Today, I am blogging the results from the second day of the NIH Consensus Development Conference on Diagnosing Gestational Diabetes. The purpose of this conference is to come to a consensus on two things: the best way to diagnose gestational diabetes (GDM), and whether we should change the criteria we use to diagnose GDM in the U.S. To read the summary from the first day of the conference, click here.
As a side note, the information presented below is the summary of the researcher’s presentations, and do not necessarily represent my opinion or the opinion of Evidence Based Birth. I am merely summarizing the presentations for you. A special thank you to the Evidence Based Birth moms and babies who provided photos for this blog article.
Review of Maternal Experience of Having Diabetes Mellitus in Pregnancy
Presented by Ilana R. Azulay Chertok, PhD, MSN, RN, IBCLC
West Virginia University School of Nursing
Dr. Chertok presented a presentation called, “In their own Voices: Experience of Women with Gestational Diabetes”
The purpose of her presentation is to describe the perspectives and feelings of women with GDM. Dr. Chertok conducted a review of the qualitative literature and found 10 studies that interviewed women to find out their perspectives on GDM. There was a wide variety of women who participated. Six of the studies took place in the US, and the other studies took place in Australia, Canada, and Sweden. The women were ethnically diverse and included immigrants and minority women.
How did women respond to the diagnosis? They were shocked, surprised, upset, depressed, disappointed, scared, and felt like they had failed (flunked).
The women had many different perceptions for the possible cause of the GD, including fate, the will of God, the “evil eye,” lifestyle, stress, diet, weight, heredity, and family history.
What these women wanted was accurate information and guidance. But they didn’t get this. Most women felt that their providers were unable to guide them appropriately. Their providers lacked accurate information, didn’t have time to explain diagnosis, symptoms, and management, and were unavailable for answers and guidance. Providers frequently used “scare tactics” and “preached” to women to make them comply with recommendations.
Overall, women felt there was a lack of tailored care for GDM:
- They were often told what they already knew, and often this information was presented in a condescending manner
- Providers did not address the information women wanted to hear
- Nutrition information that was given was not specific to culture
- Many women from different cultures felt like they were kept in the dark– Latina women in particular felt they were not given information in a language that they could understand
Women were concerned about their infants. They experienced:
- Constant worry and anxiety over their baby
- Self-blame and guilt
- More medicalization throughout the prenatal care and delivery
- In the postpartum period, women felt added guilt when infant needed additional monitoring. They continued to feel conern about infant’s health and risk for Type 2 diabetes after birth, and they were very scared by hypoglycemia episodes in the infant
Women were concerned about themselves:
- It was difficult to make a sudden lifestyle change.
- Women felt limited and restricted–“everything is so strict,” “you are not free,” “you have to follow rules,” “you can’t do this, you can’t do that.”
- They faced barriers to managing GDM– not enough money, healthy food was more expensive, self-care discomfort (having to prick finger and give shots), lack of time for exercise, culturally inappropriate instructions, lack of support or understanding by others.
- They worried about their future risk of diabetes– some women knew they were at increased risk, other women didn’t know, some got inaccurate information.
- One silver lining– women saw this as a way to make positive change– to lose weight, eat healthier, and keep in shape. There IS a positive way to frame GDM.
Importantly, women felt like they lost control:
- Women were subjected to scrutiny and lectures by family and friends about behaviors– they felt judged about everything (food, activity, testing).
- The healthcare providers usurped control. “I was constantly fighting toward the end… I really felt like they were taking all this control away from me…” “They know your body, even though it’s your body.”
- Medical needs were often disregarded– for example, one woman requested an induction and it was denied her, then she needed a C-section; other women had to fight to get insulin prescribed.
- Women felt a loss of internal control when they were unable to control blood sugar: “It was totally out of my control.”
- They weren’t informed– “It would be nice to know the tests you are going to need without coercion and with time to consider options.”
Based on women’s experiences, Dr. Chervak recommends:
- Use a multidisciplinary team and make the patient an integral part of the team
- Avoid using scare tactics, frame information positively
- Provide access to accurate, clear information about GDM
- Point women to accurate GDM websites and online support/chat rooms
- Use culturally appropriate language and information
- Encourage support from family and friends, introduce women to role models (other women with GDM)
- Discuss and write a birth plan
Evidence Based Birth® is now offering you access to an online class where you can learn in-depth facts about suspected big babies. You want to find out details about the class, including information on how you can earn continuing education contact hours. Also, because you found out about the class while reading this blog article, I will give you a special promotional deal of $5 off! Just use the coupon code “BlogArticle” to get your discount!
Pro Status Quo
Presented by Brian M. Casey, MD
Obstetrics and Gynecology
The University of Texas Southwestern Medical Center
In Dr. Casey’s opinion, GDM is really about the excessive fetal growth associated with high blood sugars. That was not the initial intent behind the original work on GDM, because back then, those researchers were most interested in predicting risk of women developing Type II diabetes later in life. However, later we “woke up” to the fact that there were more immediate consequences to GDM.
In 1979 we had the first international conference that identified GDM as an important risk factor that should be identified and treated. Since then, we’ve had multiple study groups and organizations grappling with how to make this diagnosis and how to deal with it. That leads us to today’s status quo.
What is the status quo today with GDM in the U.S.?
The status quo is a systematic 2-step approach to identify women who are at the highest risk for fetal overgrowth from maternal high blood sugars. If you look at 100 women in the U.S., 1 would have overt Type II diabetes, and 7 would be diagnosed with GDM.
So what’s the impetus for change?
Well, the HAPO study showed a continuous increase in risk for several outcomes with GDM, including birth weights and C-section. If we look at the exact relationship between the mother’s blood sugars and large for gestational age babies, you can see that the risk increases incrementally. But it’s not really until you get down to the highest blood sugars that you get to odds ratios above 3. Dr. Casey quoted an article published recently by an epidemiologist, who wrote that in observational studies, odds ratios of 1 to 3 should make us skeptical that there is really an impact on outcomes.
Then the ACHOIS trial showed that women with GDM who received treatment had a significant reduction in a combined endpoint of serious outcomes, and a decrease in fetal growth. In addition, the MFMU trial showed that there was no difference between treatment and control with regard to serious outcomes, but there was a decrease in shoulder dystocia and C-sections (although the principal investigator from the MFMU trial said that he doesn’t think that a decrease in C-sections would happen in women who are treated for lower blood sugars).
So just now we finally have evidence to support what we’ve been doing all along.
But then the International Association of Diabetes and Pregnancy Study Groups (IADPSG) recommended a single-step test so that would universalize how we screen women for GDM. With this single 75 gram 2 hour test, women would only need one abnormal value for diagnosis, and there would be lower cut-offs than what have been typically used in the past.
So let’s look at these things. First, going from a 2-step process to a 1-step process: If you give the 1-hour 50 gram test, the same patients who score positive coud re-take it, and their results would be normal about 1/3 of the time. So it’s not the most accurate test. The 3-hour 100 gram test is only reproducible about 75% of the time, meaning that the same patient would get the same results about 75% of the time. There is a lot of variability in single glucose tests, which is why it would be concerning to diagnose women based on only one value. This means that a patient with a fasting value could have a wide variety in what their results are on any given day.
If we talk about lower thresholds, the new recommendations are based on a 1.75 higher risk (odds ratio) for large infants. This means that they determined the new cut-off value for GDM because women at that glucose level had a 1.75 times higher risk of giving birth to a big baby. BUT, 70-80% of overgrown infants are born to women WITHOUT GDM. So Dr. Casey asked, is this the best way to make an impact on large babies? Another important point to keep in mind is that the ACHOIS and MFMU studies had higher average fasting blood sugars than women in the HAPO study. So we don’t really have clinical trial data to support lowering the cut-off.
Now with the new recommendations, it will increase the percentage of women with GDM from 6-7% to about 18%, which means 11 additional women per 100. We don’t have strong evidence to show that it will be beneficial to identify these additional 11 women. The women that we are already diagnosing are not at very higher risk, and so what good would it do to include these women who have glucose levels below that?
So what are potential harms? Well, we cannot underestimate the effect of “labeling” women with GDM. The label of GDM has a profound effect on how healthcare providers treat women. The ACHOIS trial had a significant increase in admissions to the neonatal nursery, probably because of the labeling effect, because care providers were overly concerned about the babies. The labeling of women with GDM changes the way physicians treat women.
Any other harms? In one study (Feig et al., 1998) among women who were diagnosed with GDM 3-5 years later, researchers found that women with GDM were more worried about their own health, rated their child’s health more poorly, and rated their general health as poorer. GDM even impacted whether or not women would pursue another pregnancy in the future.
Another study found that as you decrease blood sugars in women with GDM, you increase the workload. What are the resource implications of increasing this workload? Dr. Casey shared his own personal story. He is the medical director for a GDM clinic. Right now, he has 2 half-day clinics per week, andthey see 120-150 women with GDM per week. At his clinic, there are 8,000 clinic visits annually, and these visits are run by 8 nurses, 2 diabetes educators, 1 certified nutritionist, 10-12 APNS, 1 fellow, and 3 faculty. If we doubled or tripled the diagnosis, Dr. Casey said this would “crush us.” They couldn’t do as good a job as they are doing right now with these women who are at highest risk.
Furthermore, the cost of gestational diabetes in the U.S. would go from $636 million per year to $2 billion per year.
What are the criteria for change? These criteria must be met:
1. Women must be clearly at risk for important health outcomes associated with a positive test
- This is supported by the HAPO study.
- But, maternal weight gain and BMI are confounders
2. The diagnostic test should be reliable
- We are going to rely strongly on a single result, which will lower the specificity of the test
3. There should be proof a treatment will prevent those outcomes
- We do not have evidence to support this right now, using a lower cut-off.
4. The intervention should be cost-effective
- There will be a significant increase in costs
- Any cost-effective analyses assume that there will be positive outcomes
- But right now, we have no data to tell us what the outcomes are going to be
- Improvement in short-term outcomes for these lesser-risk women is less likely
In conclusion, the status quo should be preferred, until we have evidence to support that there will be benefits of treatment at the lower cut-off point.
Pro International Association of Diabetes and Pregnancy Study Group
Boyd E. Metzger, M.D.
Emeritus Professor, Department of Medicine
Division of Endocrinology, Metabolism and Molecular Medicine
Feinberg School of Medicine
From Dr. Metzger’s perspective, the impetus for change did not come from HAPO. It started back In 1978, when a group of people appointed by the National Institutes of Health came together to work on the diagnoses of diabetes and GDM. It took a year and a half for this group to come out with some very important recommendations. Importantly, they established a universal framework for diabetes, including the language of “Type 1” and “Type 2” diabetes. They also described the diagnostic criteria for diabetes, which were linked to outcomes. On the other hand, GDM was considered a major exception to the rule, because there was less evidence available– so they could not link diagnostic criteria to outcomes. Instead, the working group linked the diagnostic criteria for GDM to women’s risk of developing Type II diabetes in the future. There was a minority group even back then who thought that lower values should be diagnosed as GDM.
Between 1979-2013, there was much controversy and discussion over GDM. The 2 major questions are:
1) Can adverse outcomes be linked to hyperglycemia during pregnancy, or are they due to confounding factors such as obesity?
2) Is there an effective treatment that can improve outcomes of GDM?
Well, the HAPO study helped answer the first question. The results indicated that as glucose levels rise, the frequency of a whole series of adverse outcomes increases (including birth weight and high insulin in the infants, pre-term birth, preeclampsia, shoulder dystocia, and jaundice). The researchers found that these GDM-related outcomes are independent of the mother’s age, body mass index, and family history of diabetes. So the link remains, even after you take into account confounding factors. These results were consistent at all of the HAPO study locations all around the world.
So then the IADPSG took these results and used them to develop “outcome-based” criteria for diagnosing GDM.
Before they started the HAPO study, the IADPSG researchers agreed upon odds ratio that they would use for a new cut-off point (1.75). This means that if there was a 1.75 increase in the risk of high birth weight, fetal fat distribution, and high insulin levels in the infant, then that would be the glucose level where they would set the cut-off for GDM.
Why did they pick birthweight, fetal fat distribution, and high insulin levels to be the main outcomes on which they based the IADPSG diagnostic criteria? Because the HAPO study was powered to detect differences in these outcomes, and because these outcomes are linked with other outcomes (like C-sections). The group making the IADPSG recommendations also looked at the HAPO data using either 1 abnormal value or 2 abnormal values to diagnose GDM. They considered other possible combinations, other odds ratios (such as 1.5 or 2.0). In the end, because poor outcomes rise in a linear fashion corresponding to blood sugars, where you draw the line is an arbitrary decision. But they felt like this was the best place to draw the line. It should be noted that back in the 1960’s when the first cut-off for GDM was proposed (the O’Sullivan approach), this was also an arbitrary cut-off.
When Dr. Metzger addressed the concerns about the potential increased number of diagnoses, he said that the number of diagnoses of GDM should reflect what we see in the general non-pregnant female population. The most recent NHANES data showed that 2.8% of U.S. women age 18 to 44 have diagnosed diabetes, 1.7% have undiagnosed diabetes, and 26.4% have pre-diabetes (impaired glucose tolerance). Taking into account the fact that during pregnancy you see an increase in insulin resistance, we are not going to see a decrease in women with glucose abnormalities. So if 30.9% of U.S. women age 19-44 have impaired glucose tolerance, why do we consider 18% an unreliable number?
Another concern that people have raised is that the recommendations are not validated by evidence from randomized, controlled trials. Dr. Metzger compared the ACHOIS and MFMU populations with the HAPO study, and he stated that the populations were not that different. The fasting glucose was very similar, and there was only a slight difference in the 2-hour glucose values.
So why should we use the IADPSG recommendations?
- Cut-offs would actually be based on mother-baby outcomes
- Using a 75 gram 2 hour test would allow consistency around the world for both pregnant and non-pregnant patients
- Because we have an international source of data (the HAPO study) with input from multiple countries, this will enhance the adoption of these criteria
- We have been waiting for 30 years for this change, so please, let’s move forward
Presented by Edmond A. Ryan, MD
Professor, Department of Medicine
Division of Endocrinology
University of Alberta
Dr. Ryan has a website for women with GDM– you should check it out. He is from Canada and in his talk he offered an alternative way forward.
It is well-known that Type I diabetes in pregnancy has risks of congenital defects. It is also known that if Type II diabetes goes undiagnosed and untreated, then it can have risks for mother and baby. We also know that hyperglycemia (GDM) has adverse outcomes.
Dr. Ryan gave a brief overview of the ACOG-recommended 2-step process, the IADPSG 1-step process, and an alternative 2-step process that he is proposing.
In the alternative process, women would be screened with the 50 gram 1 hour test, and if women test in the upper limits than they would be presumed to have GDM. Others who screen positive would undergo a 75 gram 2 hour test, with 1 abnormal value being considered GDM.
Dr. Ryan thinks that we should be doing what’s best for babies– and what is truly best for babies based on evidence. One of the problems with the tests for GDM is that they are not reproducible. If you have a slightly abnormally result, you could take the test next week and it could be completely normal. However, if the TWO-STEP process is used, a woman with a single abnormality on a single test is less likely to have an abnormal test on repeat. In Dr. Ryan’s opinion, we need to be VERY careful about making a diagnosis based on a single lab value, because a single test is poor evidence.
The alternative that Dr. Ryan is proposing would be to use an odds ratio cut-off of 2.0 instead of 1.75. This means that whatever glucose levels result in a 2-fold risk for poor outcomes, then we should use that as a cut-off (the IADPSG group is recommending the 1.75 cut-off). Dr. Ryan talked about how obesity is a stronger predictor of large baby than GDM test results (Ricart et al., 2005). In the HAPO study with 23,316 women, there were 311 cases of shoulder dystocia and/or birth trauma. More than three-fourths of these cases occurred in women with glucose levels BELOW the IADPSG threshold– in otherwords, most cases of shoulder dystocia occur in women without GDM. Also, in the majority of women with high glucose, their outcomes were just fine.
Dr. Ryan states that when you look at the evidence in total, the current ACOG recommendations for the 2-step screening process are not based on good evidence. On the other hand, the IADPSG and the “alternative” methods are evidence-based, because they link the diagnosis with actual outcomes. So now we need to look at the benefit/harm ratio of these methods. Whichever one has the best benefit/harm profile would be the one we should pick.
What about benefits and harms of testing for GDM? Some practical drawbacks include patient time, inconvenience, and discomfort, as well as the number of blood draws. The 3 hour 100 gram test (required by ACOG in their 2-step process) requires more time off work and leads to more nausea.
Interestingly, the benefits of treating GDM are not striking. Large numbers of women would need to be treated in order to show a benefit. In the ACHOIS study, there was a reduction of the combined endpoint of serious complications– but not in each these complications individually. There were 5 stillbirths in the control group, but 1 was small for gestational age (not related to GDM), and 1 was congenital malformations (not related to GDM). The difference in stillbirths between groups was not statistically significant.
Diagnosing GDM has actual consequences. The mother would need more clinic visits and would need to undergo frequent glucose testing, therapy, obstetrical monitoring/induction, and she would have to endure the “label” of diabetes. The label of GDM increases a woman’s insurance premium in 48 states. For the baby, this could lead to unnecessary NICU admissions (due to the “label”), and early inductions.
In the U.S., our “pie” is only so big, and the “sequester pie” is even smaller. Is it really fair for society to spend so much money on these additional diagnoses of GDM, when things like obesity may be more important?
Some people say until we get definitive studies we should do nothing. But in practice, Dr. Ryan says that we need to do something. The truth is, the most important thing is to catch women with overt, undiagnosed Type II diabetes. In the meantime, we should hold onto the 50 gram screening test– systematic review supports its use, women prefer it, and it’s more cost-effective. This test would be followed by the 2 hour 75 gram test in women who screen positive (unless they have a result of more than 200, in which case they would automatically be diagnosed with GDM).
In summary: GDM is a treatable cause of large babies and it is worth detecting. However, GDM accounts for a small minority of large babies– other factors are involved, especially obesity. We need to demonstrate elevated glucose on 2 occasions. In the long-term, we need to find the real culrprit for most large babies, since it’s obviously not GDM. In the short-term, we need to determine whether treating obesity is a more acceptable and practical approach. In the mean-time, the alternative 2-step approach for diagnosing GDM is the most reasonable step forward.
Want to read more about gestational diabetes? Check out these articles written byEvidence Based Birth®: