Evaluating the clinical effectiveness of repeat screening for gestational diabetes mellitus following a diagnosis of large for gestational age and/or polyhydramnios

Abstract Objective Gestational diabetes mellitus (GDM) screening is often repeated in the third trimester when large for gestational age (LGA) and/or polyhydramnios are diagnosed. We evaluated outcomes for individuals who did versus did not rescreen, and also for individuals who passed or did not pass rescreening. Methods A retrospective cohort study from January 2011 to March 2024 of term, singleton pregnancies diagnosed with LGA and/or polyhydramnios after a normal GDM screen at 24–30 weeks was performed. Non‐inferiority analyses were performed for the comparison between those who did not rescreen and those who did rescreen (passed or did not pass) using a mixed model. In non‐inferiority analyses, if the lower bound of a one‐sided 95% confidence interval (CI) is within the non‐inferiority limit (margin of error), the test is significant, allowing non‐inferiority to be declared. The lower bound of the 95% CI and corresponding p values were reported. Outcomes in individuals who passed and who did not pass rescreening were compared using a logistic regression model. Odds ratios with respective 95% CI and p values are reported. Results Of the 1665 pregnant individuals who met the inclusion criteria, 1317 did not rescreen and 348 rescreened. In individuals who were rescreened, 13.8% were diagnosed with GDM, of which 20.8% required medication. Outcomes were not worse for those who did not rescreen than for those who did rescreen, including birthweight (BW) (3758 vs. 3813 g, −123.65, p < 0.01), BW ≥ 4000 (31.9% vs. 34.1%, −0.32, p < 0.01), cesarean delivery (21.0% vs. 28.0%, 3.15, p < 0.01), or shoulder dystocia (5.2% vs. 6.6%, −1.47, p < 0.01). Outcomes for individuals who did rescreen did not differ whether they passed or failed, aside from an increase in neonatal hypoglycemia (7.0% vs. 22.9%, 3.95 [95% CI, 1.76–8.84], p < 0.01). Conclusion Pregnant individuals who did not rescreen for GDM following the diagnosis of LGA and/or polyhydramnios in the third trimester do not appear to have worse birth outcomes than those who did rescreen. Additionally, individuals diagnosed with GDM after rescreening were not at an increased risk for adverse perinatal outcomes. This study demonstrates that rescreening all individuals with LGA and/or polyhydramnios may not be warranted.