1- or 2-step: Outcomes studied in GDM screening

The prevalence of prediabetes and diabetes together in women of gestational age is about 24 percent in the U.S., he notes. “So 20 percent is not a high number for GDM diagnosis if you think about it in those terms.” However, ACOG allows many different versions of the two-step protocol, which is another problem, in his view.

It’s difficult to predict the impact of the Kaiser study, and there may not be enough attention to the study’s limitations, Dr. Sacks says. “New England Journal papers are always very, very influential and change practices often. Even papers that have flaws in them, once they are published they might run a correction but nobody will see it. I expect there will be opinion pieces in different journals addressing the study’s limitations, but they won’t get nearly as much publicity as a New England Journal paper.”

One important aspect of the Kaiser study, in Dr. Sacks’ view, is that by focusing on patient outcomes it improves on the original criteria for diagnosing GDM, which were from the 1960s and used two standard deviations above the mean glucose tolerance as the cutoff for women likely to develop diabetes post-pregnancy. The 2008 HAPO trial on which the IADPSG guidelines are based was a perinatal outcomes-based study.

Coincidentally, the David Sacks who is a coauthor of a “Perspectives in Care” piece published in April on “Resolving the Gestational Diabetes Diagnosis Conundrum” is a specialist in neonatology and not the same person as the NIH investigator (Bilous RW, et al. Diabetes Care. 2021;44[4]:858–864). But both physicians agree on the need, expressed in that Diabetes Care editorial, for an international, multicenter, randomized controlled trial of treatment to answer the GDM screening question.

Research like the Kaiser study is centrally important to perinatologist Amy Valent, DO, director of the Diabetes and Pregnancy Program and assistant professor of obstetrics and gynecology, Oregon Health and Science University. “I specialize in endocrinopathies that occur in pregnancy, and certainly diabetes is one of the largest endocrinopathies that we see in pregnancy. Knowing who we are helping and who we’re not helping is a topic near and dear to our hearts.”

In an AACC-sponsored debate in 2019, moderated by Dr. Sacks, Dr. Valent was designated to make the case for the two-step approach to screening. But she could just as easily have advocated for the opposite point of view, because one-step is what her institution uses in practice. She is able to see both sides of the argument.

“This is always an interesting topic, if only for the fact that we still haven’t agreed on which one we should do here in the U.S. The rest of the world has adopted the one-step approach. The U.S. is really the only remaining country that still has controversies over this.” The Kaiser study and all other countries in Europe, she says, have demonstrated with their data that the one-step approach results in considerably more GDM diagnoses.

The Kaiser study is unique in looking at so many people, Dr. Valent says. “I think that’s the strongest part of the study. The population was so large and you can look at the question in a more generalizable fashion.” Like Dr. Sacks, she stresses that measuring glucose is a very sensitive thing. “How you handle the specimen, how you order the test, the time in processing, and all of that—there are a lot of variables that go into the actual value you get from that sample of blood.” Moreover, “there is interpersonal variability. Studies have shown that even if the same person were to do the same test a week apart, it could look very different.”

The cutoffs chosen for the one-step and two-step approaches are also different, she points out. “The two-step approach was to identify pregnant women who are eventually going to develop type 2 diabetes because they were at highest risk of having adverse pregnancy outcomes and maybe we can intervene sooner. Whereas the one-step approach in the HAPO study looked at actual fetal outcomes and pregnancy outcomes like C-peptide, neonatal hypoglycemia, and what are going to be the cutoffs associated with those outcomes.”

“If you’re going to start looking at outcomes, then you have to ask the question: What’s going to influence an outcome? Well, that’s glycemic control. So with insulin versus oral agents, how well the patients’ blood sugar is controlled. And then we can get to a whole rabbit hole of where our pregnancy targets should be.” The Kaiser study, in her view, wouldn’t be needed if it were “just to answer the question of which approach is going to diagnose GDM more.” It’s needed to answer the question: “If we identify more people using the one-step approach, then how can we improve outcomes? Then you have to start looking at what actually influences outcomes.”

Adherence problems did affect the Kaiser study, Dr. Valent says. “A third of the participants in the one-step approach converted to the two-step. So it’s difficult to make strong conclusions. At a person’s first prenatal visit, they got randomized right off the bat and providers were not blind to those randomization strategies. And there are providers who would say, ‘Well, I don’t think this patient is going to come back and fast for me, so I’m going to switch her over to the two-step approach so I can give her a sugar test right now.’ That was a huge limitation for this study because there was so much provider bias there. But that’s real life.”

Confirming the value of getting a screening test done is that the six percent of patients who avoided all testing have worse outcomes, she says. “It just goes to show that regardless of what you decide to do as a provider based on the evidence, you should do some sort of testing, whether it’s one-step or two-step, because the people who didn’t get any testing had worse outcomes.”

At the end of the day, Dr. Valent says, “I just talk to my referring providers and say, ‘However you decide to screen your patients, just at least screen them. That’s number one. And number two, you have to do what’s best for your practice. If you cannot sustain having twice the volume of GDM in your practice, then maybe the two-step approach is better for you right now.”

Although no pregnant mothers want to hear they have GDM, “if we can help them learn about healthy eating patterns and higher-quality food, that is a wonderful benefit. I oftentimes tell my patients with GDM that I think this is a blessing in disguise because you can change your life around and improve your metabolic health going into the future.”

The non-blindness of the Kaiser study, because the providers have to know which test their patient is receiving, has created controversy. A randomized controlled trial underway now, Dr. Valent says, does have the providers blinded as to which arm their patient is randomized. “So it will be interesting to see if they see similar results or not.”

There isn’t yet enough evidence to say which of the two approaches is better, she says. “It clearly doesn’t change outcomes to do one or the other. So should we be managing people differently? Should we be diagnosing earlier, since the fetal pancreas starts secreting insulin at about 11 to 12 weeks? By the time you’re diagnosed at 24 to 28 weeks, a lot of time has passed, especially if there’s been silent hyperglycemia. Is it a timing issue and not necessarily a test issue? How we screen is a little less important than how we should manage.”

“When you’re trying to compare a test to an outcome,” Dr. Valent says, “that makes it very challenging because you have a lot of variables that can’t be accounted for in this type of study,” such as quality of nutrition and medication management strategies. “You have to understand the limitations. Just because something doesn’t show a difference doesn’t mean it’s not important. You have to understand that maybe there are questions that weren’t able to be answered based on the study design.”

That said, she notes, there is legitimate concern about under-screening. Data released this year revealed that at one hospital in New York, only 12 of 97 women who met ACOG’s criteria for early GDM screening received it. “The ADA and ACOG both recommend screening early in pregnancy for those individuals at higher risk for having type 2 diabetes, and we aren’t doing that very well.

“What this study shows without controversy is that not screening is potentially harmful,” Dr. Valent says. “If you don’t screen, you have potentially worse outcomes. So screening is meaningful. Whether you do the one-step or the two-step, just screen. And let’s move forward and address bigger questions.”

For Dr. Hillier, the value of the Kaiser study is clear: It addresses a significant gap in the research and should advance screening and treatment of gestational diabetes. “This topic has been extremely controversial and very hotly debated because people care about how to best treat their patients and we’ve been lacking evidence. There are well-intended people on both sides trying to do the right thing for their patients,” she says.

“I feel very satisfied that we did a very, very diligent job. And during the trial we had to go longer than we expected because of the non-adherence that we didn’t predict. But I feel that my job as a researcher is to evaluate and study the two different approaches, implement that and analyze it, and then leave it to other expert organizations to decide what they want to do with the evidence.”

She hopes that the study’s findings will inform the clinical guidelines of the professional organizations as well as possibly the U.S. Preventive Services Task Force. “We’ll have to see how they interpret it.” But what the study shows is helpful, in her view. “I think anybody involved in diagnostic testing for women who may have gestational diabetes can be reassured that with either one-step or two-step, there are no differences in outcomes for women based on the screening approach.”

As for the question of when the one-step versus two-step debate will be decided once and for all, Dr. Hillier isn’t willing to guess. “I wish I had a crystal ball. But it’s like trying to predict when a pandemic will be over.”

Anne Paxton is a writer and attorney in Seattle.