Counterpoint: Enough Evidence to Treat? The American College of Obstetricians and Gynecologists Guidelines

Abstract

In the US, the most common practice for gestational diabetes mellitus (GDM)2 screening is a 2-step process: an initial 1-h (50 g) glucose tolerance test (GTT) done at 24–28 weeks, followed by a confirmatory 3-h (100 g) GTT (fasting, 1-h, 2-h, and 3-h measurements) (1). On the assumption of a 1% rate of pre-GDM and the fact that 10% of pregnant women will be of “low risk” and thus not require screening, approximately 3.6 × 106 women undergo GDM screening per year in the US (2). A diagnosis of GDM is then made if ≥2 values on the 3-h GTT are above reference limits [Carpenter and Coustan criteria: fasting, >95 mg/dL (>5.27 mmol/L); 1 h, >180 mg/dL (>9.99 mmol/L); 2 h, >155 mg/dL (>8.60 mmol/L); 3 h, >140 mg/dL (>7.77 mmol/L)]. With the current American College of Obstetricians and Gynecologists (ACOG) diagnostic criteria, approximately 7% of pregnant women are diagnosed with GDM (>250 000 women per year). Although there was a time when many experts questioned whether GDM was a “disease” that warranted treatment, data from 2 randomized clinical trials (RCTs) that demonstrated both maternal and neonatal benefit from treatment for GDM have largely quieted this argument (3, 4). Although the National Institute of Child Health and Human Development (NICHD) Maternal Fetal Medicine Units Network (MFMU) mild GDM treatment trial did not report differences between groups in the primary composite outcome (stillbirth or neonatal death, hypoglycemia, hyperbilirubinemia, increased cord blood C-peptide, and birth trauma) or between any of the individual outcomes in the composite, there were statistically significant differences in several secondary outcomes, such as macrosomia [number needed to treat (NNT) = 12], …