of 36
8/10/2019 Gestational Diabetes Mellitus508
1/36
NIH Consensus Development Conference onDiagnosing Gestational Diabetes Mellitus
NIH Consensus Development Conference Statements
Volume 29, Number 1
March 46, 2013
NATIONAL INSTITUTES OF HEALTHOffice of the Director
NIHTurning Discovery Into Health
8/10/2019 Gestational Diabetes Mellitus508
2/36
About the NIH Consensus Development Program
National Institutes of Health (NIH) consensus statements areprepared by independent panels of health professionals
and public representatives on the basis of (1) the results ofa systematic literature review prepared under contract withthe Agency for Healthcare Research and Quality (AHRQ),(2) presentations by investigators working in areas relevantto the conference questions during a 2-day public session,(3) questions and statements from conference attendeesduring open discussion periods that are part of the publicsession, and (4) closed deliberations by the panel duringthe remainder of the second day. This statement is anindependent report of the panel and is not a policystatement of NIH or the Federal Government.
The statement reflects the panels assessment of medicalknowledge available at the time the statement was written.Thus, it provides a snapshot in time of the state ofknowledge on the conference topic. When reading thestatement, keep in mind that new knowledge is inevitablyaccumulating through medical research, and that theinformation provided is not a substitute for professional
medical care or advice.
Reference Information
Individuals who wish to cite this statement should use thefollowing format:
VanDorsten JP, Dodson WC, Espeland MA, Grobman WA,Guise JM, Mercer BM, Minkoff HL, Poindexter B,Prosser LA, Sawaya GF, Scott JR, Silver RM, Smith L,Thomas A, Tita ATN. National Institutes of HealthConsensus Development Conference Statement:Diagnosing Gestational Diabetes Mellitus. NIH ConsensState Sci Statements. 2013 March 4-6; 29(1): 1-30.
8/10/2019 Gestational Diabetes Mellitus508
3/36
i
Diagnosing Gestational Diabetes MellitusNIH Consensus Development Conference on
NIH Consensus Development Conference Statements
Volume 29, Number 1
March 46, 2013
NATIONAL INSTITUTES OF HEALTHOffice of the Director
NIHTurning Discover
y Into Health
8/10/2019 Gestational Diabetes Mellitus508
4/36
ii
Disclosure Statement
All of the panelists who participated in this conferenceand contributed to the writing of this statement wereidentified as having no financial or scientific conflictof interest, and all signed forms attesting to this fact.Unlike the expert speakers who present scientific dataat the conference, the individuals invited to participateon NIH Consensus Panels are reviewed prior toselection to ensure that they are not proponents ofan advocacy position with regard to the topic andare not identified with research that could be usedto answer the conference questions.
For more information about conference procedures,please seehttp://prevention.nih.gov/cdp/about.aspx.
Archived Conference Webcast
The NIH Consensus Development Conference:
Diagnosing Gestational Diabetes Mellitus waswebcast live March 4-6, 2013. The webcast isarchived and available for viewing free of charge atprevention.nih.gov/cdp/conferences/2013/gdm/default.aspx.
http://prevention.nih.gov/cdp/about.aspxhttp://prevention.nih.gov/cdp/about.aspxhttp://localhost/var/www/apps/conversion/tmp/scratch_4/prevention.nih.gov/cdp/conferences/2013/gdm/default.aspxhttp://localhost/var/www/apps/conversion/tmp/scratch_4/prevention.nih.gov/cdp/conferences/2013/gdm/default.aspxhttp://prevention.nih.gov/cdp/about.aspx8/10/2019 Gestational Diabetes Mellitus508
5/36
1
Abstract
Objective
To provide healthcare providers, patients, and the generalpublic with a responsible assessment of currently availabledata on diagnosing gestational diabetes mellitus.
Participants
A non-U.S. Department of Health and Human Services,nonadvocate 15-member panel representing the fieldsof obstetrics and gynecology, maternal-fetal medicine,pediatrics, diabetic research, biostatistics, womens healthissues, health services research, decision analysis, healthmanagement and policy, health economics, epidemiology,and community engagement. In addition, 16 expertsfrom pertinent fields presented data to the panel andconference audience.
Evidence
Presentations by experts and a systematic review of theliterature prepared by the University of Alberta Evidence-based Practice Centre, through the Agency for HealthcareResearch and Quality (AHRQ). Scientific evidence wasgiven precedence over anecdotal experience.
Conference Process
The panel drafted its statement based on scientificevidence presented in open forum and on publishedscientific literature. The draft statement was posted athttp://prevention.nih.gov/for public comment and thepanel released a final statement approximately 10 weekslater. The final statement is an independent report of the
panel and is not a policy statement of the NIH or theFederal Government.
http://prevention.nih.gov/http://prevention.nih.gov/8/10/2019 Gestational Diabetes Mellitus508
6/36
2
Conclusions
At present, GDM is commonly diagnosed in the UnitedStates using a 1-hour screening test with a 50-gramglucose load followed by a 3-hour 100-gram glucosetolerance test (a two-step approach) for those found tobe abnormal on the screen. This approach identifiesapproximately 5% to 6% of the population as havingGDM. In contrast, newly proposed diagnostic strategiesrely on the administration of a 2-hour glucose tolerancetest (a one-step approach) with a fasting componentand a 75-gram glucose load. These strategies differon whether a 1-hour sample is included, whethertwo abnormal values are required, and the diagnosticcutoffs that are used. The International Association ofDiabetes and Pregnancy Study Groups (IADPSG) hasproposed diagnostic thresholds based on demonstratedassociations between glycemic levels and an increasedrisk of obstetric and perinatal morbidities. The panel
considered whether a one-step approach to thediagnosis of GDM should be adopted in place of thetwo-step approach. The one-step approach offers certainoperational advantages. The current two-step approachis used only during pregnancy and is largely restricted tothe United States. There would be value in a consistent,international diagnostic standard across ones lifespan.This unification would allow better standardization of
best practices in patient care and comparability ofresearch outcomes. The one-step approach also holdspotential advantages for women and their healthcareproviders, as it would allow a diagnosis to be achievedwithin the context of one visit as opposed to two.However, the one-step approach, as proposed by theIADPSG, is anticipated to increase the frequency ofthe diagnosis of GDM by twofold to threefold, to a
prevalence of approximately 15% to 20%. There areseveral concerns regarding the diagnosis of GDM inthese additional women. It is not well understood
8/10/2019 Gestational Diabetes Mellitus508
7/36
3
whether the additional women identified by thisapproach will benefit from treatment, and if so, towhat extent. Moreover, the care of these women will
generate additional direct and indirect healthcarecosts. There is also evidence that the labeling of thesewomen may have unintended consequences, such asan increase in cesarean delivery and more intensivenewborn assessments. In addition, increased patientcosts, life disruptions, and psychosocial burdens havebeen identified. Available studies do not provide clearevidence that a one-step approach is cost-effective in
comparison with the current two-step approach. Aftermuch deliberation, the panel believes that there areclear benefits to international standardization withregard to the one-step approach. Nevertheless, atpresent, the panel believes that there is not sufficientevidence to adopt a one-step approach. The panelis particularly concerned about the adoption of newcriteria that would increase the prevalence of GDM,
and the corresponding costs and interventions,without clear demonstration of improvements in themost clinically important health and patient-centeredoutcomes. Thus, the panel recommends that thetwo-step approach be continued. However, given thepotential benefits of a one-step approach, resolutionof the uncertainties associated with its use wouldwarrant revision of this conclusion.
8/10/2019 Gestational Diabetes Mellitus508
8/36
4
Introduction
Gestational diabetes mellitus (GDM) is a condition ofcarbohydrate intolerance of varying severity that beginsor is first recognized during pregnancy, and is one ofthe most common complications of pregnancy. In somecases, GDM is actually type 2 diabetes that has notpreviously been diagnosed, but for most patients theglucose intolerance disappears soon after delivery. Theprevalence of GDM varies because of different screeningand diagnostic criteria, populations, race, ethnicity, age,
and body composition. Using current testing criteria inthe United States, GDM prevalence is estimated to bebetween 5% and 6%, affecting approximately 240,000of the more than 4 million births occurring annually.Multiple studies have shown increases in GDM amongdiverse populations during the 1990s and early 2000s.This observed increase in GDM nationally is consistentwith changes in known risk factors for GDM: advanced
maternal age, family history of diabetes, higher bodymass index, and changing racial and ethic demography.All of these risk factors have changed in the past 20 yearsin such a way as to increase the prevalence of GDM; forexample, more than 20% of women in the United Statesare now obese as they enter pregnancy. Gestationaldiabetes mellitus is more common among certain ethnicgroups, such as African American, Asian, Hispanic, and
Native American women, compared with non-Hispanicwhite women, and recent patterns of immigration haveled to increases in the numbers of many of these groups.These high-risk groups are not evenly distributed in theUnited States, leading to a disproportionate prevalenceof GDM in different geographic regions.
Adverse short-term and long-term health outcomes forboth the mother and her offspring have been associatedwith the diagnosis of GDM. For the mother, theseoutcomes include gestational hypertension (pregnancy-induced high blood pressure) and preeclampsia (high
8/10/2019 Gestational Diabetes Mellitus508
9/36
5
blood pressure and proteinuria developed in pregnancy).The mother is also at increased risk for the laterdevelopment of type 2 diabetes and other long-term
metabolic complications, such as metabolic syndromeand cardiovascular disease. Excess glucose crossesthe placenta and can cause adverse fetal effects. Fetalhyperinsulinemia (high levels of insulin in the blood) canlead to excess fetal size, with increased risks of shoulderdystocia (an impacted shoulder that requires additionalobstetric manipulation during childbirth), cesareandelivery, respiratory distress syndrome, and neonatal
metabolic complications.At this time, most obstetric providers in the United Statesscreen for GDM with a 50-gram glucose challenge test(GCT, measuring serum glucose 1 hour after a womandrinks a 50-gram oral glucose solution) followed by anoral 100-gram glucose tolerance test (OGTT, in whichfour blood samples are drawn over a 3-hour period aftera woman drinks 100 grams of glucose solution) if theGCT result is abnormally high. This two-step approachhas been recommended by the American College ofObstetricians and Gynecologists (ACOG). Depending onwhich GCT cutoff is chosen, 14% to 23% of patients willrequire the diagnostic OGTT.
Despite the near uniformity of current practice in theUnited States, a number of controversies remain: the
value of routine screening, the most appropriate method(e.g., one-step compared with two-step), glycemic valueson screening tests that are used to define abnormal,the number of abnormal values needed to make adiagnosis, and the effects of treatment on the short-termand long-term outcomes for women and their children.For example, in 2008, the U.S. Preventive Services TaskForce (USPSTF) determined that the current evidence is
insufficient to assess the balance between the benefitsand harms of screening women for GDM either beforeor after 24 weeks gestation. At the same time, otherssupport liberalizing the definitions, which would categorize
8/10/2019 Gestational Diabetes Mellitus508
10/36
6
more pregnant women as having GDM. The InternationalAssociation of Diabetes and Pregnancy Study Groups(IADPSG) has proposed a one-step approach (fasting,
1-hour, and 2-hour glucose measurements), where GDMis diagnosed by only one abnormal value. Using theirproposed cutoffs, this strategy is estimated to increasethe number of women labeled as GDM twofold tothreefold and could increase personal and societal costs.Therefore, clear evidence of substantive benefits from theIADPSG approach is needed to justify a change to thatdiagnostic technique.
The National Institutes of Health (NIH) ConsensusDevelopment Program is designed to address controversialquestions of public health importance when there may bediscordance between clinical practice and the availableevidence. Consensus Development Conferences addresstargeted, carefully defined questions, which prompt athorough review of the available evidence and solicitpresentations from subject matter experts. An objectivepanel then concludes with a Consensus Statement, whichaddresses the critical questions.
By necessity, this panel, Diagnosing Gestational DiabetesMellitus, cannot address every controversy surroundingGDM and focused on diagnosis. However, the panel iscognizant of the fact that most healthcare providers inthe United States currently screen, and will continue to
screen, for this condition. The panel also is aware thathealthcare providers will continue to monitor and treatmost patients based on whatever diagnosis of GDM isused, and that those will be expensive undertakings,with potentially negative consequences for those falselycategorized as having GDM. Although those facts haveinfluenced deliberations, the panel concentrated on thediagnosis of GDM, not on the merits of routine screening
or on issues of treatment and its effects. Simultaneously,the USPSTF will reexamine the issue of routine screening.In combination, the panel hopes to clarify an approach toGDM that may resolve key controversies.
8/10/2019 Gestational Diabetes Mellitus508
11/36
7
To provide healthcare providers, public health practitioners,policymakers, and the general public with a comprehensiveassessment of diagnosing GDM, the Eunice Kennedy
Shriver National Institute of Child Health and HumanDevelopment, National Institute of Diabetes and Digestiveand Kidney Diseases, National Institute of Nursing Research,Office of Research on Womens Health, Centers for DiseaseControl and Prevention, and the NIH Office of DiseasePrevention convened a Consensus Development Conferenceon March 46, 2013, to assess the available scientificevidence. The panel included experts in the fields of
maternal and fetal medicine, healthcare economics,obstetrics and gynecology, decision analysis and biostatistics,ethics, clinical pediatrics, diabetes education, epidemiology,neonatology, endocrinology, and a public representative.A Planning Committee developed seven questions to beaddressed by the Consensus Development Panel:
1. What are the current screening and diagnosticapproaches for gestational diabetes mellitus, what arethe glycemic thresholds for each approach, and howwere these thresholds chosen?
2. What are the effects of various diabetes mellitusscreening/diagnostic approaches for patients,providers, and U.S. healthcare systems?
3. In the absence of treatment, how do health outcomes
of mothers who meet various criteria for gestationaldiabetes mellitus and their offspring compare withthose who do not?
4. Does treatment modify the health outcomes of motherswho meet various criteria for gestational diabetesmellitus and their offspring?
5. What are the harms of treating gestational diabetes,
and do they vary by diagnostic approach?
6. Given all of the above, what diagnostic approach(es) forgestational diabetes mellitus should be recommended,if any?
8/10/2019 Gestational Diabetes Mellitus508
12/36
8
7. What are the key research gaps in the diagnosticapproach of gestational diabetes mellitus?
During the first 2 days of the conference, experts
presented information on each of the key questions.After weighing the scientific evidenceincluding thedata presented by the speakers, input from attendees,and a formal evidence report commissioned through theAgency for Healthcare Research and Quality (AHRQ)an independent panel prepared and presented a draftof this Consensus Development Conference Statementaddressing the conference questions.
1. What are the current screening anddiagnostic approaches for gestationaldiabetes mellitus, what are the glycemicthresholds for each approach, and how
were these thresholds chosen?Testing for diabetes in pregnancy has been a routine partof obstetric practice since OSullivan published results forthe OGTT in pregnancy more than 40 years ago. Currently,most practices use either a one-step or two-step approachfor GDM diagnosis.
Two-step approaches, proposed by the National
Diabetes Data Group (NDDG) and Carpenter and Coustan,are commonly used in the United States and involve theadministration of a screening 50-gram GCT to the patientwithout regard to fasting (first step). If the plasma glucoselevel measured 1 hour after the load is less than a selectedcutoff (usually 130 mg/dL, 135 mg/dL, or 140 mg/dL),the woman is considered GDM-negative, and no furthertesting is required. If the glucose level is greater than the
cutoff, then a diagnostic test (second step) is needed toconfirm the diagnosis of GDM. This second step involvesa 100-gram oral glucose tolerance test (100-gram 3-hourOGTT) given while the patient is fasting; the fasting
8/10/2019 Gestational Diabetes Mellitus508
13/36
9
1-hour, 2-hour, and 3-hour post-load glucose levels aremeasured and compared with recommended diagnosticcriteria (Carpenter and Coustan or NDDG cutoffs) to
confirm or reject the diagnosis of GDM (Table 1). Thetwo-step approaches were not developed to diagnosediabetes in pregnancy per se, but rather to identifywomen at risk of developing diabetes mellitus later inlife. Of note, the Canadian Diabetes Association usesa modified two-step approach that involves a 50-gramGCT with a diagnostic cutoff at 140 mg/dL (7.8 mmol/L),followed by a 75-gram 2-hour OGTT for the second step.
Table 1:Criteria and glucose thresholds for the
diagnosis of GDM
Approach Criteria*Fastingmg/dL
1-hourmg/dL
2-hourmg/dL
3-hourmg/dL
Two-Step
(100g
load)
CarpenterandCoustan
95(5.3mmo/L)
180(10.0mmol/L)
155(8.6mmol/L)
140(7.8mmol/L)
NDDG 105(5.8mmol/L)
190(10.6mmol/L)
165(9.2mmol/L)
145(8.0mmol/L)
CDA 95(5.3mmol/L)
191(10.6mmol/L)
160(8.9mmol/L)
OneStep
(75gload) WHO 126
(7.0mmol/L)140(7.8mmol/L)
IADPSG 92(5.1mmol/L)
180(10mmol/L)
153(8.5mmol/L)
* NDDG = National Diabetes Data Group; CDA = Canadian DiabetesAssociation; WHO = World Health Organization; IADPSG = InternationalAssociation of Diabetes and Pregnancy Study Groups.
Single-step approachesproposed by the World HealthOrganization (WHO) and the IADPSG are commonlyused outside of the United States to diagnose GDM.In the single-step approach, a 75-gram 2-hour OGTTis administered to the fasting woman. Using the WHOapproach, fasting and 2-hour post-load glucose levelsare measured, and using the IADPSG approach, fasting,
8/10/2019 Gestational Diabetes Mellitus508
14/36
10
1-hour, and 2-hour glucose levels are evaluated againstrecommended criteria to confirm or refute the diagnosis ofGDM. Table 1 summarizes the GDM diagnostic glycemic
cutoffs for these criteria. Both the WHO and the IADPSGconsider any single abnormal value as diagnostic of GDM,but the IADPSG consensus cutoffs are the only onesthat are based on pregnancy outcomes (glucose valuesassociated with a 1.75-fold increase in selected adversepregnancy outcomes).
2. What are the effects of various diabetesmellitus screening and diagnosticapproaches for patients, providers,and U.S. healthcare systems?
Patients
Adopting the IADPSG criteria would substantiallyincrease the proportion of women diagnosed with GDM.Changing to a screening and diagnostic approach thatrequires every pregnant woman to undergo an OGTTis burdensome. In addition, the fasting state may bedifficult and uncomfortable for some women. Thediagnosis of GDM carries considerable inconveniencefor patients, regardless of the criteria used. They must
self-monitor their blood glucose levels several times aday and carefully monitor what they eat. They will needto meet with a registered dietitian, a diabetes educator,or both, resulting in additional appointments. Also (anddespite a lack of clear efficacy), they often undergofetal testing such as non-stress testing and additionalobstetric ultrasonography. These extra proceduresand provider visits require additional time and create
additional challenges regarding transportation, childcare, or employment and may result in additional out-of-pocket costs. These problems are likely enhancedfor vulnerable populations.
8/10/2019 Gestational Diabetes Mellitus508
15/36
11
Providers
Increasing the proportion of women with GDM bytwofold to threefold has considerable implications forhealthcare providers. Two randomized clinical trials havedemonstrated that when additional women are diagnosedwith GDM, the number of prenatal visits or visits to ahealthcare provider increases. These visits would requireadditional clinical resources as well as the services ofregistered dietitians and diabetes educators. In onestudy of two large hospitals in Australia, it was estimatedthat the workload would increase approximately 30%if new diagnostic criteria for GDM were implemented.One estimate is that the IADPSG criteria would resultin 450,000 more patient education visits, one millionmore clinic visits, and one million more prenatal testingappointments each year in the United States.
U.S. Healthcare Systems
The additional outpatient visits and testing describedabove will affect hospitals and payers. There may becapacity constraints relating to additional volume oflaboratory testing. Other more difficult to quantify factorsinclude increased responsibility on labor and deliverysuites due to inductions of labor and increased timeon postpartum rounds due to potentially more frequentcesarean deliveries.
In 2009, it was estimated that the annual cost in theUnited States for the care of GDM would increase from$636 million to $2 billion. Other published results suggestthat direct medical and patient time costs would both behigher if the IADPSG protocol were adopted, even whenaccounting for the savings associated with the potentialprevention of related complications. The results of
economic analyses that weigh the tradeoff betweencosts, health benefits, and potential harms vary widely,and these analyses do not provide sufficient informationto compare the various approaches likely due to
8/10/2019 Gestational Diabetes Mellitus508
16/36
12
uncertainty regarding the health benefits of increaseddiagnosis of GDM. Cost-effectiveness analyses are morefavorable to screening if it is assumed that screening
leads to the prevention of future type 2 diabetes inwomen previously diagnosed with GDM. Long-termfollow-up studies confirming this benefit, however,are lacking.
3. In the absence of treatment, how do
health outcomes of mothers who meetvarious criteria for gestational diabetesmellitus and their offspring comparewith those who do not?
Many high-quality studies have evaluated maternal andfetal outcomes among women with untreated GDMcompared with those without GDM. Although these
studies employed various diagnostic criteria, severalfindings have been consistent. In terms of maternaloutcomes, studies have shown that GDM increasesrisks of cesarean delivery, preeclampsia, and gestationalhypertension. Studies also indicate increased risk forthe later development of type 2 diabetes and otherlong-term metabolic complications.
In terms of fetal outcomes, methodologically strongstudies have shown a continuous relationship betweenincreasing glucose levels and increasing incidenceof large-for-gestational-age infants and infants withmacrosomia (a condition in which the newbornis significantly larger than average). In addition, aconsistently higher risk of shoulder dystocia hasbeen found among women with a diagnosis of GDM
compared with those without; shoulder dystocia canlead to rare, but important, outcomes such as brachialplexus injury. Some studies report neonatal hypoglycemia
8/10/2019 Gestational Diabetes Mellitus508
17/36
13
(low blood glucose) and hyperbilirubinemia (excessbilirubin in the blood) among neonates born to womenwith GDM, although the evidence supporting these
associations has not been consistent. A relationshipbetween GDM and subsequent childhood obesityhas been found in some but not all studies. The effecton longer term outcomes in the offspring, includingtype 2 diabetes mellitus, is unclear. Of note, the maternaland fetal risks have been largely defined using thetraditional two-step diagnostic approach for GDM.Milder forms of GDM diagnosed through other
strategies may not be associated with theseadverse outcomes to the same degree.
The Hyperglycemia and Adverse Pregnancy Outcomes(HAPO) study demonstrated that the magnitudes ofmaternal and fetal risks increase with the severity ofmaternal hyperglycemia (high blood glucose). The HAPOstudy evaluated glucose tolerance at 24 weeks to 32weeks during pregnancy in 25,505 pregnant women from15 centers in nine countries, providing information on aheterogeneous, multinational, ethnically diverse group ofwomen. For women with hyperglycemia during pregnancythat was not so severe as to require unmasking in orderto inititiate treatment, i.e. plasma glucose levels greaterthan 105 mg/dL [fasting], greater than 200 mg/dL [2hour], or greater than 160 mg/dL [random], increasing
maternal glucose levels were related to increasedinfant birth weight, body fat, and cord C-peptide (ameasure of insulin production in the infant) above the90th percentile, and increased primary cesarean deliveryrates. In addition, these women also had increasedrisks for premature delivery, preeclampsia, shoulderdystocia or birth injury, and hyperbilirubinemia. Neonatalhypoglycemia and admissions to neonatal intensive
care units also were more common in infants born tomothers meeting the IADPSG criteria (used in theHAPO study) for GDM.
8/10/2019 Gestational Diabetes Mellitus508
18/36
14
4. Does treatment modify the healthoutcomes of mothers who meet variouscriteria for gestational diabetes mellitusand their offspring?
Very few well-designed, high-quality studies haveattempted to estimate the benefit of treatment of GDM.These treatments included self-monitoring of bloodglucose, medical nutrition therapy, and insulin in somepatients. Criteria for the diagnosis of GDM varied, buttwo-step approaches were used in the two largest
randomized trials. Women with more severe forms ofGDM were not included in the studies.
Maternal Outcomes
Treatment of GDM reduced the risk for hypertensivedisorders of pregnancy by approximately 40%.Shoulder dystocia risk was reduced with treatment
by approximately 60%; however, as shoulder dystociawas a rare event, the absolute risk changed from only3.5% (untreated) to 1.5% (with treatment). Anotherconsistent finding among the studies was that thetreatment of GDM did not increase the risk ofcesarean delivery.
Results were not consistent among studies for maternal
weight gain and risk for induction of labor; therefore,the panel could draw no conclusions on the effect oftreatment on these two maternal outcomes. Evidencewas absent or insufficient to conclude whether thereis an effect of treatment of GDM on birth trauma,body mass index at delivery, or long-term maternaloutcomes including type 2 diabetes mellitus, obesity,and hypertension.
8/10/2019 Gestational Diabetes Mellitus508
19/36
15
Fetal, Neonatal, and Child Outcomes
A pooled meta-analysis of five randomized clinical trialsfound a 50% reduction in macrosomia in infants bornto mothers who received treatment for GDM, althoughthe absolute difference in mean birth weight was lessthan 150 grams in the two largest studies. Similarly,randomized trials have demonstrated that infants ofmothers who received treatment for GDM were lesslikely to be large for gestational age (absolute riskreduction 6%). Randomized trials, however, have notshown a decrease in neonatal hypoglycemia in responseto maternal treatment of GDM. There are no sufficientdata available to conclude whether treatment of GDMmodifies neonatal morbidities such as prematurity,admission to neonatal intensive care units, or mortality.More studies are needed to evaluate the long-termmetabolic outcomes (obesity and risk of type 2 diabetesmellitus) of children born to women with GDM.
The panel strongly recommends caution when applyingthese results to clinical practice. First, participants inclinical trials typically are highly motivated individualswho are eager to adhere to even complex protocolsin academic medical center venues with very favorablestaff-to-patient ratios. These conditions are rarelypresent in the average clinical practice. Second, notall treatments employed in current daily practice were
studied. Oral anti-diabetic agents, such as glyburideand metformin, are notably absent. Third, differingthresholds for criteria to diagnose GDM may changethe size of the effect of the treatments for the entiregroup in unpredictable ways. Milder forms of GDMmay not benefit from treatment. Finally, application oftreatments purely for the sake of the benefits withoutregard for the costs would be inappropriate.
8/10/2019 Gestational Diabetes Mellitus508
20/36
16
5. What are the harms of treatinggestational diabetes, and do they varyby diagnostic approach?
A potential harm of the increased diagnosis of mild GDMis patient anxiety. It is generally accepted that patientsexperience short-term stress and anxiety when receivinga new diagnosis of a serious condition, including GDM,which could adversely affect their health. Nonetheless, itis unclear if long-term stress and anxiety are increased.In part, this is due to a paucity of data. Also, it is possible
that women may adapt to their diagnosis with diabetesmanagement, thereby decreasing their anxiety level.In addition to anxiety, women with a new diagnosis ofGDM have reported feelings of loss of control, shock,depression, fear, and disappointment.
Few studies directly addressed the emotional impact ofscreening for and diagnosis of GDM. One study noted a
lower sense of well-being, less positive experience oftheir pregnancies, and more concern about their health.Another study noted that women with GDM had increasedconcern about their babys health, and their own health,as well as a fear of losing personal control over theirhealth. Also, the over-diagnosis of GDM may lead tothe medicalization of pregnancy, which transforms anotherwise normal pregnancy into a disease.
There is considerable variability in the 2-hour glucosetolerance test. Results may differ in as many as 25% ofwomen if performed at different times. Thus, a one-steptest is likely to result in more false positive resultsthan a two-step test. In turn, positive tests will furtherincrease cost, inconvenience, and anxiety.
The harms of medical therapy for GDM are well known.
Medications such as insulin and oral hypoglycemicagents may cause hypoglycemia and other side effects.There are also obstetric harms associated with anincrease in labeling patients with the diagnosis of GDM.
8/10/2019 Gestational Diabetes Mellitus508
21/36
17
One randomized controlled trial has shown higher ratesof induction of labor in women with GDM comparedwith normal controls. Women with GDM are more likely
to undergo increased maternal and fetal monitoring.Subjective interpretation of ultrasound findings and fetalnon-stress tests produces a high rate of false positivesand unnecessary inductions of labor, failed inductions,and cesarean delivery. However, the literature has notyet addressed this possibility.
Cesarean rates may be higher in women with GDM,and it is uncertain whether treatment can mitigatethis increase. Cesarean delivery is associated witha higher rate of short- and long-term complications.There is concern about the rising cesarean rate bymany groups; the present rate in the United States is32.9%. Since the vaginal-birth-after-cesarean rate isnow less than 10%, most women, who delivered byprimary cesarean, will again deliver by repeat cesarean.With each subsequent pregnancy, the rate of placentaprevia (which occurs when a placenta partially or totallycovers the mothers cervix) and placenta accreta (aserious pregnancy condition that occurs when bloodvessels and other parts of the placenta grow too deeplyinto the uterine wall) increase dramatically. Theseconditions result in serious complications such ashemorrhage, infection, emergency hysterectomy,
and even death.A diagnosis of GDM may lead to more intensiveneonatal care, potentially separating mother and infant.One study indicated that infants born to mothers withthe diagnosis of GDM were more frequently admittedto an intermediate care nursery. (It is important tonote that protocols for increased surveillance varyamong hospitals.) Though there is theoretical risk for
small-for-gestational-age fetuses in patients treatedfor GDM, the two largest randomized clinical trialshave not confirmed this risk.
8/10/2019 Gestational Diabetes Mellitus508
22/36
18
6. Given all of the above, what diagnosticapproach(es) for gestational diabetesmellitus should be recommended, if any?
At present, GDM is commonly diagnosed in the UnitedStates using a 1-hour screening test with a 50-gramglucose load followed by a 3-hour 100-gram glucosetolerance test (a two-step approach) for those foundto be abnormal on the screen. This approach identifiesapproximately 5% to 6% of the population as havingGDM. The diagnostic threshold criteria for this test were
originally predicated not on perinatal outcomes, but onthe likelihood that a woman would develop diabetesmellitus in future years. More recently, evidence hasaccumulated that GDM identified by this system isassociated with an increased risk of adverse maternaland perinatal outcomes.
In contrast, newly proposed diagnostic strategies rely on
the administration of a 2-hour glucose tolerance test (aone-step approach). Each of these strategies is basedon a one-step approach with a fasting component, a75-gram glucose load, and 2 hours of testing. However,these tests differ on whether a 1-hour sample is included,whether two abnormal values are required, and thediagnostic cutoffs that are used. Most recently, theIADPSG has proposed diagnostic thresholds based on
demonstrated associations between glycemic levels andan increased risk of obstetric and perinatal morbidities.
The panel considered whether a one-step approach tothe diagnosis of GDM should be adopted in place of thetwo-step approach. The one-step approach offers certainoperational advantages. The current two-step approachis used only during pregnancy and is largely restricted tothe United States. There would be value in a consistent,international diagnostic standard across ones lifespan.This unification would allow better standardization of best
8/10/2019 Gestational Diabetes Mellitus508
23/36
19
practices in patient care and comparability of researchoutcomes. The one-step approach also holds potentialadvantages for women and their healthcare providers,
as it would allow a diagnosis to be achieved within thecontext of one visit as opposed to two.
To determine whether the advantages of the one-stepapproach should lead to its adoption, several criterianeed to be fulfilled:
There should be evidence that the additional women,who are identified by the one-step approach, have
an increased frequency of maternal morbidities,perinatal morbidities, or both.
There should be evidence that these morbidities,particularly among the new subset of women whoare identified by the one-step approach but whowould have been considered normal using thecurrent two-step process, can be decreased
by intervention. There should be evidence that the benefits of the
decrease in morbidities outweigh the potentialharms (maternal, perinatal, and societal).
While there is good evidence that increasing glycemiclevels during pregnancy is associated with more maternaland perinatal morbidities, there is no single cutoff below
which these associations are absent. These associationshave been best demonstrated for the outcomes ofshoulder dystocia, cesarean delivery, macrosomia,large-for-gestational-age birth weight, neonatal adiposity,neonatal hypoglycemia, and elevated umbilical cordblood C-peptide. It is not as clear whether associationsexist for other important outcomes such as brachialplexus palsy, perinatal mortality, childhood obesity, or
subsequent maternal metabolic complications.
8/10/2019 Gestational Diabetes Mellitus508
24/36
20
There also is evidence that treatment of women withGDM, diagnosed either by the one-step or two-stepapproach, may improve some outcomes: macrosomia,
large-for-gestational-age birth weight, shoulderdystocia, and hypertensive disease of pregnancy.Despite improvements in these intermediate outcomes,the frequencies of composite neonatal morbidity andcesarean delivery have not been consistently improvedwith treatment. Long-term outcomes for mothers andtheir offspring have not been improved in the fewstudies that have been performed.
The one-step approach, as proposed by the IADPSG,is anticipated to increase the frequency of the diagnosisof GDM by twofold to threefold, to a prevalence ofapproximately 15% to 20%. There are several concernsregarding the diagnosis of GDM in these additionalwomen. There is insufficient evidence that the resultsof the published randomized clinical trials of treatmentof mild GDM can be generalized to women who wouldbe diagnosed with GDM by the IADPSG criteria. It is notwell understood whether the additional women identifiedby this approach will benefit from treatment, and if so,to what extent. Moreover, the care of these women willgenerate additional direct and indirect healthcare costs.Such costs include increased utilization of registereddietitians and diabetes educators, prenatal care visits,
and fetal assessments (ultrasonography and prenataltesting). There is also evidence in some studies thatthe labeling of these women may have unintendedconsequences, such as an increase in cesareandelivery and more intensive newborn assessments.In addition, increased patient costs, life disruptions,and psychosocial burdens have been identified.Therefore, available studies do not provide clear
evidence that a one-step approach is cost-effectivein comparison with the current two-step approach.
8/10/2019 Gestational Diabetes Mellitus508
25/36
8/10/2019 Gestational Diabetes Mellitus508
26/36
22
Determine whether the additional women categorizedas having diabetes by the IADPSG model, who wouldbe considered normal in the two-step strategy, accrue
significant benefits from treatment. This question maybe best answered by a randomized controlled trialthat, ideally, would use clinically important healthand patient-centered outcomes.
Conduct cost-benefit, cost-effectiveness, and cost-utility analyses to more fully understand the resourceimplications of changing the thresholds for a diagnosisof GDM.
Given that the different approaches representdifferent burdens for patients, conduct research tounderstand patient preferences and the psychologicalconsequences of the diagnosis of GDM.
Perform well-conducted prospective cohort studiesof the real world impact of GDM treatment on care
utilization and practice patterns. Assess lifestyle interventions during pregnancy, such
as nutrition and exercise, that may improve maternaland fetal outcomes in women with GDM. Considerfactors, other than GDM (e.g., obesity), that contributeto fetal macrosomia in order to develop a risk modelfor macrosomia.
Assess the long-term impact that a label of GDMmay have for future pregnancy planning, futurepregnancy management, and future insurability.
Conduct further study of the long-term metabolic,cardiovascular, developmental, and epigenetic(inherited changes in phenotype [appearance]caused by mechanisms other than changes inDNA) impact on offspring whose mothers havebeen treated for GDM.
8/10/2019 Gestational Diabetes Mellitus508
27/36
23
Assess interventions to decrease the subsequentrisk of the occurrence of metabolic syndrome,diabetes, and cardiovascular disease in women
with GDM.Footnote:While we do not believe that current evidencejustifies a departure from the current two-step approachto screening and diagnosis, we do believe that a singlestandard both for screening (130 mg/dL, 135 mg/dL, or140 mg/dL for the GCT) and for the diagnostic thresholdson the GTT (Carpenter and Coustan or NDDG) should beadopted by appropriate professional organizations.
8/10/2019 Gestational Diabetes Mellitus508
28/36
Consensus Development Panel
24
James Peter VanDorsten, M.D.Panel and Conference
ChairpersonLawrence L. Hester, Jr. ProfessorDivision of Maternal-Fetal
MedicineDepartment of Obstetrics
and GynecologyMedical University of
South Carolina
Charleston, South Carolina
William C. Dodson, M.D., FACOGChiefDivision of Reproductive
Endocrinology and InfertilityMilton S. Hershey Medical CenterProfessorDepartment of Obstetrics
and Gynecology
Penn State College of MedicineHershey, Pennsylvania
Mark A. Espeland, Ph.D., FASA, FSCTProfessor of Public Health
SciencesDepartment of Biostatistical
SciencesDivision of Public Health SciencesWake Forest School of Medicine
Winston-Salem, North Carolina
William A. Grobman, M.D., M.B.A.Professor and Vice-ChairDepartment of Obstetrics
and GynecologyNorthwestern UniversityFeinberg School of MedicineChicago, Illinois
Jeanne Marie Guise, M.D., M.P.H.DirectorBuilding Interdisciplinary
Research Careers in WomensHealth (BIRCWH) andMentored Clinical ScientistsComparative EffectivenessDevelopment K12 Programs
Oregon Institute for Patient-Centered Comparative
EffectivenessCo-DirectorOregon Health & Science
University SimulationAssociate DirectorScientific Resource Center for
Evidence-based PracticeCenter and DevelopingEvidence To Inform Decisions
About Effectiveness Program
Community Practice andResearch for the OregonClinical and TranslationalResearch Institute
ProfessorDepartments of Obstetrics
and Gynecology, MedicalInformatics and ClinicalEpidemiology, and PublicHealth and Preventive Medicine
Oregon Health & ScienceUniversity Portland, Oregon
Brian M. Mercer, M.D.Professor and ChairmanReproductive BiologyCase Western Reserve
UniversityMetroHealth CampusChairmanDepartment of Obstetrics and
GynecologyMetroHealth Medical CenterCleveland, Ohio
8/10/2019 Gestational Diabetes Mellitus508
29/36
8/10/2019 Gestational Diabetes Mellitus508
30/36
26
Speakers
William H. Barth, Jr., M.D.
ChiefDivision of Maternal Fetal MedicineObstetrics and Gynecology
ServiceMassachusetts General Hospital
Associate Professor ofObstetrics, Gynecology
and Reproductive BiologyHarvard Medical SchoolBoston, Massachusetts
William M. Callaghan, M.D., M.P.H.ChiefMaternal and Infant Health BranchDivision of Reproductive HealthNational Center for Chronic
Disease Prevention andHealth Promotion
Centers for Disease Controland Prevention
Atlanta, Georgia
Brian M. Casey, M.D.Gillette ProfessorshipObstetrics and Gynecology
The University of TexasSouthwestern Medical Center
Dallas, Texas
Patrick M. Catalano, M.D.
ProfessorReproductive BiologyDirectorCenter for Reproductive HealthDepartment of Obstetrics
and GynecologyMetroHealth Medical CenterCase Western Reserve UniversityCleveland, Ohio
Aaron B. Caughey, M.D., Ph.D.,M.P.P., M.P.H.Professor and ChairDepartment of Obstetrics
and GynecologyOregon Health & Science UniversityPortland, Oregon
Ilana R. Azulay Chertok, Ph.D.,
M.S.N., R.N., IBCLCAssociate ProfessorWest Virginia University
School of NursingMorgantown, West Virginia
Donald R. Coustan, M.D.Professor of Obstetrics
and GynecologyWarren Alpert Medical School
Brown UniversityDivision of Maternal-Fetal MedicineWomen & Infants Hospital of
Rhode IslandProvidence, Rhode Island
Timothy Cundy, M.D.Professor of MedicineFaculty of Medical and
Health Sciences
The University of AucklandAucklandNew Zealand
Lois E. Donovan, M.D., FRCPCClinical Associate Professor
and Medical DirectorDiabetes in PregnancyDivision of Endocrinology and
Metabolism
Department of Obstetrics andGynecologyUniversity of Calgary
Alberta Health ServicesCalgary, AlbertaCanada
Matthew W. Gillman, M.D., S.M.DirectorObesity Prevention Program
ProfessorDepartment of Population MedicineHarvard Medical SchoolHarvard Pilgrim Health
Care InstituteBoston, Massachusetts
8/10/2019 Gestational Diabetes Mellitus508
31/36
27
Alan E. Guttmacher, M.D.
DirectorEunice Kennedy Shriver
National Institute ofChild Health andHuman Development
National Institutes of HealthBethesda, Maryland
Lisa Hartling, Ph.D.Assistant Professor
Department of PediatricsDirectorUniversity of Alberta Evidence-
based Practice CentreAlberta Research Centre for
Health EvidenceUniversity of AlbertaEdmonton, AlbertaCanada
Mark B. Landon, M.D.Richard L. Meiling Professor
and ChairDepartment of Obstetrics
and GynecologyThe Ohio State University
College of Medicine andWexner Medical Center
Columbus, Ohio
Boyd E. Metzger, M.D.Emeritus ProfessorDepartment of MedicineDivision of Endocrinology,
Metabolism andMolecular Medicine
Northwestern UniversityFeinberg School of MedicineChicago, Illinois
David M. Murray, Ph.D.
Associate Director forPrevention and
DirectorOffice of Disease PreventionNational Institutes of HealthBethesda, Maryland
Wanda Nicholson, M.D.,M.P.H., M.B.A.
Director
Diabetes and Obesity CoreCenter for Womens
Health ResearchAssociate ProfessorDepartment of Obstetrics
and GynecologyThe University of North Carolina
School of MedicineChapel Hill, North Carolina
David J. Pettitt, M.D.Senior ScientistSansum Diabetes
Research InstituteSanta Barbara, California
Edmond A. Ryan, M.D.ProfessorDepartment of MedicineDivision of Endocrinology
University of AlbertaEdmonton, AlbertaCanada
Catherine Y. Spong, M.D.Associate DirectorExtramural ProgramsEunice Kennedy Shriver
National Institute ofChild Health and
Human DevelopmentNational Institutes of HealthBethesda, Maryland
8/10/2019 Gestational Diabetes Mellitus508
32/36
Planning Committee
Catherine Y. Spong, M.D.
ChiefPregnancy and
Perinatology BranchEunice Kennedy Shriver
National Institute ofChild Health andHuman Development
National Institutes of HealthBethesda, Maryland
William H. Barth, Jr., M.D.ChiefDivision of Maternal
Fetal MedicineObstetrics and
Gynecology ServiceMassachusetts General HospitalBoston, Massachusetts
Lisa Begg, Dr.P.H., R.N.
Director of Research ProgramsOffice of Research onWomens Health
Office of the DirectorNational Institutes of HealthBethesda, Maryland
Patrick M. Catalano, M.D.ProfessorReproductive Biology
DirectorCenter for Reproductive HealthMetroHealth Medical
Center/Case WesternReserve University
Cleveland, Ohio
Christine S. Chang, M.D., M.P.H.Medical OfficerCenter for Outcomes
and EvidenceAgency for HealthcareResearch and Quality
Rockville, Maryland
28
Donald R. Coustan, M.D.
Professor of Obstetricsand Gynecology
Brown University WarrenAlpert Medical School
Women & Infants Hospitalof Rhode Island
Providence, Rhode Island
Patricia Dietz, Dr.P.H.Team Leader
Research and Evaluation TeamApplied Sciences BranchDivision of Reproductive HealthCenters for Disease Control
and PreventionAtlanta, Georgia
Judith E. Fradkin, M.D.DirectorDivision of Diabetes,
Endocrinology, andMetabolic DiseaseNational Institute of
Diabetes and Digestiveand Kidney Diseases
National Institutes of HealthBethesda, Maryland
Gilman Drew Grave, M.D.Acting Director
Endocrinology, Nutritionand Growth BranchEunice Kennedy Shriver
National Institute ofChild Health andHuman Development
National Institutes of HealthBethesda, Maryland
Planning Committee members provided their input at a meeting heldApril 35, 2011. The information provided here was accurate atthe time of that meeting.
8/10/2019 Gestational Diabetes Mellitus508
33/36
29
Stephen C. Groft, Pharm.D.
Acting DirectorOffice of Medical Applications
of ResearchOffice of the DirectorNational Institutes of HealthBethesda, Maryland
Alan E. Guttmacher, M.D.DirectorOffice of the Director
Eunice Kennedy ShriverNational Institute ofChild Health andHuman Development
National Institutes of HealthBethesda, Maryland
Suchitra Iyer, Ph.D.Health Scientist Administrator
Agency for Healthcare
Research and QualityRockville, Maryland
Mark B. Landon, M.D.Richard L. Meiling Professor
and ChairDepartment of Obstetrics
and GynecologyThe Ohio State University
College of Medicine
Wexner Medical CenterColumbus, Ohio
Deborah Langer, M.P.H.Communications AdvisorOffice of Medical Applications
of ResearchOffice of the DirectorNational Institutes of HealthBethesda, Maryland
Kelli K. Marciel, M.A.
Communications DirectorOffice of Medical Applications
of ResearchOffice of the DirectorNational Institutes of HealthBethesda, Maryland
Boyd E. Metzger, M.D.Tom D. Spies Professor of
Metabolism and Nutrition
Northwestern UniversityFeinberg School of MedicineChicago, Illinois
Elizabeth Neilson, M.S.N., M.P.H.Communications AdvisorOffice of Medical Applications
of ResearchOffice of the DirectorNational Institutes of Health
Bethesda, Maryland
Susanne Olkkola, M.Ed., M.P.A.Senior AdvisorConsensus Development ProgramOffice of Medical Applications
of ResearchOffice of the DirectorNational Institutes of HealthBethesda, Maryland
Emily Oken, M.D., M.P.H.Associate ProfessorDepartment of
Population MedicineHarvard Medical School
and Harvard PilgrimHealth Care Institute
Boston, Massachusetts
Planning Committee members provided their input at a meeting heldApril 35, 2011. The information provided here was accurate atthe time of that meeting.
8/10/2019 Gestational Diabetes Mellitus508
34/36
30
Dwight J. Rouse, M.D.Professor of Obstetrics
and GynecologyBrown University Warren Alpert
Medical School Division ofMaternal-Fetal MedicineWomen & Infants Hospital
of Rhode IslandProvidence, Rhode Island
Edmond A. Ryan, M.D.ProfessorDepartment of MedicineDivision of Endocrinology
University of AlbertaEdmonton, AlbertaCanada
Caroline Signore, M.D., M.P.H.Medical OfficerPregnancy and
Perinatology BranchEunice Kennedy Shriver
National Institute of
Child Health andHuman DevelopmentNational Institutes of HealthBethesda, Maryland
Philip Franklin Smith, M.D.Deputy DirectorDivision of Diabetes,
Endocrinology, andMetabolic Diseases
Research ProgramsNational Institute ofDiabetes and Digestiveand Kidney Diseases
National Institutes of HealthBethesda, Maryland
William N. Spellacy, M.D.ProfessorDepartment of Obstetrics
and GynecologyUniversity of South FloridaCollege of Medicine
Tampa, Florida
Xenia T. Tigno, Ph.D.Program OfficerNational Institute of
Nursing Research
National Institutes of HealthBethesda, Maryland
James Peter VanDorsten, M.D.Lawrence L. Hester, Jr. ProfessorChairmanDepartment of Obstetrics
and GynecologyMedical University of
South Carolina
Charleston, South Carolina
Paris WatsonSenior AdvisorConsensus Development ProgramOffice of Medical Applications
of ResearchOffice of the DirectorNational Institutes of HealthBethesda, Maryland
Tracy Wolff, M.D., M.P.H.Medical OfficerU.S. Preventive Services
Task Force ProgramCenter for Primary Care,
Prevention, andClinical Partnership
Agency for HealthcareResearch and Quality
Rockville, Maryland
Planning Committee members provided their input at a meeting heldApril 35, 2011. The information provided here was accurate atthe time of that meeting.
8/10/2019 Gestational Diabetes Mellitus508
35/36
Conference Sponsors
31
Eunice Kennedy ShriverNational
Institute of Child Health andHuman Development, NIH
Office of Disease Prevention, NIH
Conference Cosponsors
National Institute of Diabetesand Digestive and KidneyDiseases, NIH
National Institute of NursingResearch, NIH
Office of Research onWomens Health, NIH
Conference Partners
Centers for Disease Controland Prevention
Health Resources andServices Administration
8/10/2019 Gestational Diabetes Mellitus508
36/36
partmentof
ndHumanS
ervices
Institutesof
Health
DiseasePre
vention,OD
ecutiveBoulevard
B03MSC75
23
a,MD
20892-7523
usiness
orPrivateUse$300
FirstClassMail
Postagea
ndFeesPaid
NIH/OD
PermitNumberG-802