A Pragmatic, Randomized Clinical Trial of Gestational Diabetes Screening

Abstract

Background: Gestational diabetes mellitus is common and is associated with an increased risk of adverse maternal and perinatal outcomes. Although experts recommend universal screening for gestational diabetes, consensus is lacking about which of two recommended screening approaches should be used.

Methods: We performed a pragmatic, randomized trial comparing one-step screening (i.e., a glucose-tolerance test in which the blood glucose level was obtained after the oral administration of a 75-g glucose load in the fasting state) with two-step screening (a glucose challenge test in which the blood glucose level was obtained after the oral administration of a 50-g glucose load in the nonfasting state, followed, if positive, by an oral glucose-tolerance test with a 100-g glucose load in the fasting state) in all pregnant women who received care in two health systems. Guidelines for the treatment of gestational diabetes were consistent with the two screening approaches. The primary outcomes were a diagnosis of gestational diabetes, large-for-gestational-age infants, a perinatal composite outcome (stillbirth, neonatal death, shoulder dystocia, bone fracture, or any arm or hand nerve palsy related to birth injury), gestational hypertension or preeclampsia, and primary cesarean section.

Results: A total of 23,792 women underwent randomization; women with more than one pregnancy during the trial could have been assigned to more than one type of screening. A total of 66% of the women in the one-step group and 92% of those in the two-step group adhered to the assigned screening. Gestational diabetes was diagnosed in 16.5% of the women assigned to the one-step approach and in 8.5% of those assigned to the two-step approach (unadjusted relative risk, 1.94; 97.5% confidence interval [CI], 1.79 to 2.11). In intention-to-treat analyses, the respective incidences of the other primary outcomes were as follows: large-for-gestational-age infants, 8.9% and 9.2% (relative risk, 0.95; 97.5% CI, 0.87 to 1.05); perinatal composite outcome, 3.1% and 3.0% (relative risk, 1.04; 97.5% CI, 0.88 to 1.23); gestational hypertension or preeclampsia, 13.6% and 13.5% (relative risk, 1.00; 97.5% CI, 0.93 to 1.08); and primary cesarean section, 24.0% and 24.6% (relative risk, 0.98; 97.5% CI, 0.93 to 1.02). The results were materially unchanged in intention-to-treat analyses with inverse probability weighting to account for differential adherence to the screening approaches.

Conclusions: Despite more diagnoses of gestational diabetes with the one-step approach than with the two-step approach, there were no significant between-group differences in the risks of the primary outcomes relating to perinatal and maternal complications. (Funded by the Eunice Kennedy Shriver National Institute of Child Health and Human Development; ScreenR2GDM ClinicalTrials.gov number, NCT02266758.).

Copyright © 2021 Massachusetts Medical Society.

Figures

Figure 1:. Consort Diagram of Randomization of GDM Screening Methods and Analytical Comparison Groups

☆All pregnancies were randomized to 1-step or 2-step GDM screening strategies within the EMR as part of clinical care at their first prenatal visit. The 1 step (75g 2 hour OGTT) approach diagnosed GDM based on IADPSG criteria, and the 2-step screening approach by Carpenter and Coustan criteria. †Percentages do not add up to 100% as some pregnancies met multiple exclusion criteria. The major reason for exclusion was miscarriage (as randomization occurred at the 1st prenatal visit, in many cases this visit also determined non-viability, or miscarriage, on the same day of randomization and before any GDM screening was ordered); terminations were also included in this exclusion category. Change of insurance during pregnancy was an exclusion as we were unable to evaluate outcomes in these pregnancies. ‡Intention-to-treat analyses (ITT) were planned. Due to unanticipated lower adherence to the fasting 1-step at both sites, we continued randomizing until we enrolled enough pregnancies with 1-step screening to have adequate statistical power, and conducted additional analyses - inverse probability (IP) weighted ITT - both with and without adjustment for factors related to non-adherence.,, Factors related to lower adherence included both maternal and provider characteristics as well as provider reliance on non-fasting tests to ensure that GDM screening was completed at a visit. These pragmatic barriers to adherence could not be adequately addressed without putting patients at risk of not receiving GDM screening. §Among the 1,503 pregnancies that did not receive either 1-step or 2-step screening, 1,450 (6.1%) were unscreened (778 [6.5%] randomized to 1-step and 672 (5.7%) randomized to 2-step), and these pregnancies presented on average at a mean of 18.9 weeks’ gestation compared to 10.5 weeks for pregnancies with screening. There were also 53 pregnancies that had other clinically recommended screening in the first trimester (HbA1c or FPG) but did not have either randomized GDM screening approach.