Diabetes during pregnancy is associated with significant risk of complications to the mother, fetus and newborn. We reviewed the potential impact of early detection and control of diabetes mellitus during pregnancy on stillbirths for possible inclusion in the Lives Saved Tool (LiST).
A systematic literature search up to July 2010 was done to identify all published randomized controlled trials and observational studies. A standardized data abstraction sheet was employed and data were abstracted by two independent authors. Meta-analyses were performed with different sub-group analyses. The analyses were graded according to the CHERG rules using the adapted GRADE criteria and recommendations made after assessing the overall quality of the studies included in the meta-analyses.
A total of 70 studies were selected for data extraction including fourteen intervention studies and fifty six observational studies. No randomized controlled trials were identified evaluating early detection of diabetes mellitus in pregnancy versus standard screening (glucose challenge test between 24th to 28th week of gestation) in pregnancy. Intensive management of gestational diabetes (including specialized dietary advice, increased monitoring and tailored dietary therapy) during pregnancy (3 studies: 3791 participants) versus conventional management (dietary advice and insulin as required) was associated with a non-significant reduction in the risk of stillbirths (RR 0.20; 95% CI: 0.03-1.10) (‘moderate’ quality evidence). Optimal control of serum blood glucose versus sub-optimal control was associated with a significant reduction in the risk of perinatal mortality (2 studies, 5286 participants: RR=0.40, 95% CI 0.25- 0.63), but not stillbirths (3 studies, 2469 participants: RR=0.51, 95% CI 0.14-1.88). Preconception care of diabetes (information about need for optimization of glycemic control before pregnancy, assessment of diabetes complications, review of dietary habits, intensification of capillary blood glucose self-monitoring and optimization of insulin therapy) versus none (3 studies: 910 participants) was associated with a reduction in perinatal mortality (RR=0.29, 95% CI 0.14 -0.60). Using the Delphi process for estimating effect size of optimal diabetes recognition and management yielded a median effect size of 10% reduction in stillbirths.
Diabetes, especially pre-gestational diabetes with its attendant vascular complications, is a significant risk factor for stillbirth and perinatal death. Our review highlights the fact that very few studies of adequate quality are available that can provide estimates of the effect of screening for aid management of diabetes in pregnancy on stillbirth risk. Using the Delphi process we recommend a conservative 10% reduction in the risk of stillbirths, as a point estimate for inclusion in the LiST.