Permissive intrapartum glucose control: an equivalence randomized control trial (PERMIT)

Background: There is limited high-quality data on the best practices for maternal blood glucose management during labor. Objective: We compared permissive care (target maternal blood glucose 70–180 mg/dL) to usual care (blood glucose 70–110 mg/dL) among laboring individuals with diabetes. Study Design: This was a two-site equivalence randomized control trial for individuals with diabetes (pregestational or gestational) at ≥34 weeks in labor. Individuals were randomly allocated to usual care or permissive care. Maternal blood glucose was evaluated by capillary blood glucose monitoring in latent and active labor every 4 and 2 hours. Insulin drip was initiated if maternal blood glucose exceeded the upper bounds of the allocated target. The primary outcome was the first neonatal heel stick glucose within 2 hours of birth before feeding. We assumed a mean first neonatal blood glucose of 50±10 mg/dL. To ensure that the use of permissive care did not increase or decrease the first neonatal blood glucose >10 mg/dL (2-tailed: a=0.05, b=0.1), 96 total participants were required. We calculated adjusted relative risk and 95% confidence intervals in an intention-to-treat analysis. A preplanned Bayesian analysis was used to estimate the probability of equivalence with a neutral informative prior. Results: Of deliveries with diabetes assessed for eligibility (from October 2022 to June 2023), 280 of 511 (54.8%) met eligibility criteria, and 96 of 280 (34.3%) agreed and were randomized. In the usual care group, 17% required an insulin drip compared with none in permissive care. There was equivalence in the primary outcome between usual and permissive care (57.9 vs 57.1 mg/dL; adjusted mean difference, −0.72 [95% confidence interval, −8.87 to 7.43]). Bayesian analysis indicated a 98% posterior probability of the mean difference not being >10 mg/dL. The rate of neonatal hypoglycemia was 25% in the usual care group and 29% in the permissive group (adjusted relative risk, 1.14; 95% confidence interval, 0.60–2.17). There was no difference in other neonatal or maternal outcomes. Conclusion: In this randomized control trial, although almost 1 in 6 individuals with diabetes required an insulin drip with usual intrapartum maternal blood glucose care, permissive care was associated with equivalent neonatal blood glucose.