Metformin is comparable to insulin for pharmacotherapy in gestational diabetes mellitus: A network meta-analysis evaluating 6046 women.

Abstract

CONTEXT\nThe comparative efficacy of gestational diabetes (GDM) treatments lack conclusive evidence for choice of first-line treatment.\n\n\nOBJECTIVES\nThe aim of this study was to compare the efficacy of metformin and glibenclamide to insulin using a core outcome set (COS) to unify outcomes across trials investigating the treatment of gestational diabetes mellitus.\n\n\nSTUDY DESIGN\nA network meta-analysis (NMA) was conducted.\n\n\nDATA-SOURCE\nPubMed, Embase, and Cochrane Controlled Register of Trials were searched from inception to January 2020.\n\n\nSTUDY SELECTION\nRCTs that enrolled pregnant women who were diagnosed with GDM and that compared the efficacy of different pharmacological interventions for the treatment of GDM were included.\n\n\nMETA-ANALYSIS\nA generalized pairwise modelling framework was employed.\n\n\nRESULTS\nA total of 38 RCTs with 6046 participants were included in the network meta-analysis. Compared to insulin, the estimated effect of metformin indicated improvements for weight gain (WMD -2·39kg; 95% CI -3·31 to -1·46), maternal hypoglycemia (OR 0·34; 95% CI 0·12 to 0·97) and LGA (OR 0·61; 95% CI 0·38 to 0·98). There were also improvements in estimated effects for neonatal hypoglycemia (OR 0·48; 95% CI 0·19 to 1·25), pregnancy induced hypertension (OR 0·63; 95% CI 0·37 to 1·06), and preeclampsia (OR 0·74; 95% CI 0·538 to 1·04), though with limited evidence against our model hypothesis of equivalence with insulin for these outcomes.\n\n\nCONCLUSION\nMetformin is, at least, comparable to insulin for the treatment of GDM. Glibenclamide appears less favorable, in comparison to insulin, than metformin.