GRAPEVINE, TEX. – Early postpartum glucose tolerance testing for women with gestational diabetes resulted in a 99% adherence rate, with similar sensitivity and specificity as the currently recommended 4- to 12-week postpartum testing schedule.
“Two-day postpartum glucose tolerance testing has similar diagnostic utility as the 4- to 12-week postpartum glucose tolerance test to identify impaired glucose metabolism and diabetes at 1 year postpartum,” said Erika Werner, MD, speaking at the meeting sponsored by the Society for Maternal-Fetal Medicine.
Overall, 29% of women studied had impaired glucose metabolism at 2 days postpartum, as did 25% in the 4- to 12-weeks postpartum window. At 1 year, that figure was 35%. The number of women meeting diagnostic criteria for diabetes held steady at 4% for all three time points.
The findings warrant “consideration for the 2-day postpartum glucose tolerance test (GTT) as the initial postpartum test for women who have gestational diabetes, with repeat testing at 1 year,” said Dr. Werner, a maternal-fetal medicine physician at Brown University, Providence, R.I.
Glucose testing for women with gestational diabetes mellitus (GDM) is recommended at 4-12 weeks postpartum by both the American Diabetes Association and the American College of Obstetricians and Gynecologists.
Testing can allow detection and treatment of impaired glucose metabolism, seen in 15%-40% of women with a history of GDM. Up to 1 in 20 women with GDM will receive a postpartum diagnosis of type 2 diabetes.
However, fewer than one in five women will actually have postpartum glucose testing, representing a large missed opportunity, said Dr. Werner.
Several factors likely contribute to those screening failures, she added. In addition to the potential for public insurance to lapse at 6 weeks postpartum, the logistical realities and time demands of parenting a newborn are themselves a significant barrier.
“What if we changed the timing?” and shifted glucose testing to the early postpartum days, before hospital discharge, asked Dr. Werner. Several pilot studies had already compared glucose screening in the first few days postpartum with the routine schedule, finding good correlation between the early and routine GTT schedule.
Importantly, the earlier studies achieved an adherence rate of more than 90% for early GTT. By contrast, fewer than half of the participants in the usual-care arms actually returned for postpartum GTT in the 4- to 12-week postpartum window, even under the optimized conditions associated with a medical study.
The single-center prospective cohort study conducted by Dr. Werner and collaborators enrolled 300 women with GDM. Women agreed to participate in glucose tolerance testing as inpatients, at 2 days postpartum, in addition to receiving a GTT between 4 and 12 weeks postpartum, and additional screening that included a glycosylated hemoglobin (HbA1c) test at 1 year postpartum.
The investigators obtained postpartum day 2 GTTs for all but four of the patients. A total of 201 patients returned in the 4- to 12-week postpartum window, and 168 of those participants returned for HbA1c testing at 1 year. Of the 95 patients who didn’t come back for the 4- to 12-week test, 33 did return at 1 year for HbA1c testing.
Dr. Werner and her coinvestigators included adult women who spoke either fluent Spanish or English and had GDM diagnosed by the Carpenter-Coustan criteria, or by having a blood glucose level of 200 mg/dL or more in a 1-hour glucose challenge test.
The early GTT results weren’t shared with patients or their health care providers. For outpatient visits, participants were offered financial incentives and received multiple reminder phone calls and the offer of free transportation.
For the purposes of the study, impaired glucose metabolism was defined as fasting blood glucose of 100 mg/dL or greater, a 2-hour GTT blood glucose level of 140 mg/dL or greater, or HbA1c of 5.7% or greater.
Participants were diagnosed with diabetes if they had a fasting blood glucose of 126 mg/dL or greater, a 2-hour GTT blood glucose level of 200 mg/dL or greater, or HbA1c of 6.5% or greater.
Dr. Werner and colleagues conducted two analyses of their results. In the first, they included only women in both arms who had complete data. In the second analysis, they looked at all women who had data for the 1-year postpartum mark, assuming that interval GTTs were negative for women who were missing these values.
The statistical analysis showed that, for women with complete data, both early and later postpartum GTTs were similar in predicting impaired glucose metabolism at 1 year postpartum (areas under the receiver operating curve [AUC], 0.63 and 0.60, respectively).
For identifying diabetes at 1 year, both early and late testing had high negative predictive value (98% and 99%, respectively), but the later testing strategy had higher sensitivity and specificity, yielding an AUC of 0.83, compared with 0.65 for early testing.
Turning to the second analysis that included all women who had 1-year postpartum HbA1c values, negative predictive values for diabetes were similarly high (98%) for both the early and late testing strategies. For identifying impaired glucose metabolism at 1 year in this group, both the positive and negative predictive value of the early and late strategies were similar.
Patients were about 32 years old at baseline, with a mean body mass index of 31.7 kg/m2. More than half of patients (52.3%) had private insurance, and 22% had GDM in a pregnancy prior to the index pregnancy. Black patients made up about 9% of the study population; 54% of participants were white, and 23% Hispanic. About one-third of patients were nulliparous, and two-thirds had education beyond high school.
During their pregnancies, about 44% of patients managed GDM by diet alone, 40% required insulin, with an additional 1% also requiring an oral agent. The remainder required oral agents alone. Patients delivered at a mean 38.3 weeks gestation, with about 40% receiving cesarean deliveries.
Some of the study’s strengths included its prospective nature, the diverse population recruited, and the fact that participants and providers were both blinded to the 2-day GTT results. Although more than half of participants completed the study – besting the previous pilots – 44% of patients still had incomplete data, noted Dr. Werner.
The American Diabetes Association sponsored the study. Dr. Werner reported no other conflicts of interest.
SOURCE: Werner E et al. SMFM 2020. Abstract 72.