1973-LB: Postpartum Dysglycemia Screening with Continuous Glucose Monitoring

Introduction & Objective: Up to one third of birthing persons with gestational diabetes (GDM) will have persistent dysglycemia, and more than half do not complete the recommended postpartum oral glucose tolerance test (OGTT). This study assessed the feasibility of postpartum continuous glucose monitoring (CGM) to detect dysglycemia by assessing sensor return rates and participant experience and calculated the predictive power of CGM to predict OGTT results.

Methods: CGM was placed on postpartum day 1-3 after pregnancy complicated by GDM and worn at home for up to 10 days. Participants were encouraged to undergo standard of care 6-week OGTT. CGM data was analyzed according to time-above cutoffs. Analysis with removal of data within 4 half-lives of steroid administration was also performed.

Results: Fifty women (36±6 years; 40% non-Hispanic White, 24% non-Hispanic Black, 22% Asian, 14% Hispanic; 34% Medicaid insured) were consented; 86% returned CGM and 66% completed OGTT. All CGM downloads included at least 72 hours of data, and 19% provided data after recent steroid administration. Fasting data was available on 88% of CGM and 42% of OGTT. Mean CGM glucose was 122±14 mg/dL. Median time >140, >180, and >200 mg/dL was 17%, 1.9%, and 0.6%, respectively; and 91%, 21%, and 33% had time >140 mg/dL of >5%, >180 mg/dL of >5%, and >200 mg/dL of >1%. Mean fasting CGM glucose was 113±15 mg/dL and was 100-125 mg/dL in 51% and >125 mg/dL in 23%. Dysglycemia on OGTT (7 impaired glucose tolerance, 1 diabetes) was well predicted by percent time >180 of >4% (sensitivity 86%, specificity 85%, positive predictive value 60%, negative predictive value 96%). Removal of steroid affected data did not improve performance. CGM and OGTT were considered burdensome by 7% and 52%, respectively.

Conclusion: Completion rates and acceptability scores were better for CGM than OGTT. Percent time >180 mg/dL had strong predictive power for OGTT. Time above range for this population was also found to be higher than reported for people without diabetes.