Continuous Glucose Monitoring in Pregnancies Complicated by Type 1 Diabetes Mellitus [20K

Abstract

INTRODUCTION: Continuous glucose monitors (CGM) is recommended for use in type 1 diabetes mellitus (T1DM) but studies in pregnancy are limited. We compare the number of hypoglycemic and hyperglycemic episodes on CGM to traditional self-monitoring blood glucose (SMBG) in women with T1DM. METHODS: Patients with T1DM presenting for care at UC San Diego <20 weeks were approached for participation. Patients underwent routine diabetes care and wore a blinded CGM (Dexcom G4) for 7 days every 4 weeks. Dexcom provided CGM. The primary outcome was the number of hours of hypoglycemia and hyperglycemia detected by each modality. Missed episodes of hypoglycemia/hyperglycemia were those not detected by one testing modality or the other. Mann Whitney-U, χ2 analysis, and Kruskal Wallis Test were performed where appropriate. RESULTS: Ten patients were recruited for this study. Two terminated their pregnancies. Two withdrew their participation. Thirty nine thousand five hundred seventy two glucose values (3,625 hours of glycemic data) were collected through CGM. Per day, the mean hours of reported hypoglycemia (4.3±1.9 for CGM versus 0.9±0.4 for SMBG, P<.01) and hyperglycemia (10.8±2.3 for CGM versus 2.7±0.9 for SMBG, P<.01) were significantly higher for the CGM group. For hypoglycemic and hyperglycemic hours, CGM demonstrated a lower mean proportion of missed hypoglycemia (3.7 versus 0.2 by CGM [P<.01]) and hyperglycemia (9.2 versus 0.1 by CGM [P<.01]). There was no correlation between A1c or trimester and missed hypoglycemic/hyperglycemic events. CONCLUSION: CGM detected a significant amount of time of unreported hypoglycemia/hyperglycemia over SMBG. Further study is needed to see if CGM can improve glycemic control and perinatal outcomes.