2 Step Gestational Diabetes Screen Prevails

Congratulations on the pregnancy; now here is a giant glass of sugar water to drink followed by a needle stick! Women in America have become accustomed to this necessary and yucky rite of passage in pregnancy: the gestational diabetes (GDM) glucose test. If you have a high number on the first test, you get the 3 hour oral glucose tolerance test- better bring a book!  Where did this glucose testing scheme come from anyway?

Dr. Landon of the Ohio State University College of Medicine provides an informative editorial in the January 2016 Green Journal providing a historical timeline of GDM screening recommendations. It all started in 1982 when Dr. O’Sullivan published the original GDM research. He diagnosed GDM based on a women’s likelihood of developing diabetes in the future, using a 100 gram glucose load test. Importantly, his diagnostic criteria were not based on perinatal or pregnancy-related outcomes.

Naturally, obstetricians wanted to know that their diagnoses of GDM were meaningful within the peripartum period. So, the 2008 Hyperglycemia and Adverse Pregnancy Outcome study, known as the HAPO study, sought to address this issue. They studied the international standard for screening for GDM- the one step 75 gram 2-hour oral glucose tolerance test. 23,316 non-diabetic women participated, and the authors found that increases in the three blood glucose levels taken for the test increased the likelihood of large for gestational age infants, primary c-section, fetal insulin levels, and neonatal adiposity. Unfortunately, the relationship was graded and there were no obvious cut-off points at which the risks of these outcomes jumped. So, in 2008, the results of the HAPO study in hand, the International Association of the Diabetes and Pregnancy Study Groups (IADPSG), came up with thresholds. In 2010, the IADPSG made it’s recommendations: everyone should have a one-step 2 hour 75 gram oral glucose tolerance test and if a woman has even one value that is off by an odds ratio of 1.75 times the mean, she gets the diagnosis of GDM. 

But wait- these new criteria drastically increased the cost of providing prenatal care. So many more women would be diagnosed with GDM!  If we diagnose all these extra women with GDM and treat them, will it improve their pregnancy outcomes?

In the United States, the Eunice Kennedy Shriver National Institute of Child Health and Human Development (the NICHD) convened a Consensus Development Conference to address these new concerns. No, they concluded in 2013, we should not all switch to the one-step test. We should continue to do the two-step, 50 gram glucose screen + the three hour oral glucose tolerance test, that is popular in the US. In 2013, the American College of Obstetricians and Gynecologsts, ACOG, agreed- we should keep doing the 2-step approach until we have evidence to show that the expensive, IADPSG approach actually improves perinatal outcomes.

This month the researchers at Kaiser Permanente answered the call. It would be an impossible task to randomize patients to two different screening groups, but the clever researchers, led by Drs. Feldman, Tieu and Yasumura, created a before-after cohort study design from 2010 to 2013. The before group simply used the current testing method- the two-step process using the Carpenter-Coustan thresholds for screening and diagnosing gestational diabetes. There were 2,972 patients in the before group and 17% were diagnosed with GDM by the two-step screening criteria. The after group were screened by the IADPSG one-step process. Of the 3,094 women in the after group, 27% were diagnosed with GDM. As everyone expected, the IADPSG criteria led to significantly more GDM diagnoses. Everyone diagnosed with GDM received intensive teaching, home glucose monitoring, and medication as medically indicated.

But now on to the interesting part- did diagnosing more women lead to better outcomes? The simple answer was no. Despite the IADPSG group receiving more GDM diagnoses, more glyburide treatment, and gaining about 1.5 pounds less during pregnancy, there was no difference in large for gestational age infants, macrosomic infants (babies greater than 4000 grams), NICU admissions, preterm delivery, preeclampsia, or high bilirubin in the infants. Only one major outcome was different: c-sections. The IADPSG group had significantly more c-sections, 20% versus 16%. They authors even adjusted for potential confounders like BMI.  The authors conclude that their results do not support using the IADPSG guidelines, which would lead to more women receiving a GDM diagnosis, more cost, more c-sections, and no benefit to the pregnancies.