An intravenous insulin protocol designed for pregnancy reduces neonatal hypoglycaemia following betamethasone administration in women with gestational diabetes

Abstract

Marked hyperglycaemia is common following betamethasone administration in women with gestational diabetes (GDM), and may contribute to neonatal hypoglycaemia. Validated protocols to deliver glycaemic stability following betamethasone are lacking. We hypothesized that an intravenous insulin (IVI) protocol for pregnancy-specific glycaemic targets (Pregnancy-IVI) would achieve greater at-target glycaemic control than a generic adult intravenous insulin protocol (Adult-IVI), and may reduce neonatal hypoglycaemia. A retrospective cohort study of the performance Adult-IVI and Pregnancy-IVI following betamethasone in GDM, sequentially implemented at a tertiary hospital, without change in indication for IVI. Cases were identified by electronic record search. Primary outcome was percentage of on-IVI time with at-target glycaemia [blood glucose level (BGL) 3.8-7mmol/l]. Secondary outcomes were time with critical hyperglycaemia (BGL >10mmol/l), occurrence of maternal hypoglycaemia (BGL <3.8mmol/l), and incidence of neonatal hypoglycaemia (BGL ≤2.5mmol/l) if betamethasone was administered within 48h of birth. The cohorts comprised 151 women (Adult-IVI n=86; Pregnancy-IVI n=65). The primary outcome was 68% time-at-target [95% confidence interval (CI) 64-71%) for Pregnancy-IVI compared with 55% (95% CI 50-60%) for Adult-IVI (P=0.0002). Critical maternal hyperglycaemia (0% vs. 2%, P=0.02) and hypoglycaemia (2% vs. 12%, P=0.02) were both lower with Pregnancy-IVI than Adult-IVI. Neonatal hypoglycaemia was less common after Pregnancy-IVI (29%) than after Adult-IVI (54%, P=0.03). A multiple logistic regression model adjusting for potential confounders gave an odds ratio for neonatal hypoglycaemia with Pregnancy-IVI of 0.27 (95% CI 0.10-0.76, P=0.01). An IVI protocol designed for pregnancy effectively controlled maternal hyperglycaemia following betamethasone administration in GDM. This is the first intervention to show a reduction in betamethasone-associated neonatal hypoglycaemia, linked with optimum maternal glycaemic control.