The gestational diabetes tsunami: Can we survive it?

At the time of writing, the front page banner of the Sun Herald describes the likely overwhelming of the health system (in western Sydney) by a ‘tsunami’ of diabetes. Clearly our obstetric services are well and truly in the path of this tidal wave with major workload implications. Population changes such as sedentary lifestyles, poor diets, increasing levels of obesity, increasing maternal age and migration to Australia of women of high-risk ethnicities are clearly contributing factors. Added in are the new – often contentious - approaches to the screening, diagnosis and management of gestational diabetes mellitus (GDM). This edition of ANZJOG includes several papers which illustrate a number of these issues and controversies. As background, in light of several randomised controlled trials (RCTs) demonstrating benefit from GDM management1, 2 and most notably the outcomes of the Hyperglycemia and Pregnancy Outcomes (HAPO) study,3 the International Association of Diabetes in Pregnancy Study Groups (IADPSG) established new diagnostic criteria for GDM,4 which have been endorsed by the Australasian Diabetes in Pregnancy Society (ADIPS), RANZCOG, the Royal College of Pathologists of Australia (RACP) and the World Health Organization (WHO). An increased prevalence of GDM was predicted following adoption of the new criteria. Two papers from Wollongong provide recent data on the ‘size of the problem’.5, 6 The paper by Zheng et al.6 illustrates the inadequacies of data collection, with consequent ‘major implications for healthcare planning’ and resource allocation. By cross-referencing the Midwives’ Data Collection from a private hospital with results of pregnancy oral glucose tolerance tests (OGTTs) from the local pathology provider, they found a major discrepancy in the GDM prevalence of the Data Collection (6.9%) and real GDM prevalence (~15.0%).6 Another 1.3% of the women were identified as having pre-existing diabetes. The authors refer to two other Australian studies which report similar under-reporting of the prevalence of GDM.6 In the other Wollongong paper, Moses et al.5 look closely at the total prevalence of GDM (ie inclusive of the public and private sectors), a jurisdiction they state to be very representative of the Australian population. With an estimated 87.7% Caucasians in their region, other jurisdictions may argue differently. However, using WHO terminology,7 they found on reviewing OGTT data that the 2012-2014 prevalence of hyperglycaemia in pregnancy (HIP) using the new criteria was 13.1%, with 12.7% of these having GDM and 0.4% having diabetes in pregnancy (DIP). Of further note, in those tested before 20 weeks gestation, 22% had HIP (21% GDM, 1% DIP).5 The authors argue for more local health economic data, as this high prevalence has health resource implications. They refer most notably to the St Carlos Gestational Diabetes Study,8 which showed improved pregnancy outcomes and ‘markedly reduced healthcare costs’ subsequent to the application of the new diagnostic criteria. The acceptance of the new ADIPS/RANZCOG/WHO diagnostic guidelines has been anything but unanimous, with considerable contention surrounding their application. The Endocrine Society of Australia (ESA), the Society of Obstetric Medicine of Australia and New Zealand (SOMANZ) and the Royal Australasian College of General Practitioners (RACGP) have not adopted these recommendations. In May 2015, Flack and Ross9 surveyed attitudes to these guidelines of members of the National Association of Diabetes Centres (NADC). Interestingly, abandonment of the 50 g glucose challenge test (GCT) has been universal, but 24.2% of the Centres had not adopted the new diagnostic criteria. One of their respondents’ rationale for non-adoption was ‘lack of evidence on treatment benefit of GDM diagnosed using these criteria’. A further issue of contention identified related to the recommendation for early testing for women with risk factors: various respondents use different approaches for ‘whom to test’ and for ‘which test to use’ (OGTT, fasting or random glucose, HbA1c).9 It was pointed out that early diagnosis requires increased health resources for a longer period through pregnancy. A sense of confusion, frustration and concern surrounding the lack of consensus was conveyed from the survey participants.9 A recurring theme highlighted in these papers is the workforce and economic implications of the increasing prevalence of GDM.5, 6, 9 It is worth pointing out that a number of studies do suggest that the economic outcome of application of the new criteria is a beneficial one, particularly when possible lifelong benefits for mother and infant are considered.8, 10 One of the possible methods of ‘lightening the load’ that we have implemented in the Australian Capital Territory is to stratify or ‘triage’ women to levels of risk. We have recently audited outcomes (Zhang et al., unpublished data) and the findings suggest that allocation of well-controlled GDM on diet alone to ‘usual antenatal care pathways’ (ie midwifery or general-practitioner shared care) is safe. We advocate that these low-risk GDM women should be managed within routine antenatal care. In support of stratification of GDM patients controlled on diet only to less intensive management pathways is the paper by Jeffery et al.11 They have evaluated the role of routine cardiotocography (CTG) in GDM patients. CTGs in those managed on diet alone, without a second pregnancy complication, did not change management (cost >$21 280 per DCM (definitive change in management)); for those on diet only with a second pregnancy complication, 161 CTGs were required to change management ($8063 per DCM); and for those treated with the combination of diet and medication (ie insulin and/or metformin), 43 CTGs were required to change management ($2660 per DCM). The authors concluded that routine CTGs should not be recommended for the low-risk group and should be for women with GDM requiring medication.11 For those with GDM on diet only with a secondary complication, they recommended that CTGs be performed based upon the nature of the complication only.11 It is important to point out though, that this study did not have the statistical power to ‘comment on the efficacy of CTG to prevent fetal demise’, which is ultimately the reason for such surveillance.11 The authors do not describe the actual outcomes of the DCMs which occurred. Given the known false positive rate of CTG screening, this may have been interesting information. In addition, ‘unnecessary’ intervention based on false positive antenatal testing may well have added further to costs and use of resources. Only one of the five articles on diabetes in pregnancy related to pre-existing diabetes. Padmanbhan et al.12 provide a comparison of insulin requirements of type 1 (T1DM) and type 2 (T2DM) diabetic women during the course of pregnancy in a retrospective cohort study of 222 pregnancies (67 T1DM, 155 T2DM) from two large tertiary-hospitals in western Sydney. The pattern of change in insulin dose was different between the two groups. Insulin requirements in T1DM women decreased by about 4% toward the end of the first trimester, increased by about 18% and 23% through the second and third trimesters, respectively, and fell in the last 4 weeks by about 4% again.12 In contrast, insulin requirements of T2DM women increased progressively by about 15%, 36% and 44% through the first, second and third trimesters. The percentage increase in the first trimester was greater in those T2DM women not on insulin pre-pregnancy. Surprisingly, the total daily dose expressed as units/kg was equivalent in T1DM and T2DM patients.12 The authors discuss the potential pathophysiological reasons for the findings. A major difference of course is that T1DM women do not have any significant endogenous insulin production, whereas T2DM women do. Therefore, the T1DM women have to more closely match the total insulin requirement through pregnancy with exogenous insulin, with little room for error. For the clinicians, this study confirms previous studies showing greater risk of hypoglycaemia in T1DM women from 7–13 weeks and from 36 weeks gestation, and that they should expect greater rises in insulin doses throughout pregnancy in T2DM women.12 In addition to the insulin requirement data, the greater than double prevalence of T2DM compared to T1DM, and their older age and greater body mass indices is worth noting.12 With the increasing prevalence of T2DM in pregnancy, all of us need to be aware of the imperative to provide pre-pregnancy planning to this group, as their risks for adverse pregnancy outcomes are at least as high as those for T1DM mothers. To sum up, the future tsunami of T2DM facing western Sydney is very real. For all of us involved in antenatal care, the tsunami of GDM is already here. So how do we survive it? Do we go back and change the diagnostic criteria to make our workload manageable? Or do we accept that the wave is upon us and deal with it with the potential of reducing the size of the next wave? Flack and Ross9 gave this advice: ‘It would be timely to encourage universal adoption of these criteria to remove the confusion and frustration being experienced, while at the same time trialling new models of care and collaborating with healthcare providers to ensure the workload and workforce issues are addressed’.