Continuous Glucose Monitoring for Diabetes Management During Pregnancy: Evidence, Practical Tips, and Common Pitfalls

Abstract

Continuous glucose monitoring (CGM) has the potential to revolutionize diabetes management during pregnancy by providing detailed and real-time data to patients and clinicians, overcoming many of the limitations of self-monitoring of blood glucose. Although there are limited data on the role of CGM to improve pregnancy outcomes in patients with type 2 diabetes or gestational diabetes, CGM has been shown to reduce pregnancy complications in patients with type 1 diabetes. Despite the limited data in some populations, given its ease of use and recent FDA approval with expanding insurance coverage, CGM has gained widespread popularity among pregnant patients with all types of diabetes. It is critical for obstetric clinicians to understand how CGM can be successfully integrated into clinical practice. We present a practical, step-wise approach to CGM data interpretation that can be incorporated into diabetes management during pregnancy as well as common CGM pitfalls and solutions. Although technology will continue to advance with newer generation CGM devices and diabetes technology such as automated insulin delivery (not covered here), these key principles form a basic foundation for understanding CGM technology and its utility for pregnant people.

Precis:

Continuous glucose monitoring for people with diabetes is a valuable tool that can be successfully incorporated into obstetric practice using a practical, step-wise approach.

Introduction and Background

Pregestational and gestational diabetes mellitus increase the risk of adverse maternal and neonatal outcomes ^1–5. The risk of these adverse outcomes correlates with maternal glycemic control, and evidence demonstrates that these risks are reduced with treatments targeted to improve maternal glycemic control ^2,6,7. The American College of Obstetricians and Gynecologists (ACOG) and the American Diabetes Association (ADA) have recommended monitoring maternal glucose levels at least 4 times daily with self-monitoring of capillary blood glucose (SMBG)^1,5,8, commonly known as fingersticks. In pregnancy, it is recommended to monitor fasting glucose and either 1-hour or 2-hour postprandial glucose after each meal^1,5,8. However, SMBG is uncomfortable, inconvenient, and only provides snapshots of glycemic control at discrete periods during the day without assessment overnight. SMBG not only lacks the ability to provide data on glycemic response to activities such as snacks or exercise, but evidence supports issues regarding discomfort, scheduling and workplace accommodations, disruption of daily activities, limited health literacy, and challenges with equipment contribute to suboptimal adherence to the recommended SMBG plan^9–12.

Continuous glucose monitoring (CGM) systems overcome many of the limitations of SMBG with an easy-to-use, discreet, wearable device that measures interstitial glucose continuously throughout the day. All CGM systems include three basic components (Figure 1). The CGM sensor is a thin, flexible filament that is inserted under the skin into the interstitial space to measure glucose levels continuously. The CGM transmitter sits on top of the skin, collects glucose readings from the sensor and sends an average glucose to the third component, the CGM receiver or compatible smartphone, which stores and displays the glucose data for both patient and clinician use. Real-time CGM systems (rtCGM) automatically transmits and displays the glucose data on the smartphone app or CGM receiver, whereas intermittently-scanned CGM systems (isCGM) require the receiver or smartphone be physically waived over the sensor/transmitter at least once every 8 hours in order to display and store the glucose data. Currently available CGM systems provide glucose values up to every 1–5 minutes per day, totaling up to 288 to 1,440 values per day. In addition, they use intuitive arrows to provide information about the predicted direction and rate of glucose change in the immediate future allowing the individual to not only know what their glucose is currently, but anticipate glucose stability or how fast glucose is rising or falling. Finally, CGM data can be viewed by clinicians on web-based dashboards that provide summative assessments including reviewing glucose data over specific periods of time, comparing week to week glucose metrics, and multiple CGM statistical metrics.

Figure 1.

Basic components of continuous glucose monitoring systems.

Evidence for CGM in Pregnancy

Data demonstrate the superiority of CGM compared to SMBG in pregnancy for people with type 1 diabetes (T1D). In the CONCEPTT trial, a multicenter, open-label, randomized controlled trial of 215 pregnant individuals with T1D, CGM in addition to SMBG resulted in greater maternal HbA1c reduction from baseline to 34 weeks, compared to SMBG alone (6.83% to 6.35% with CGM vs 6.95% to 6.53% with SMBG, P=0.04)¹³. CGM reduced maternal hyperglycemia (27% vs 32% time above range) without increasing hypoglycemia (3% vs 4% time below range). Most notably, CGM reduced the incidence of large-for-gestational age infants (53% vs 69%, P=0.02), neonatal hypoglycemia (15% vs 28%. P=0.03), and neonatal intensive care unit admissions (27% vs 43%, P=0.02)¹³. Ultimately, CGM for management of T1D in pregnancy was found to be cost-saving from the National Health Service of England perspective given numbers needed to treat of only 6 to 8 to prevent these adverse neonatal outcomes¹⁴. As a result of the CONCEPTT trial and due to the rising use of CGM for management of T1D among non-pregnant individuals, many pregnant individuals with T1D will enter pregnancy already using CGM.

However, there are no definitive clinical trials demonstrating the role of CGM for management of type 2 diabetes (T2D) or gestational diabetes (GDM) in pregnancy. A recent systematic review and meta-analysis identified only 3 randomized controlled trials conducted from 2003 to 2015 that tested the efficacy of CGM in pregnant individuals with T2D with a combined total of only 137 participants¹⁵. While there was no evidence of benefit with CGM use in pregnant individuals with T2D, the meta-analysis was restricted by the limited amount of data available, and the authors concluded that further research was needed in the T2D population.

There are also no large, randomized clinical trials evaluating the efficacy of CGM for management of GDM. Observational studies suggest high feasibility and acceptability as well as increased detection of differences in nocturnal hyperglycemia related to higher large-for-gestational-age infants, which would be missed by SMBG, and superior detection of hypo- and hyperglycemic episodes compared with SMBG¹⁶. Ultimately, large clinical trials are required to establish whether CGM should be standard of care for the GDM population.

Despite the lack of evidence of improved pregnancy outcomes among patients with T2D or GDM at this time, CGM may still provide other benefits including improved safety with alerts, safe facilitation of rigorous multi-dose insulin adjustments, and improved adherence with glucose monitoring in circumstances where SMBG is unable to be regularly performed¹⁷.

At this time the ACOG does not have any specific recommendations about what populations should be offered or recommended CGM in pregnancy.^1,5 The ADA recommends CGM use in pregnant individuals with T1D, but emphasizes that CGM should be used in addition to pre- and postprandial SMBG and not as a substitute for SMBG.^8,17 The ADA has concluded that data are insufficient to recommend CGM for all pregnant individuals with T2D or GDM, and that the decision to use CGM in these populations should be individualized.⁸

Data to support the use of CGM to diagnose GDM in pregnancy are also limited, but actively being studied.

CGM Use in Clinical Practice

Obstetric care clinicians can integrate CGM into their practice at varying levels (Table 1)¹⁸. Professional use CGMs are practice-based CGM devices that are given to the patient from the office practice, and patients return the equipment to the office to download and analyze the data together with the patient. These devices are intended for use over a short period of time (10–14 days) to use more detailed glycemic patterning to enhance patient teaching and/or guide medication adjustments. The CGM data can be blinded or unblinded to the patient during this 10–14 day period. Viewing rtCGM data in addition to utilizing CGM features to log events, medications, activity and meals provides pregnant individuals with insight into their metabolic health. Although professional CGM use has not been well elucidated in the obstetric literature, it has been shown to improve patient engagement and adherence to diabetes care in non-pregnant populations^19,20.

Table 1:

CGM in obstetric clinical practice

	Professional use CGM	Personal use CGM	Advanced integration of CGM with insulin pump
CGM Systems	• Dexcom G6 Pro (blinded or unblinded) • Freestyle Libre Pro • Medtronic iPro2	• Dexcom G6 • Dexcom G7 • Eversense E3* • Freestyle Libre 2 • Freestyle Libre 3 • Medtronic Guardian 3	• Medtronic 670/770G & 780G with Guardian 3 & 4 • Omnipod 5 with Dexcom G6 or Freestyle Libre 2 Plus • Tandem t:slim X2 Insulin Pump with Dexcom G6 and G7 or Freestyle Libre 2 Plus
Patient demographic	• GDM, T2D, or T1D • History of obesity surgery with dumping syndrome or reactive hypoglycemia • Management challenges: patient reported hypoglycemia, significant glycemic variability (i.e. high and low glucose values same time daily), variable eating/sleeping schedule, identification of nocturnal hyperglycemia • Insufficient SMBG data to conclusively provide treatment recommendations	• GDM, T2D, or T1D • Management challenges: patient reported hypoglycemia, significant glycemic variability (i.e. high and low glucose values same time daily), variable eating/sleeping schedule, identification of nocturnal hyperglycemia • Insufficient SMBG data to conclusively provide treatment recommendations	• T1D, T2D (in select cases) on insulin therapy
Clinical provider expertise	• Basic understanding of the different CGM platforms and reports • Retrospective interpretation of CGM data • Treatment adjustments based on glycemic data • Insulin therapy†	• Intermediate understanding of the different CGM platforms and reports • Retrospective interpretation of CGM data • Treatment adjustments based on glycemic data • Skilled in basal-bolus insulin therapy†	• Advanced understanding the different CGM + pump platforms and reports • Retrospective interpretation of CGM + pump data • Insulin treatment adjustments based on CGM + pump data • Experience and extensive knowledge about automated insulin delivery system use, limitations in pregnancy, and informing patients on strategies for optimizing glycemic control • Managing with an experienced MFM or endocrinologist is advised
Office involvement	• Minimal involvement • Device placement and removal • Download CGM data • CGM billing‡	• Moderate involvement • Device placement and removal • Instructing patients on device use including CGM features • Download CGM data • Troubleshooting patient device and CGM data challenges • Educating patients on potential interfering substances that can affect accuracy (i.e. hydroxyurea, acetaminophen, vitamin C) and side effects (i.e., rashes, other skin changes, or filament fracture) • CGM pharmacy or medical durable ordering • CGM billing‡	• Extensive involvement • Device placement and removal • Download CGM + pump data • Troubleshooting patient devices and CGM data challenges • Educating patients on potential interfering substances that can affect accuracy (i.e. hydroxyurea, acetaminophen, vitamin C) and side effects • Insulin pump education including possible automated insulin delivery system use • CGM pharmacy or medical durable ordering • Pump supply ordering • CGM / pump billing‡

^*Eversense E3 system requires provider training for insertion/removal and office support for troubleshooting

T1D are recommended to have CGM systems for personal use

^†CGM provides data to personalize insulin therapy when pharmacotherapy is indicated. Being skilled or co-managing with providers skilled in insulin therapy, particularly basal-bolus therapy, can maximize achieving pregnancy-specific glycemic targets

^‡Current CGM billing codes include 95250 (Professional CGM – billed maximum of once per month), 95249 (Personal CGM Startup/Training – billed once only), and 95251 (Personal CGM Interpretation – billed maximum of once per month)

Table modified from Hirsch et al 2021¹⁸

In contrast to professional CGM, which may be used in blinded or unblinded modes, personal CGM is always unblinded so that patients can track their glucose in using rtCGM or isCGM and understand the effects of nutritional content, quality, and portion; meal-timing; intensity and duration of physical activity; insulin-timing and other medications as well as other factors such as stress and mood; cultural habits; and sleep quality and duration. Personal CGM use including both rt- and isCGM data can also be shared with providers, family or friends, which can reduce barriers to glycemic optimization when data is readily available.

Although advanced integration of CGM with insulin pumps in sensor-augmented pump therapy or hybrid closed loop therapy should be managed by an experienced MFM or endocrinologist, trained obstetricians and obstetric care providers can use professional and personal CGM in their practice. Obstetric clinicians may be personally skilled in prescribing and managing insulin or other oral medications or may utilize co-management with an MFM or endocrinologist to achieve maternal glycemic control during pregnancy.

The newest generation CGM devices provide accuracy similar or better to that of commonly used glucometers both during and outside of pregnancy.^21–23 However, CGM readings measure interstitial glucose levels while SMBG measures capillary glucose and thus the CGM reading may lag behind that of SMBG by approximately 10–15 minutes as glucose equilibrates across the two compartments (Table 2, Figure 2). CGM trend arrows should be used to anticipate a rise or fall in glucose. For example, if glucose is rising after eating a meal the SMBG result will be higher than the CGM reading. Conversely, if glucose is falling after administering insulin or exercise, the SMBG result will be lower than the CGM reading. The SMBG and CGM reading should be within an acceptable range (20–30 mg/dL) of each other when a person’s glucose is stable and within a normal range. At extreme values (high or low), there may be a greater discrepancy between SMBG and CGM readings.

Table 2 –

Common challenges with CGM technology

Pitfall or challenge	Problem	Solutions
CGM and SMBG are discordant	• CGM accuracy slightly declines at low or high blood glucose ranges (<70 or >180m/dL) • CGM sensors can misfire during placement which interferes with the accuracy of the data during the sensor wear period • Certain substances that may be used during pregnancy (acetaminophen, vitamin C >500mg, and hydroxyurea) can interfere with the accuracy of some CGM systems	• If patients are using a CGM system that can be calibrated, educate patients to check SMBG when CGM arrows are stable () to determine if SMBG and CGM continue to be >20–30mg/dL discrepant from one another • If discrepant by >20–30mg/dL while glucose levels are stable (indicated by CGM arrows (), patients can calibrate CGM sensors but should limit repeated calibration to avoid further accuracy challenges • SMBG should be performed if CGM data is discordant from their symptoms (i.e. CGM alert low glucose but asymptomatic) prior to responding with a self-treatment and use SMBG to guide management17 • Patients who continue to have discordant readings or lost CGM signals should remove and replace with a new sensor • Check medications and supplements regularly with patients
CGM lag-time	• CGM measures interstitial fluid glucose, which takes time to equilibrate with blood glucose. During rapid changes in glucose trajectory (i.e. increasing after eating a meal or decreasing from insulin correction), CGM data can lag the capillary blood glucose measured by SMBG by 10–15 minutes (Figure 2). • Patients who continue to check SMBG while wearing a CGM can get frustrated with differences in CGM readings compared to SMBG readings.	• Educating patients that CGM data 1) will not be the same as their SMBG glucose value if their glucose is rising or falling; 2) should be used to identify trends in glycemia and anticipate future glucose changes based on trajectory arrows • SMBG should be checked with any discordance between CGM data and clinical symptoms (i.e. symptoms of hypoglycemia with CGM data in normal range) • Patients should not repeatedly calibrate CGM with SMBG. Some systems (Dexcom and Medtronic) can be calibrated, but it is generally discouraged unless >20–30 mg/dL off when the glucose trajectory is stable
CGM alerts low during sleep	• CGM sensor is a metallic filament that detects signaling currents that interact within the interstitial tissue. • The signal can be disrupted when there is pressure on the sensor device falsely reading glucose in a lower range also known as a “compression low”.	• Educate patients to remove the pressure off of the sensor and notice if it resumes back up to previous baseline in the next few minutes. • SMBG should be checked with any discrepancy between CGM data and clinical symptoms before treating for hypoglycemia17.
Focusing only on time-in-range (TIR; 63–140 mg/dL) and only trying to achieve 70% TIR	• The current recommendation for TIR for pregnant patients with type 1 diabetes is >70%25. • Evidence to guide TIR targets for T2D and GDM is limited but likely possible to safely achieve higher TIR than 70%. • Only focusing on the TIR can overlook periods of hyperglycemia (i.e. fasting hyperglycemia) or significant glucose variability (i.e. swings from hypo- to hyperglycemia)	• For pregnant patients who have safely achieved 70% TIR without significant hypoglycemia based on time below range (TBR <4%)25, additional lifestyle or medication adjustments may be considered to further improve glycemic control. • Review the CGM data in the fasting period for the patient and consider adjustments if ≥95mg/dL. • Review the daily CGM profiles in order to identify repeated patterns in glucose excursions that should be addressed with treatment recommendations in contrast to one-time outliers. • Assess mean glucose with goal of <120mg/dL (ideally 80–100 mg/dL) or the lowest glucose which can be safely achieved without increasing hypoglycemia32 • Assess coefficient of variance to determine the stability of a patient’s glycemic profiles (goal ≤ 36%)25
Alert fatigue	• Patient with “mild” or lifestyle managed GDM, in particular, may have overall lower glucose values that are not pathologic, but may represent normal pregnancy physiology, especially overnight, and do not require treatment. • Additionally, patients who set their glucose “high” alerts at pregnancy target ranges such as 140 mg/dL may go into that range several times per day or week. • Having excessive low or high alert notifications can lead to alert fatigue where patients ignore reacting appropriately to the alerts.	• Alerts should be chosen based on shared-decision making between the patient and the provider based on what glucose levels are actionable for the patient. • Patients with diet-controlled diabetes can set low alerts at the lowest CGM value (i.e. below the low range 60mg/dL) given they are not taking medications that would cause hypoglycemia. • Patients can consider setting their high alert to be higher than pregnancy target range, particularly if their glycemic control is currently not well controlled, with adjustments in the high alert threshold as glycemic control improves. • Some CGMs have features to delay first alert for a period of time (e.g. 15 min to 4 hours) • Some CGMs have a silent or vibrate alerts

Figure 2.

Lag time of continuous glucose monitoring and trend arrows. CGM, continuous glucose monitoring.

Despite advancements in diabetes technology, SMBG still plays an important role in conjunction with CGM at this time (Table 2)^8,17. For example, if CGM readings and symptoms are discordant (i.e. CGM hypoglycemia alert but asymptomatic) or with hyperglycemia outside of pregnancy-specific targets that could impact medication administration, it is recommended to confirm blood glucose via SMBG¹⁷. CGM readings can be falsely low, often overnight, due to compression of the CGM sensor while sleeping. This low CGM reading in the absence of symptoms of hypoglycemia should prompt confirmation with SMBG before treating hypoglycemia. While CGM alerts can be helpful to prevent both high and low glucose excursions, these should be set at thresholds that are actionable and may need to be individualized based on diabetes type and current degree of glycemic control with gradual modification over time to prevent frustration with CGM technology or avoiding these important alerts altogether (Table 2).

Step-Wise Approach to Evaluating CGM Data

Despite the limited data demonstrating clinical superiority of CGM for management of diabetes in pregnant individuals with T2D and GDM, the FDA has approved multiple CGM systems for use in pregnancy (Freestyle Libre 2, Freestyle Libre 3, and Dexcom G7). As such, payors are increasingly covering CGM costs, despite clinical equipoise in some populations, and patients with T1D, T2D, and GDM increasingly requesting to use CGM for glucose monitoring in pregnancy. Thus, it is important for obstetric care providers to understand how to interpret and utilize the data provided from CGM systems. Overall, we recommend evaluating the data provided by CGM with a similar approach and similar frequency to SMBG data assessment. We propose using 4 basic steps, which are aligned with the approach proposed for CGM review in a non-pregnant population²⁴, but acknowledge that there may be other effective methods of CGM data evaluation and use.

Step 1: Are there enough data available?

It is important to first assess the completeness of the CGM data available. Adequacy of CGM data can be determined by looking at the Ambulatory Glucose Profile (AGP) Report (Figure 3) provided through the CGM systems’ online portals. Non-pregnant populations are recommended to have at least 10–14 days of data available²⁵. However, given the need for more frequent glycemic monitoring and medication adjustment in pregnancy, having at least 5–7 days of CGM data available provides summative decision-making capabilities. Additionally, the proportion of time when CGM was actively collecting and transmitting data should be at least 70%²⁵. Medical decision-making based on less data than this should be approached with caution, similar to making decisions when there are multiple days or timepoints of missing SMBG data. Additionally, we recommend investigating the reason for missing data when there is <5–7 days or <70% active CGM time. Reasons for incomplete CGM data may include failure to replace CGM sensor after previous sensor expired (~10–14 days) or fell off, sensor malfunction, or loss of Bluetooth connection between the transmitter and receiver or smartphone. A major benefit of CGM is that as long as the sensor is worn and data is transmitting via Bluetooth to a smartphone app or handheld receiver, ample data should exist without the requirement to perform, record, and bring in a SMBG log to all visits.

Figure 3.

Step-wise approach to evaluating the Ambulatory Glucose Profile report. CGM, continuous glucose monitoring; GMI, glucose management indicator.

Step 2: Is there a problem and what is it?

The next step is to evaluate the CGM data to determine if glucose values are meeting pregnancy-specific glucose targets or if there are recurring hyperglycemia or hypoglycemia episodes. Standardized CGM metrics found in the AGP report can help providers make formative assessments with their patients to identify areas that may need improvement (Figure 3). It is important to remember to change the CGM target range from factory target range settings for non-pregnant individuals (i.e. 70–180mg/dL) to reflect the recommended pregnancy-specific range of 63–140mg/dL in order to ensure appropriate calculation of CGM metrics²⁵. Currently, this can be done by accessing the settings from the mobile application or website for some CGM devices. Key CGM metrics include the proportion of time spent in target range (TIR), above range (TAR), and below range (TBR) as well as average glucose, coefficient of variation (a measure of glycemic variability), and glycemic management indicator (an estimation of hemoglobin A1c based on average glucose)²⁵.

For pregnant individuals with T1D, >70% TIR, <25% TAR, and <4% TBR are recommended based on the results of the CONCEPTT trial^13,25. There are no specific CGM metrics for pregnant patients with T2D or GDM due to insufficient data available, but it has been suggested that higher TIR may be possible given overall lower risk of hypoglycemia compared to pregnant individuals with T1D. Observational studies have consistently demonstrated associations between excursions from these recommended CGM targets and adverse pregnancy outcomes in individuals with all types of diabetes^26,27. The limited existing data from real-world studies support the goals of >70% TIR, <25% TAR and <4% TBR while noting that often these goals are not achieved before delivery among patients with T1D^13,28. It remains unclear whether the CGM targets should vary by diabetes type, trimester or other clinical factors. As such, we recommend reviewing the TIR, TAR, and TBR in the AGP report at each prenatal visit and assessing if patients are meeting the goals recommended for T1D in pregnancy. If there is evidence of suboptimal glycemic control (TIR≤70%), it should be determined if this is related to hyperglycemia (TAR≥25%) or hypoglycemia (TBR≥4%) or both. It is important to meet a patient at their current TIR and encourage stepwise improvements in TIR whenever possible as studies from pregnant patients with T1D and T2D have shown that every 5% increase in TIR is associated with improved perinatal outcomes^13,27,29.

Step 3: Where is the problem?

If there is evidence of dysglycemia due to hyperglycemia and/or hypoglycemia, the next step is to determine when over the course of the day these are occurring. The summary graph provided in the AGP report provides information about mean glucose, glucose variability, and patterns of when glucose goes out of range (Figure 3). This graph provides the median (50^th percentile) glucose level over the 24-hour day along with the 5^th, 25^th, 75^th, and 95^th percentiles. The shaded area between the 25^th and 75^th percentiles can be used to understand the range of glucose excursions 50% of the time. Evaluating if and when the 5^th percentile line crosses the lower CGM target of 63mg/dL can highlight areas of significant hypoglycemia. When the 75^th percentile line crosses the upper CGM target of 140mg/dL can highlight areas of significant hyperglycemia. Notably, fasting glucose levels above the SMBG target of 95mg/dL can be easily overlooked as these will fall between the target range of 63–140mg/dL and should require additional attention.

Evaluation of daily CGM profiles are helpful to identify consistent patterns of glycemic excursions and differentiate them from one-time outliers. In clinical practice, we find that reviewing these summary and/or daily graphs with patients allows an opportunity to identify behavioral patterns that may explain frequent excursions; for example, patients with hyperemesis gravidarum who report worse nausea in the morning followed by rebound carbohydrate intake commonly have CGM graphs that reflect this pattern. In contrast, patients with shift-work schedules may have nocturnal hyperglycemia that reflect altered sleeping and eating schedules. Patients fearful of hypoglycemia may wait to give mealtime insulin until after eating. Clarifying patient schedules and behaviors, and understanding which behaviors represent a “chronic state” as opposed to an occasional excursion for a non-recurrent reason, is an important element of reviewing problems.

Step 4: How do we fix the problem?

After identification of what the problem is (hyperglycemia, hypoglycemia, or both) and when the problem is occurring (overnight, fasting, postprandial, or universal), a plan should be made with the patient to fix the problem. Again, the approach to treatment of glucose excursions should be similar to the principles used to manage diabetes when using SMBG data, and thus changes in diet and lifestyle and medication adjustment based on hyper- or hypoglycemia are only briefly reviewed here. Hypoglycemia should be addressed first at it is a major maternal safety issue. Fasting hypoglycemia should prompt consideration of decreasing basal insulin or adding a balanced bedtime snack. Postprandial hypoglycemia should prompt evaluation of short-acting mealtime insulin timing, dose, and any pre- or postprandial corrections, as well as consideration of meal content. In real time, low alerts on CGM or concern for hypoglycemia based on symptoms should be confirmed with SMBG by the patient.

After addressing hypoglycemia, hyperglycemia should be addressed given its association with adverse perinatal outcomes. High fasting glucose on CGM or gradual increase in median glucose on CGM during the overnight period should be treated in the same way high fasting glucose on SMBG would be treated through initiation or increase in nighttime long-acting insulin. Abrupt increases in median glucose on CGM after eating that remains elevated in the postprandial hours should be treated in the same way high postprandial glucoses on SMBG would be with pre-prandial insulin initiation or adjustment. Compared to SMBG, CGM can better characterize a glucose response to meals. Large spikes in glucose that return to pre-prandial levels within 2–3 hours may indicate that mealtime insulin is not being given early enough before meals to mitigate the rapid glucose rise with meal consumption. This may be particularly evident in the third trimester of pregnancy when insulin absorption can be delayed by up to 45–60 minutes³⁰. Overall, the approach to treatment and aggressiveness with medication adjustments will vary based on a number of factors including type of diabetes, risk of hypoglycemia, response to prior treatment, gestational age, and patient preferences. Detailed discussion of insulin pump management including hybrid closed loop systems with automated insulin delivery is beyond the scope of this review, but may be found elsewhere³¹.

Conclusion

Despite clinical equipoise about the use of CGM for pregnant individuals with GDM or T2D as well as those without diabetes, the widespread popularity of these unobtrusive, convenient, and informative devices likely means they will soon become permanent fixtures of the obstetrician’s office. CGM data provides the opportunity to overcome many of the limitations of SMBG, improve health engagement and adherence with glucose monitoring, and introduce the concept of personalized medicine to obstetric care. As with all technology, as CGM systems advance, recommendations for use in pregnancy will also continue to evolve. To date, CGM has been shown to improve maternal and neonatal outcomes among pregnant individuals with T1D. Evidence for the role of CGM in pregnant people with T2D and GDM is currently lacking but clinical trials are underway or anticipated soon. Ultimately CGM has the potential to revolutionize antenatal care by offering the opportunity to use data-driven strategies to achieve euglycemia and improve pregnancy outcomes.

Pregnancy is a motivating time and obstetric providers have a significant advantage to impact transgenerational health. As more data are collected throughout pregnancy, CGM has the capability to provide insight to pregnancy physiology and metabolic health that are not currently appreciated with SMBG and point-of-care testing such as oral glucose tolerance tests and hemoglobin A1C. Using CGM technology to not only optimize diabetes management but also educate and empower patients on the effects of behavioral and lifestyle modifications such as healthy eating patterns and physical activity can result in lifelong health rewards. With the rapid clinical expansion of CGM use and the potential benefits of this strategy, it is increasingly important for obstetricians to understand CGM technology and metrics.