Advertisement

Two-day postpartum compared with 4- to 12-week postpartum glucose tolerance testing for women with gestational diabetes

      Background

      A majority of women in the United States with gestational diabetes mellitus do not undergo the recommended 4- to 12-week postpartum glucose tolerance test.

      Objective

      This study aimed to compare the diagnostic value of the 2-day postpartum glucose tolerance test with the 4- to 12-week postpartum glucose tolerance test to identify impaired glucose metabolism at 1 year after delivery among women with gestational diabetes.

      Study Design

      Postpartum women who delivered at 1 hospital between January 2017 and July 2018 were offered enrollment in a prospective cohort if they had gestational diabetes mellitus diagnosed by Carpenter-Coustan criteria or a 1-hour glucose challenge test result of ≥200 mg/dL, spoke English or Spanish, and planned to remain in the hospital for at least 2 days after delivery. Participating women underwent a 75-gram 2-hour glucose tolerance test on postpartum day 2 and were incentivized to have a 4- to 12-week glucose tolerance test and measurement of glycosylated hemoglobin at 1 year after delivery. Participants and providers were blinded to the 2-day postpartum results. The diagnostic value of an abnormal 2-day postpartum glucose tolerance test (fasting result of ≥100 mg/dL or 2-hour glucose tolerance test result of ≥140 mg/dL) was compared with that of an abnormal 4- to 12-week glucose tolerance test to identify impaired glucose metabolism (≥5.7% glycosylated hemoglobin) and diabetes (≥6.5% glycosylated hemoglobin) at 1 year after delivery. Receiver operating characteristic (ROC) curves were also compared at 2 days and 4–12 weeks after delivery.

      Results

      Of the 300 recruited women, 296 (99%) completed the 2-day postpartum glucose tolerance test, and 202 (68%) returned for the 4- to 12-week glucose tolerance test. Approximately 1 year after delivery, 203 (68%) women had their glycosylated hemoglobin measured, of whom 35% had impaired glucose metabolism and 4% had diabetes. The study population was diverse (46% nonwhite). Furthermore, 56% were obese (mean body mass index, 32 kg/m2), and 55% had received medication to control their glucose during pregnancy. There were no significant differences between the 2-day and 4- to 12-week postpartum glucose tolerance tests in predicting impaired glucose metabolism based on ≥5.7% glycosylated hemoglobin in 1 year after delivery: sensitivity (46% vs 36%); specificity (79% vs 84%); positive predictive value (52% vs 53%); and negative predictive value (75% vs 72%). There was also no difference between the 2-day and the 4- to 12-week glucose tolerance tests in identifying diabetes at 1 year after delivery. Both the 2-day and 4- to 12-week glucose tolerance tests had similar ROC curves in identifying impaired glucose metabolism and diabetes at 1 year after delivery.

      Conclusion

      Two-day postpartum glucose tolerance tests have similar diagnostic value as 4- to 12-week postpartum glucose tolerance tests in predicting impaired glucose metabolism and diabetes at 1 year after delivery and are associated with nearly 100% adherence to the test. Thus, changing the timing of the glucose tolerance test should be considered.

      Key words

      Introduction

      In the United States, more than 6% of pregnancies are complicated by gestational diabetes mellitus (GDM).
      Committee on Practice Bulletins—Obstetrics
      ACOG Practice Bulletin No. 190: gestational diabetes mellitus.
      Without intervention, 240,000 (70%) women diagnosed with GDM annually will develop diabetes mellitus (DM) over their lifetime, resulting in serious health consequences and escalating healthcare costs.
      Committee on Practice Bulletins—Obstetrics
      ACOG Practice Bulletin No. 190: gestational diabetes mellitus.
      However, there may be a simple, cost-effective change to postpartum care that could result in up to a 40% decrease in DM incidence in women previously diagnosed with GDM.
      • Ratner R.E.
      • Christophi C.A.
      • Metzger B.E.
      • et al.
      Prevention of diabetes in women with a history of gestational diabetes: effects of metformin and lifestyle interventions.
      ,
      • Aroda V.R.
      • Christophi C.A.
      • Edelstein S.L.
      • et al.
      The effect of lifestyle intervention and metformin on preventing or delaying diabetes among women with and without gestational diabetes: the Diabetes Prevention Program outcomes study 10-year follow-up.
      The Diabetes Prevention Program randomized 3234 participants with impaired glucose tolerance, including 350 women with a history of GDM, to metformin, lifestyle modification, or placebo. Among the women with a history of GDM, both lifestyle modification and metformin compared with placebo resulted in DM risk reduction at 3 and 10 years after randomization.
      • Ratner R.E.
      • Christophi C.A.
      • Metzger B.E.
      • et al.
      Prevention of diabetes in women with a history of gestational diabetes: effects of metformin and lifestyle interventions.
      ,
      • Aroda V.R.
      • Christophi C.A.
      • Edelstein S.L.
      • et al.
      The effect of lifestyle intervention and metformin on preventing or delaying diabetes among women with and without gestational diabetes: the Diabetes Prevention Program outcomes study 10-year follow-up.
      This opportunity for intervention is part of the reason that both the American Diabetes Association (ADA) and the American College of Obstetricians and Gynecologists (ACOG) recommend postpartum glucose testing at 4–12 weeks after delivery for all women with GDM.
      Committee on Obstetric Practice
      ACOG Committee Opinion No. 435: postpartum screening for abnormal glucose tolerance in women who had gestational diabetes mellitus.
      ,
      American Diabetes Association
      14. Management of diabetes in pregnancy: standards of medical care in diabetes—2019.
      However, in the current system, a minority of women receive this recommended testing.
      • Hale N.L.
      • Probst J.C.
      • Liu J.
      • Martin A.B.
      • Bennett K.J.
      • Glover S.
      Postpartum screening for diabetes among Medicaid-eligible South Carolina women with gestational diabetes.
      • Bennett W.L.
      • Chang H.Y.
      • Levine D.M.
      • et al.
      Utilization of primary and obstetric care after medically complicated pregnancies: an analysis of medical claims data.
      • Werner E.F.
      • Has P.
      • Kanno L.
      • Sullivan A.
      • Clark M.A.
      Barriers to postpartum glucose testing in women with gestational diabetes mellitus.

      Why was this study conducted?

      A minority of women with gestational diabetes return at the recommended 4–12 weeks postpartum, and therefore we miss a huge opportunity to identify women with impaired glucose metabolism or diabetes before complications occur.

      Key findings

      Glucose tolerance testing at 2 days post-partum is not only feasible but has similar diagnostic value as glucose tolerance testing at 4–12 weeks postpartum when the goal is to identify impaired glucose metabolism or diabetes at 1 year postpartum among women with gestational diabetes.

      What does this add to what is known?

      This is first study to examine the diagnostic value of the 2-day postpartum glucose tolerance test to identify impaired glucose metabolism and diabetes at 1 year postpartum.
      Qualitative studies suggest that there are numerous reasons for the routine failure to perform glucose testing after delivery on women with GDM.
      • Werner E.F.
      • Has P.
      • Kanno L.
      • Sullivan A.
      • Clark M.A.
      Barriers to postpartum glucose testing in women with gestational diabetes mellitus.
      • Bennett W.L.
      • Ennen C.S.
      • Carrese J.A.
      • et al.
      Barriers to and facilitators of postpartum follow-up care in women with recent gestational diabetes mellitus: a qualitative study.
      • Van Ryswyk E.
      • Middleton P.
      • Hague W.
      • Crowther C.
      Clinician views and knowledge regarding healthcare provision in the postpartum period for women with recent gestational diabetes: a systematic review of qualitative/survey studies.
      • Smirnakis K.V.
      • Chasan-Taber L.
      • Wolf M.
      • Markenson G.
      • Ecker J.L.
      • Thadhani R.
      Postpartum diabetes screening in women with a history of gestational diabetes.
      In an attempt to address barriers such as transportation, insurance, and childcare, our group and others have investigated shifting the timing of postpartum glucose testing to 2 days after delivery when women are still hospitalized.
      • Werner E.F.
      • Has P.
      • Tarabulsi G.
      • Lee J.
      • Satin A.
      Early postpartum glucose testing in women with gestational diabetes mellitus.
      • Dinglas C.
      • Muscat J.
      • Heo H.
      • Islam S.
      • Vintzileos A.
      Immediate postpartum glucose tolerance testing in women with gestational diabetes: a pilot study.
      • Carter E.B.
      • Martin S.
      • Temming L.A.
      • Colditz G.A.
      • Macones G.A.
      • Tuuli M.G.
      Early versus 6–12 week postpartum glucose tolerance testing for women with gestational diabetes.
      These pilot studies have revealed that 2-day postpartum glucose testing for women with a history of GDM not only is feasible but also may have a similar sensitivity in diagnosing DM as 4- to 12-week postpartum testing. However, the diagnostic value of postpartum day 2 glucose testing to identify impaired glucose metabolism and diabetes over a longer time course (at least 1 year after delivery) is needed before changing practice. Therefore, we sought to prospectively compare the diagnostic value of glucose tolerance testing during postpartum hospitalization with glucose tolerance testing at 4–12 weeks after delivery with a goal of identifying impaired glucose metabolism and DM at 1 year after delivery among women with a history of GDM.

      Materials and Methods

      We performed a prospective cohort study at the Women and Infants Hospital in Providence, Rhode Island, with institutional review board approval (IRB number 836907). Women were recruited on the postpartum units between January 2017 and July 2018. Women were eligible for enrollment if they were diagnosed with GDM during their pregnancy either by a 1-hour glucose challenge test value of ≥200 mg/dL or by the Carpenter-Coustan criteria using the 3-hour 100-g glucose tolerance test (GTT), were ≥18 years of age, and were fluent in English or Spanish. We excluded women intending to move outside of the state of Rhode Island during the subsequent year and women who could not tolerate a 75-gram GTT (eg, women with a history of gastric bypass).
      We obtained written informed consent for study participation after delivery, typically on postpartum day 1. Following consent, each participant completed a demographic survey, and study staff ordered a 2-hour 75-gram GTT for the morning of postpartum day 2. The test was dispensed by pharmacy staff, and both a fasting serum glucose and 2-hour serum glucose were drawn by a hospital phlebotomist. Study staff, participants, and clinicians were blinded to the results of the postpartum day 2 GTT unless the glucose value exceeded 250 mg/dL in which case the patient’s obstetrician was notified and study participation terminated because this value was deemed to represent overt diabetes. Postpartum care was not otherwise perturbed by study participation.
      Each participant was followed up by study staff at approximately 3–5 weeks after delivery to arrange the standard of care at 4–12 weeks GTT. This was ordered at the hospital laboratory, the same laboratory that processed the inpatient postpartum day 2 GTT. Results of the 4–12 week GTT were shared with the participant and their obstetrical clinician.
      At 3 months after delivery, all participants were asked to complete a survey by phone, and outpatient postpartum electronic medical records (EMRs) were queried. At 10 months after delivery, each study participant was followed up by study staff to arrange a 1-year postpartum study visit. At that visit, a glycosylated hemoglobin (HgbA1c) was measured.
      To maximize compliance with the study visits, the research team scheduled all study visits on the most convenient day for the participant including weekend days. Participants received $25 at the time of each GTT (2 days after delivery and 4–12 weeks after delivery) and $50 at the 1-year postpartum visit.
      Data on covariates were collected on all enrolled participants. Women completed surveys at enrollment in which they self-identified their race and ethnicity, education history, and insurance status. We also collected data from the EMR regarding maternal age, prepregnancy body mass index (BMI), weight and height at the first and last prenatal visit, GDM management, and delivery information. Around 1 year after delivery, we collected data on their postpartum care, including nutrition referrals and medications that were initiated.
      The primary goal of this study was to compare the diagnostic value of the 2-hour GTT on postpartum day 2 with the standard of care GTT at 4–12 weeks after delivery to forecast impaired glucose metabolism at 1 year after delivery. Impaired glucose metabolism at 1 year after delivery was defined as ≥5.7% HgbA1c. A secondary goal of the study was to assess the diagnostic value of the 2-hour GTT on postpartum day 2 compared with 4- to 12-week postpartum GTT with regard to predicting DM at 1 year after delivery, defined as ≥6.5% HgbA1c.
      Because the study goals required data on HgbA1c at 1 year after delivery, participants who did not return in 1 year (±4 months) after delivery to have HgbA1c drawn were excluded. Characteristics of participants who completed the study and were included were compared with the characteristics of participants who enrolled but were lost to follow-up before 1 year after delivery to assess for bias in the sample. For the primary analysis, we limited the cohort to only women who completed glucose evaluation at all 3 time points (2 days after delivery, 4–12 weeks after delivery, and 1 year after delivery). All women with fasting GTT values of ≥100 mg/dL or 2-hour GTT values of ≥140 mg/dL at 2 days after delivery or 4–12 weeks after delivery were considered to have impaired glucose metabolism at that time point. We also planned a secondary analysis in which women with missing data at 2 days or 4–12 weeks after delivery were included as long as they had HgbA1c drawn at 1 year after delivery. In this analysis, a participant with a fasting GTT value of ≥100 mg/dL or 2-hour GTT value of ≥140 mg/dL at 2 days after delivery or 4–12 weeks after delivery was considered to have impaired glucose metabolism at that time point, but all other women, including those who did not complete the GTT at either 2 days or 4–12 weeks after delivery, were categorized as having normal glucose metabolism. We opted to include women who did not complete the 2 day or 4–12 postpartum GTT and women who had negative results at those time points in an attempt to mimic real-world clinical practice in which women who do not get tested remain undiagnosed.
      Using dichotomous categories (abnormal GTT at 2 days after delivery or no known abnormality; abnormal GTT at 4–12 weeks after delivery or no known abnormality; and abnormal HgbA1c or normal HgbA1c), sensitivity, specificity, and predictive value of the 2-hour GTT on postpartum day 2 was compared with the 4- to 12-week postpartum GTT to predict impaired glucose metabolism in 1 year after delivery. The postpartum day 2 GTT was then compared with the 4- to 12-week postpartum GTT to predict DM at 1 year after delivery in a similar fashion. Finally, the area under the ROC curves for the 2-day postpartum GTT and the 4- to 12-week postpartum GTT were compared with regard to ≥5.7% HgbA1c and ≥6.5% HgbA1c.
      With a planned enrollment of 300, assuming as many as 50% of women would be lost to follow-up, we had 88% power to detect a significant difference in ROC curves of 0.80 vs 0.60 at the alpha=0.05 level for a 2-tailed test, when the correlation between the 2 day and 4- to 12-week postpartum GTTs is zero (unlikely) or 96% power if the correlation was 0.3 (at the low range of plausible relationships) or as high as 99% if the correlation is 0.6 (a plausible upper range value in such analyses).

      Results

      Of the 300 women who agreed and were consented to participate in the study, 296 (99%) completed a GTT on postpartum day 2. Three women who initially consented to the study refused the test on postpartum day 2, and 1 woman had a fasting glucose drawn and received the 75-gram glucose drink, but a laboratory error resulted in her 2-hour serum glucose not being obtained. Of the 300 women recruited for the study, 202 (67%) returned at 4–12 weeks for a GTT, and 203 (68%) returned in 1 year after delivery for HgbA1c (Figure 1).
      Figure thumbnail gr1
      Figure 1Testing adherence among study participants by time points
      GTT, glucose tolerance testing.
      Werner. Two-day postpartum compared to 4–12 week postpartum glucose tolerance testing for women with gestational diabetes. Am J Obstet Gynecol 2020.
      The cohort was diverse with 23% of women self-identifying as Hispanic and 9% self-identifying as African American (Table 1). Most enrolled women were also overweight or obese with a mean BMI of 31.7 kg/m2. Women who did not return at 4–12 weeks after delivery compared with those who did were younger, heavier, less educated, and less likely to have insurance. Women who did not return in 1 year after delivery differed only from those who did with regard to insurance; women who did not return were less likely to be privately insured (41.2% vs 57.6%; P=.009).
      Table 1Maternal characteristics of study cohort
      VariableAll participants (N=300)Complete data (n=168)HgbA1c in 1 year after delivery (n=203)
      Age in years (mean [SD])31.7 (5.3)32.3 (4.9)31.9 (5.1)
      Race and ethnicity (%)
       White54.048.848.4
       Black or African American9.38.39.9
       Hispanic or Latino23.026.826.6
       Asian7.08.36.9
       Multiracial, other, or unknown6.77.77.9
      Nulliparous (%)34.034.531.5
      Education beyond high school (%)68.375.070.9
      Private insurance (n, %)52.360.757.6
      Body mass index (kg/m2) (mean [SD])31.7 (7.9)30.7 (7.3)31.1 (7.5)
      GDM in prior pregnancy (%)22.021.423.2
      Weight gain with IOM guidelines (%)28.928.828.1
      GDM management (%)
       Diet only43.743.542.9
       Insulin40.037.539.9
       Glyburide15.317.916.3
       Insulin plus oral agent1.01.21.0
      Gestational weeks at delivery (mean [SD])38.3 (1.9)38.638.3 (1.7)
      Cesarean delivery (%)39.735.737.4
      GDM, gestational diabetes mellitus; HgbA1c, glycosylated hemoglobin; IOM, Institute of Medicine; SD, standard deviation.
      Werner. Two-day postpartum compared to 4–12 week postpartum glucose tolerance testing for women with gestational diabetes. Am J Obstet Gynecol 2020.
      At 2 days after delivery, 29% of women tested had impaired glucose metabolism (fasting glucose value of ≥100 mg/dL or 2-hour glucose value of ≥140 mg/dL), and 4% had results suggestive of diabetes (fasting glucose value of ≥126 mg/dL or 2-hour glucose value of ≥200 mg/dL). At 4–12 weeks after delivery, 25% of the women who returned for testing had impaired GTT results, and 4% had findings suggestive of diabetes. None of these women were started on any treatments for DM. Finally, at 1 year after delivery, 35% of women tested had abnormal HgbA1c (≥5.7%), and 4% of women had HgbA1c consistent with DM (≥6.5%).
      Among the 168 women with complete glucose data, there was no difference between the 2-day postpartum GTT and the 4- to 12-week postpartum GTT with regard to test characteristics in identifying impaired glucose metabolism or DM (Table 2). When ROC curves were used to compare diagnostic value, there was no difference in the 2-day postpartum GTT area under the curve (AUC) and the 4- to 12-week postpartum GTT AUC with regard to predicting impaired glucose metabolism or diabetes in 1 year after delivery (Figure 2). When the cohort was expanded to all women who returned in 1 year after delivery, assuming normal GTT results for all missing 2-day postpartum or 4- to 12-week postpartum tests. The diagnostic values of the tests also did not differ (Table 3 and Figure 3).
      Table 2Diagnostic values of the GTT among participants with complete data 1 year after delivery
      Abnormal HgbA1c (≥5.7%)Diabetes (≥6.5% HgbA1c)
      2-day postpartum testing, % (95% CI) (N=168)4- to 12-week postpartum testing, % (95% CI) (N=168)2-day postpartum testing, % (95% CI) (N=168)4- to 12-week postpartum testing, % (95% CI) (N=168)
      Sensitivity46.4 (33.0–60.3)35.7 (23.4–49.6)57.1 (18.4–90.1)85.7 (42.1–99.6)
      Specificity78.6 (69.8–85.8)83.9 (75.8–90.2)71.4 (63.8–78.3)80.1 (73.1–86.0)
      Positive predictive value52.0 (37.4–66.3)52.6 (35.8–69.0)8.0 (2.2–19.2)15.8 (6.0–31.3)
      Negative predictive value74.6 (65.7–82.1)72.3 (63.8–79.8)97.5 (92.7–99.5)99.2 (95.8–100)
      CI, confidence interval; GTT, glucose tolerance test; HgbA1c, glycosylated hemoglobin.
      Werner. Two-day postpartum compared to 4–12 week postpartum glucose tolerance testing for women with gestational diabetes. Am J Obstet Gynecol 2020.
      Figure thumbnail gr2
      Figure 2Graphs depicting 2-day AUC and 4- to 12-week AUC among women with complete testing data
      A, Comparing 2-day postpartum glucose tolerance testing with 4- to 12-week postpartum glucose tolerance testing to identify impaired glucose metabolism in 1 year after delivery among women with complete testing data. B, Comparing 2-day postpartum glucose tolerance testing with 4- to 12-week postpartum glucose tolerance testing to identify diabetes in 1 year after delivery among women with complete testing data.
      AUC, area under the curve; CI, confidence interval.
      Werner. Two-day postpartum compared to 4–12 week postpartum glucose tolerance testing for women with gestational diabetes. Am J Obstet Gynecol 2020.
      Table 3Diagnostic values of the GTT among participants who had HgbA1cs obtained 1 year after delivery
      Abnormal HgbA1c (≥5.7%)Diabetes (≥6.5% HgbA1c)
      2-day postpartum testing, % (95% CI) (N=203)4- to 12-week postpartum testing, % (95% CI) (N=203)2-day postpartum testing, % (95% CI) (N=203)4- to 12-week postpartum testing, % (95% CI) (N=203)
      Sensitivity40.8 (29.3–53.2)28.2 (18.1–40.1)66.7 (29.9–92.5)66.7 (29.9–92.5)
      Specificity78.0 (70.0–84.8)87.1 (80.2–92.3)73.2 (66.4–79.3)84.0 (78.1–88.9)
      Positive predictive value50.0 (36.6–63.4)54.1 (36.9–70.5)10.3 (3.9–21.2)16.2 (6.2–32.0)
      Negative predictive value71.0 (62.9–78.3)69.3 (61.7–76.2)97.9 (94.1–99.6)98.2 (94.8–99.6)
      CI, confidence interval; GTT, glucose tolerance test; HgbA1c, glycosylated hemoglobin.
      Werner. Two-day postpartum compared to 4–12 week postpartum glucose tolerance testing for women with gestational diabetes. Am J Obstet Gynecol 2020.
      Figure thumbnail gr3
      Figure 3Graphs depicting 2-day AUC and 4- to 12-week AUC among women with 1 year postpartum data
      A, Comparing 2-day postpartum glucose tolerance testing with 4- to 12-week postpartum glucose tolerance testing to identify impaired glucose metabolism in 1 year after delivery among women with 1 year postpartum data. B, Comparing 2-day postpartum glucose tolerance testing with 4- to 12-week postpartum glucose tolerance testing to identify diabetes in 1 year after delivery among women with 1 year postpartum data.
      AUC, area under the curve; CI, confidence interval.
      Werner. Two-day postpartum compared to 4–12 week postpartum glucose tolerance testing for women with gestational diabetes. Am J Obstet Gynecol 2020.

      Discussion

      In routine practice, a minority of women with GDM returned in 4–12 weeks after delivery for a GTT despite ACOG and ADA recommendations.
      • Hale N.L.
      • Probst J.C.
      • Liu J.
      • Martin A.B.
      • Bennett K.J.
      • Glover S.
      Postpartum screening for diabetes among Medicaid-eligible South Carolina women with gestational diabetes.
      • Bennett W.L.
      • Chang H.Y.
      • Levine D.M.
      • et al.
      Utilization of primary and obstetric care after medically complicated pregnancies: an analysis of medical claims data.
      • Werner E.F.
      • Has P.
      • Kanno L.
      • Sullivan A.
      • Clark M.A.
      Barriers to postpartum glucose testing in women with gestational diabetes mellitus.
      Even in this study in which women were incentivized financially to return, received free transportation, and numerous reminder phone calls, one-third of women did not complete glucose testing at 4–12 weeks after delivery compared with <1% of women not completing testing at 2 days after delivery. Researchers have investigated means of improving compliance with the recommended 4- to 12-week GTT including hiring nurses dedicated to scheduling the test, incentivizing providers to refer women for the test, incentivizing patients to complete the test, and changing the type of test from a GTT to a point of care test. At best, a few of these protocols have achieved rates similar to this study.
      • Mendez-Figueroa H.
      • Daley J.
      • Breault P.
      • et al.
      Impact of an intensive follow-up program on the postpartum glucose tolerance testing rate.
      In actual practice, most states report between a 3% and 35% compliance with postpartum glucose testing.
      • Hale N.L.
      • Probst J.C.
      • Liu J.
      • Martin A.B.
      • Bennett K.J.
      • Glover S.
      Postpartum screening for diabetes among Medicaid-eligible South Carolina women with gestational diabetes.
      • Bennett W.L.
      • Chang H.Y.
      • Levine D.M.
      • et al.
      Utilization of primary and obstetric care after medically complicated pregnancies: an analysis of medical claims data.
      • Werner E.F.
      • Has P.
      • Kanno L.
      • Sullivan A.
      • Clark M.A.
      Barriers to postpartum glucose testing in women with gestational diabetes mellitus.
      In contrast, compliance with GTTs at 2 days after delivery have exceeded 95% in all studies in the United States thus far.
      • Werner E.F.
      • Has P.
      • Tarabulsi G.
      • Lee J.
      • Satin A.
      Early postpartum glucose testing in women with gestational diabetes mellitus.
      • Dinglas C.
      • Muscat J.
      • Heo H.
      • Islam S.
      • Vintzileos A.
      Immediate postpartum glucose tolerance testing in women with gestational diabetes: a pilot study.
      • Carter E.B.
      • Martin S.
      • Temming L.A.
      • Colditz G.A.
      • Macones G.A.
      • Tuuli M.G.
      Early versus 6–12 week postpartum glucose tolerance testing for women with gestational diabetes.
      ,
      • Waters T.P.
      • Kim S.Y.
      • Werner E.
      • et al.
      Should women with gestational diabetes be screened at delivery hospitalization for type 2 diabetes?.
      Furthermore, in this study, we demonstrated that the 2 day test does not differ substantially from the 4-12 week test with regard to sensitivity, specificity, or negative or positive predictive value in identifying women who will develop impaired glucose metabolism and diabetes at 1 year after delivery. Thus, if we are going to recommend postpartum glucose testing in women with GDM, it is prudent to consider routinely offering the test at 2 days after delivery during hospitalization.
      This study and its results have limitations. This was a single-center study. Other sites have found similar 2-day postpartum GTT compliance rates, but we are the only study to compare the 2-day GTT and the 4- to 12-week postpartum GTT to 1-year results. Thus, although our population is diverse, any single-center study should be confirmed at other centers across the study. Our study is also limited by the length of follow-up. In an ideal study with no financial constraints, we would have followed these women for their lifetimes to determine which GTT best predicts diabetes over the life course. In addition, despite financial incentivizes, recurrent reminders, and attempt to make follow-up as convenient as possible, we still lost one-third of the enrolled participants in 1 year after delivery. Although this points to the importance of capitalizing on the postpartum hospitalization for testing and intervention, it does raise concern for selection bias. We compared the retained population with the lost to follow-up population to address this and found a difference only with regard to insurance status. Other predictors of diabetes risk, including BMI, age, and history of recurrent GDM, did not differ between groups. Furthermore, Hispanics and African Americans were more likely to return in 1 year than white women, suggesting that the study design did not limit the diversity of the study population analyzed. Finally, our study does not address qualitatively whether participants preferred the 2-day postpartum GTT. Although we have performed qualitative studies previously identifying barriers to the 4- to 12-week postpartum test,
      • Werner E.F.
      • Has P.
      • Kanno L.
      • Sullivan A.
      • Clark M.A.
      Barriers to postpartum glucose testing in women with gestational diabetes mellitus.
      we did not repeat those surveys in this study.
      Despite the limitations of this study, we believe it addresses a vital gap in knowledge, which is how to pragmatically identify women at higher risk for impaired glucose metabolism and DM after a pregnancy complicated by GDM. We have indicated that reassuring results of a GTT on either 2 days or 4–12 weeks after delivery are associated with a high likelihood of normal glucose testing results in 1 year after delivery. This suggests that the in-hospital 2-day GTT is an appropriate initial test to determine who needs to be retested in the first year after delivery and who can wait until several years after delivery to be retested. This will allow for more personalized postpartum counseling so that those at higher risk can receive evidenced-based, time-intensive interventions to reduce their DM risk.
      We have previously indicated that GDM screening in pregnancy is cost-effective at the population level only if GDM diagnosis leads to DM prevention over the life course.
      • Werner E.F.
      • Pettker C.M.
      • Zuckerwise L.
      • et al.
      Screening for gestational diabetes mellitus: are the criteria proposed by the International Association of Diabetes and Pregnancy Study Groups cost-effective?.
      As currently managed, GDM leads to little postpartum follow-up. The 2-day postpartum GTT is the first step in changing postpartum care so that a GDM diagnosis leads to postpartum risk mitigation. The 2-day postpartum GTT can be achieved in more than 90% of women with GDM, and its high negative predictive value will allow for DM risk–reducing interventions to be targeted at the minority of women who have abnormal GTT values at 2 days after delivery. These women will leave the hospital knowing that they are at high risk for progression to impaired glucose metabolism and DM and need testing 1 year after delivery. They also can be made aware of the possible interventions (metformin and lifestyle modification) that can reduce their DM risk. Even modest reductions in DM risk for these women will result in universal GDM screening being cost-effective at the population level.

      References

        • Committee on Practice Bulletins—Obstetrics
        ACOG Practice Bulletin No. 190: gestational diabetes mellitus.
        Obstet Gynecol. 2018; 131: e49-e64
        • Ratner R.E.
        • Christophi C.A.
        • Metzger B.E.
        • et al.
        Prevention of diabetes in women with a history of gestational diabetes: effects of metformin and lifestyle interventions.
        J Clin Endocrinol Metab. 2008; 93: 4774-4779
        • Aroda V.R.
        • Christophi C.A.
        • Edelstein S.L.
        • et al.
        The effect of lifestyle intervention and metformin on preventing or delaying diabetes among women with and without gestational diabetes: the Diabetes Prevention Program outcomes study 10-year follow-up.
        J Clin Endocrinol Metab. 2015; 100: 1646-1653
        • Committee on Obstetric Practice
        ACOG Committee Opinion No. 435: postpartum screening for abnormal glucose tolerance in women who had gestational diabetes mellitus.
        Obstet Gynecol. 2009; 113: 1419-1421
        • American Diabetes Association
        14. Management of diabetes in pregnancy: standards of medical care in diabetes—2019.
        Diabetes Care. 2019; 42: S165-S172
        • Hale N.L.
        • Probst J.C.
        • Liu J.
        • Martin A.B.
        • Bennett K.J.
        • Glover S.
        Postpartum screening for diabetes among Medicaid-eligible South Carolina women with gestational diabetes.
        Womens Health Issues. 2012; 22: e163-e169
        • Bennett W.L.
        • Chang H.Y.
        • Levine D.M.
        • et al.
        Utilization of primary and obstetric care after medically complicated pregnancies: an analysis of medical claims data.
        J Gen Intern Med. 2014; 29: 636-645
        • Werner E.F.
        • Has P.
        • Kanno L.
        • Sullivan A.
        • Clark M.A.
        Barriers to postpartum glucose testing in women with gestational diabetes mellitus.
        Am J Perinatol. 2019; 36: 212-218
        • Bennett W.L.
        • Ennen C.S.
        • Carrese J.A.
        • et al.
        Barriers to and facilitators of postpartum follow-up care in women with recent gestational diabetes mellitus: a qualitative study.
        J Womens Health (Larchmt). 2011; 20: 239-245
        • Van Ryswyk E.
        • Middleton P.
        • Hague W.
        • Crowther C.
        Clinician views and knowledge regarding healthcare provision in the postpartum period for women with recent gestational diabetes: a systematic review of qualitative/survey studies.
        Diabetes Res Clin Pract. 2014; 106: 401-411
        • Smirnakis K.V.
        • Chasan-Taber L.
        • Wolf M.
        • Markenson G.
        • Ecker J.L.
        • Thadhani R.
        Postpartum diabetes screening in women with a history of gestational diabetes.
        Obstet Gynecol. 2005; 106: 1297-1303
        • Werner E.F.
        • Has P.
        • Tarabulsi G.
        • Lee J.
        • Satin A.
        Early postpartum glucose testing in women with gestational diabetes mellitus.
        Am J Perinatol. 2016; 33: 966-971
        • Dinglas C.
        • Muscat J.
        • Heo H.
        • Islam S.
        • Vintzileos A.
        Immediate postpartum glucose tolerance testing in women with gestational diabetes: a pilot study.
        Am J Perinatol. 2017; 34: 1264-1270
        • Carter E.B.
        • Martin S.
        • Temming L.A.
        • Colditz G.A.
        • Macones G.A.
        • Tuuli M.G.
        Early versus 6–12 week postpartum glucose tolerance testing for women with gestational diabetes.
        J Perinatol. 2018; 38: 118-121
        • Mendez-Figueroa H.
        • Daley J.
        • Breault P.
        • et al.
        Impact of an intensive follow-up program on the postpartum glucose tolerance testing rate.
        Arch Gynecol Obstet. 2014; 289: 1177-1183
        • Waters T.P.
        • Kim S.Y.
        • Werner E.
        • et al.
        Should women with gestational diabetes be screened at delivery hospitalization for type 2 diabetes?.
        Am J Obstet Gynecol. 2020; 222: 73.e1-73.e11
        • Werner E.F.
        • Pettker C.M.
        • Zuckerwise L.
        • et al.
        Screening for gestational diabetes mellitus: are the criteria proposed by the International Association of Diabetes and Pregnancy Study Groups cost-effective?.
        Diabetes Care. 2012; 35: 529-535