Design: In an observational cohort of pregnant women with pregestational type 2 diabetes on metformin monotherapy before the last menstrual period (LMP), we emulated a target trial with two treatment strategies: 1) insulin monotherapy: discontinue metformin and initiate insulin within 90 days after LMP, or 2) insulin plus metformin: continue metformin and initiate insulin within 90 days after LMP.

Setting: US Medicaid health care administration database (2000-2018).

Fig1.sas: flow chart and baseline characteristics of the study population

Fig2.sas: Comparison of estimated risk ratio of congenital malformations among live births using different analytic approaches,

Table3_tt1.sas: Estimated risk of non-live births and non-chromosomal major congenital malformations under insulin monotherapy or insulin plus metformin therapy during the first trimester among women with type 2 diabetes (target trial with stricter eligibility)

Table3_tt2.sas: Estimated risk of non-live births and non-chromosomal major congenital malformations under insulin monotherapy or insulin plus metformin therapy during the first trimester among women with type 2 diabetes (target trial with broader eligibility)