hypoglycaemia and improving glycaemic control and quality of life for patients with type 1 diabetes (46). In certain countries rtCGM is now commonly used before and throughout pregnancy and has been studied during labour and delivery (47).

A summary of the commonly used terminology in rtCGM (specific to pregnancy) is shown in Box 1.

Although rtCGM has been studied in diabetes since the 1970s and studied in diabetes in pregnancy since at least 2008 (49), the seminal trial in this field was published in 2017. The Continuous glucose monitoring in pregnant women with type 1 diabetes (CONCEPTT) study was a landmark randomized controlled trial in diabetes in pregnancy (50). In this study, 325 women with type 1 diabetes who were pregnant or planning pregnancy were randomized to either rtCGM or self-monitoring of blood glucose (SMBG). Although no improvement was seen in hypoglycemia, a reduction in hyperglycemia in the rtCGM group resulted in a greater number of hours spent in range between 3.5-7.8 mmol/L (TIR). This resulted in improved fetal outcomes including a reduction in LGA births (odds ratio 0·51, 95% CI 0·28 to 0·90; p=0·0210), a reduction in neonatal hypoglycemia (odds ratio 0·45; 0·22 to 0·89) and a reduction in time spent in neonatal intensive care units (NICU). Further post-hoc analyses of the CONCEPTT study demonstrated:

Similar observational studies have demonstrated that each additional 8.5% TIR in pregnancy correlates to a 1% or an 11 mmol/mol decrease in HbA1c (54).

These improvements in fetal outcomes prompted assessments of potential cost benefits and cost-analyses which demonstrated favorable results. The degree of cost-effectiveness differs internationally as the cost associated with NICU varies (55). For example in one Canadian study rtCGM was cost effective when paid for by the individual patient, and cost neutral when paid for by the healthcare provider or government (in this instance, the patient pays for the initial cost outlay and the government does not have to incorporate the cost of purchasing the rtCGM into their budget). In UK based cost-analysis, rtCGM was calculated to be very cost efficient (56). It is important to note that this study included the use of real time CGM rather than retrospective CGM. Retrospective CGM only allows retrospective evaluation and review of glucose readings and does not allow the patient to make changes to their insulin dose at the time of hyperglycemia. Retrospective CGM is very helpful in evaluating daily trends in hyperglycemia, however it has shown conflicting results in terms of fetal macrosomia (57).

In women with type 2 diabetes, studies of rtCGM have been small and findings are inconsistent. One study which enrolled 46 women with type 1 and 25 women with type 2 diabetes found an improvement in birth weight and macrosomia in the rtCGM group, although an exact breakdown was not given (49). A larger Danish study which enrolled 123 women with type 1 and 31 women with type 2 diabetes found no difference between rtCGM and SMBG users (58).

From a patient perspective, rtCGM appears to be generally well tolerated. The main drawbacks of rtCGM include sleep disturbance, discomfort/adhesive issues at the insertion site and false readings of hypoglycemia related to positioning (compression hypoglycemia) (59). Despite this more than 80% would recommend rtCGM and a similar proportion would use it again (60). This increased use and popularity of rtCGM also carries implications for healthcare providers (61). The analysis of rtCGM is resource intense and requires dedicated time with each patient. Patients derive the most benefit from rtCGM when it is reviewed by a clinician and adjustments can be made (62). As such the effective use of rtCGM relies on appropriately trained staff and regular patient contact.

Lastly, although rtCGM unequivocally confers benefits to women with type 1 diabetes, due to its cost it is not universally available and the inequity of available healthcare remains a major issue for people living with diabetes worldwide (63).

Intermittently scanned CGM (isCGM) or flash CGM provides immediate information about a patient’s current and predicted interstitial glucose. It differs from rtCTG as patients have to scan a wearable device with a device reader or use an app on their mobile phone (whereas rtCGM provides constant information without scanning) (64). Unlike rtCGM, first generation isCGMs do not have alarms and do not need to be calibrated against SMBG.

The glucose targets for isCGM are the same as for rtCGM and isCGM devices have also been studied and approved for use in pregnancy.

One measurement tool which is being increasingly used in discussions about real time and intermittent scanning CGM is the mean absolute relative difference (MARD). MARD is a single number which represents the accuracy of the glucose monitor. It is calculated using the difference between the CGM readings and the values measured at the same time by the reference measurement system eg central laboratory level (65).

When compared to SMBG in pregnancy, the MARD between isCGM and SMBG is 11%, and this same study found high rates of patient satisfaction (66). It is worth noting that this was a single study of 74 pregnancy women, however the MARD range is similar to that quoted by the manufactures and real life data (67, 68).

isCGM also has a number of benefits over SMBG including less hypoglycaemia and greater TIR (69, 70). Despite these advantages, discrepancies exist between isCGM and SMBG. In one study isCGM under-estimated glucose levels and potentially lead to differences in treatment decisions in up to 30% of instances (71). Finally improved fetal outcomes have not been demonstrated when compared to SMBG (69). In larger observational studies of over 300 women with type 1 diabetes, isCGM resulted in better glycaemic control in trimester 2 compared to conventional SMBG - however this translated to higher rates of neonatal hypoglycaemia and no improvement in rates of LGA or prematurity (72). The authors have suggested that the tendency of isCGM to under-read glucose readings could make patients more likely to overtreat hypoglycaemia, and thus expose themselves to rebound hyperglycaemia. It may also lead both patients and clinicians to be more cautious about insulin adjustments.

When compared to rtCGM, isCGM has some disadvantages. In one study of 20 pregnant women with type 1 diabetes, isCGM users reported a great TBR overnight and similar results have been found in more recent studies (73, 74). Although there was no significant difference in maternal or fetal outcomes such studies raise the question of using isCGM for treatment decisions, especially overnight. Other studies which evaluated clinical outcomes using older rtCGM and isCGM found better glycaemic control in the first trimester in rtCGM users. Despite better control in this critical period of organogenesis, the clinical outcomes were comparable and patients in both groups improved their glycaemic control (75).

The culmination of this evidence has resulted in the recommendation to offer isCGM to all pregnant women with PGDM who are unable or unwilling to use rtCGM (76).

Despite the clear benefits of CSII outside of pregnancy, in pregnancies complicated by type 1 diabetes the use of CSII has been associated with some disadvantages and conflicting evidence. While initial studies showed benefit in neonatal hypoglycaemia, caesarean delivery, preterm delivery and the Apgar score at five minutes (77), other results have been disappointing. Some studies have failed to demonstrate glycaemic benefits and have found that CSII users require more support and staff resources (78). Others have shown better glycaemic control and less hypoglycaemia, but this has not translated to improved fetal outcomes (79–82). One meta-analysis of four randomised and 43 observational studies found that although glycaemic control and insulin requirements were better in CSII versus multiple daily injections (MDI), rates of gestational weight gain, large and small for gestational age (LGA, SGA) and second and third trimester glycaemic control were inferior in CSII users (83). An older systematic review of RCTs found no differences between CSII and MDI (84). Other studies similarly found higher rates of gestational weight gain and LGA in CSII users despite better glycaemic control (85).

A pre-specified analysis of the CONCEPTT study found that MDI users had better glycaemic control at 24 and 34 weeks, had less hypertension and less neonatal hypoglycaemia and NICU admissions (86). A similar study which also evaluated CSII and MDI in rtCGM users found no significant difference in outcomes (87).

It is worth noting that much of this evidence comes from observational studies and the lack of RCTs in this area mean that results must be interpreted in this light.

CSII have also been studied at the time of delivery and women who consistently use CSII throughout pregnancy and delivery have better TIR than those who switch to an intravenous insulin infusion in labour. It should be noted however that the decision to use CSII during labour is multifactorial and depends on staff familiarity with different diabetes technology and the ability to frequently and reliably check for ketones if needed (88).

Telemedicine or telehealth is defined as “technology-based virtual platform to deliver various aspects of health information, prevention, monitoring, and medical care” (105). It was already increasingly used prior to the Sars-CoV-2 pandemic however its use soared during this time due to universal lockdown restrictions (106). Telemedicine offers increased convenience and flexibility and is generally acceptable to patients however to be beneficial in PGDM hard outcomes like glycaemic control and fetal outcomes need to be evaluated (107). A systematic review of the use of telemedicine in treating both PGDM and gestational diabetes (GDM) did not find any benefit in maternal or fetal outcomes, although very few studies evaluated women with PGDM (108). In studies that did include small numbers of patients with PGDM, telemedicine resulted in fewer GP and nurse visits (109), improved satisfaction and quality of life (110) and some slight improvements in glycaemic control (111).

Further evaluation is needed before telemedicine can become a routine part of care for pregnant women with PGDM.

Smart pens are reusable insulin delivery devices which can help patients track their timing and doses of insulin. This prevents inadvertent insulin delivery when patients forget about previously administered doses and improves glycaemic control by helping to reduce the frequency of missed doses. Such interventions have been shown to improve HbA1c levels and are cost-effective in non-pregnant population (112). They are increasingly used and are highly acceptable to patients and improve confidence in diabetes self-management in both type 1 and type 2 diabetes (113, 114). Although it is reasonable to assume that these benefits would translate to women with PGDM there is no evidence in this area.

Another innovation which has shown promise in the management of type 1 diabetes is the use of applications which facilitate carbohydrate counting and bolus calculation. A number of applications are available to facilitate carbohydrate counting, however more recent developments include the launch of food identification software that calculates the carbohydrate content of food using photographs (115, 116). This software can improve carbohydrate counting and HbA1c in a population of young adults, although its use is limited and it remains unvalidated in pregnancy (117).