Impact of Real Time, Flash and Retrospective Continuous Glucose Monitoring on Feto-Maternal Outcomes in Women With Gestational Diabetes and Pre-Gestational Diabetes: A Systematic Review and Meta-Analysis With Trial Sequential Analysis of Randomized Controlled Trials.

AIMS: Prior systematic reviews and meta-analyses (SRMs) evaluating continuous glucose monitoring (CGM) in pregnancy have yielded heterogeneous results due to pooling gestational diabetes mellitus (GDM) with pregestational diabetes, combining different CGM technologies (real-time, flash and retrospective) and mixing randomized controlled trials (RCTs) with observational designs. Several large RCTs have recently been published, necessitating an updated, design-restricted evaluation.

METHODS: Electronic databases were systematically searched for RCTs assessing CGM versus self-monitoring of blood glucose (SMBG) in pregnant women with GDM or pregestational diabetes. CGM modalities (real-time, flash and retrospective) and diabetes subtypes were analysed separately. Primary outcomes were neonatal intensive-care unit (NICU) admission and large-for-gestational-age (LGA) infants; secondary outcomes included other neonatal outcomes, maternal complications and CGM-derived glycaemic metrics.

RESULTS: Seventeen RCTs were included (13 GDM, 4 pregestational-diabetes). Real-time CGM significantly reduced NICU admissions in GDM compared to SMBG (OR 0.55, 95% CI 0.34-0.90; p = 0.02; I 2 = 0%). However trial sequential analysis showed that the current evidence has lack of power but not lack of effect with regards to reduction in NICU admission. Retrospective CGM in GDM was associated with lower risk of LGA (OR 0.46, 95% CI 0.22-0.98; p = 0.048; I 2 = 0%). Other feto-maternal outcomes like birthweight, macrosomia, Caesarean delivery, hypertensive disorders, neonatal hypoglycaemia and preterm-birth showed no significant difference between CGM and SMBG. No CGM modality conferred measurable benefit on feto-maternal outcomes in pregestational diabetes.

CONCLUSION: RT-CGM probably reduces NICU admission (moderate-certainty), but TSA indicates that current evidence is still information-size limited, so further large RCTs are warranted. Use of other CGM models in GDM and pregestational diabetes was largely neutral with regard to feto-maternal outcomes.