Broadening the diagnosis of gestational diabetes mellitus (GDM) with a one-step screening approach does not lead to significant differences in maternal or perinatal outcomes, compared with a two-step approach. Investigators reported these findings in the New England Journal of Medicine after testing the two screening methods in more than 23,000 pregnant women.
GDM affects 6%-25% of pregnant women, increasing the risk of neonatal death and stillborn births. It can also lead to serious complications such as fetal overgrowth. Clinical guidelines recommend GDM screening between 24 and 28 weeks’ gestation to improve outcomes in mothers and infants. However, the scientific community has struggled to reach a consensus on testing approach.
For decades, clinicians used a two-step screening approach: a nonfasting 1-hour glucose challenge test and a longer 3-hour fasting oral glucose tolerance test to diagnose GDM; roughly 20% who test positive on this glucose challenge test require the second step. Results of a large study led to new diagnostic criteria on a one-step 75-g oral glucose tolerance test. The Hyperglycemia and Adverse Pregnancy Outcome (HAPO) study “found a linear relationship with hyperglycemia and outcomes – the higher the glucose, the worse the outcomes,” said Teresa Hillier, MD, MS, an endocrinologist and investigator with Kaiser Permanente Center for Health Research Northwest and CHR-Hawaii. The International Association of the Diabetes and Pregnancy Study Groups (IADPSG) made a clinical recommendation on the one-step approach, now a common screening tool in the United States.
A focus on rare GDM outcomes
The IADPSG fasting one-step criteria typically identifies women with milder symptoms as having gestational diabetes, a factor expected to increase diagnosis rates by two- or threefold, said Dr. Hillier. “The unknown question was whether diagnosing and treating more women would be associated with any differences in any of the multiple GDM-associated outcomes for mother and baby.”
She and her colleagues conducted a large-scale randomized trial at two Kaiser sites to assess multiple maternal and perinatal outcomes including rare but important GDM-associated outcomes such as stillbirth and neonatal death between the two screening methods.
They randomized 23,792 pregnant women 1:1 to the one- or two-step gestational diabetes test at their first prenatal visit. Primary outcomes included diagnosis of gestational diabetes; large-for-gestational-age infants; primary cesarean section, and gestational hypertension or preeclampsia; and a composite perinatal outcome of any stillbirth, neonatal death, shoulder dystocia, bone fracture, or arm or hand nerve palsy related to birth injury.
Most participants (94%) completed screening, although there was lower adherence to screening in the one-step approach. The reasons for this aren’t entirely clear, said Dr. Hillier. Convenience may be a factor; patients have to fast for several hours to complete the one-step test, whereas the first test of the two-step screening approach can be done at any time of day, and most patients pass this test.
Corroborating HAPO’s results, twice as many women in the one-step group (16.5%) received a GDM diagnosis, compared with 8.5% in the two-step group (unadjusted relative risk, 1.94; 97.5% confidence interval, 1.79-2.11). However, for the other primary outcomes, investigators found no significant differences in incidences or unadjusted risks. Perinatal composite outcomes for the one- and two-step groups were 3.1% and 3.0%, respectively, and primary cesarean section outcomes were 24.0% and 24.6%.
In the one-step group, 8.9% experienced large-for-gestational-age infants outcomes, compared with 9.2% in the two-step group (RR, 0.95; 97.5% CI, 0.87-1.05). Among those diagnosed with gestational diabetes, similar percentages of women in the one- and two-step groups received insulin or hypoglycemic medication (42.6% and 45.6%, respectively).
Dr. Hillier and colleagues also reported comparable results among the two groups on safety outcomes and secondary outcomes such as macrosomia incidence, small-for-gestational-age infants, and factors such as neonatal hypoglycemia and respiratory distress.
“Although we did not find increased harms associated with the diagnosis and treatment of gestational diabetes in many more women with the one-step approach, some retrospective observational cohort studies have shown higher incidences of primary cesarean delivery and neonatal hypoglycemia with one-step screening after conversion from two-step protocols, with no substantive improvement in outcomes,” Dr. Hillier and colleagues noted.
The trial had several limitations. Adjustments made to address lower adherence to the one-step approach might not have accounted for all nonadherence differences. Another issue is the two sites didn’t use identical thresholds for the glucose challenge test in the two-step cohort. Demographically, the study lacked Black and American Indian representation.
“Moreover, the potential long-term benefits of increased diagnoses of gestational diabetes – such as the identification of more women at high risk for subsequent diabetes who might benefit from risk-reduction strategies – were not addressed by the trial,” Brian Casey, MD, wrote in a related editorial. Based on the study’s findings, “the perinatal benefits of the diagnosis of gestational diabetes with the use of the IADPSG single-step approach appear to be insufficient to justify the associated patient and health care costs of broadening the diagnosis” of GDM, added Dr. Casey, a professor with the department of obstetrics and gynecology at the University of Alabama at Birmingham.