O N L I N E L E T T E R S

women having “shared care” between their can be first detected or predicted by testing
OBSERVATIONS family doctors and the antenatal clinic, who during pregnancy. Given these two major
would subsequently deliver in the public considerations, it is reasonable to consider
hospital, and approximately one-half were determining the percentage of women to be
Diagnosis women who would be delivered in a pri- diagnosed with GDM and adjusting the
of Gestational vate hospital and attending one of several diagnostic criteria accordingly.
private obstetricians. Women attending for
Diabetes Mellitus: A private pathology were older than women
Different Paradigm attending the public hospital. NADIA MANZOOR, MD
ROBERT G. MOSES, MD
to Consider With the IADPSG criteria compared
with the ADIPS criteria, the prevalence From the Endocrine Registrar, Illawarra Shoalhaven
of GDM increased from 8.6 to 9.1% for Local Health District, West Wollongong, New

T
antenatal clinic patients, from 10.5 to 16.2% South Wales, Australia.
he recently completed National In- for private patients, and overall from 9.6 Corresponding author: Robert G. Moses, robert.
stitutes of Health (NIH) (1) consen- to 13.0%. This overall prevalence was sim-
moses@sesiahs.health.nsw.gov.au.
sus development conference on the DOI: 10.2337/dc13-1433
ilar to the 12.1 and 13.0% from a post hoc © 2013 by the American Diabetes Association.
diagnosis of gestational diabetes mellitus analysis of two Australian sites participat- Readers may use this article as long as the work is
(GDM) found insufficient evidence to ing in the HAPO study. properly cited, the use is educational and not for
change from current practice to that rec- With an OR of 2.0, the glucose profit, and the work is not altered. See http://
ommended by the American Diabetes As- creativecommons.org/licenses/by-nc-nd/3.0/ for
tolerance test diagnostic values were fast- details.
sociation (2). The NIH final statement ing $5.3 mmol/L, 1 h $10.6 mmol/L,
identified some key research gaps, the and 2 h $9.0 mmol/L (5). The prevalence
first mentioned of which was to “evaluate of GDM with these criteria was 5.6% for
the diagnostic thresholds associated with Acknowledgments—No potential conflicts of
antenatal clinic patients, 8.4% for private interest relevant to this article were reported.
an odds ratio (OR) for adverse outcomes patients, and 7.1% overall. Whereas with N.M. researched the data. N.M. and R.G.M.
of 2.0 in the [Hyperglycemia and Adverse an OR of 1.75, in our predominantly Cau- wrote the manuscript. R.G.M. is the guarantor
Pregnancy Outcomes] HAPO study (as casian population, 57% of women would of this work and, as such, had full access to all
opposed to the OR of 1.75 that is have been diagnosed based on the fasting the data in the study and takes responsibility
currently recommended by the Interna- glucose alone, this reduced to 33.7% with for the integrity of the data and the accuracy of
tional Association of Diabetes in Preg- an OR of 2.0. the data analysis.
nancy Study Groups [IADPSG]).” We The strengths of the original study
currently have data that may assist in were that it was prospective and included c c c c c c c c c c c c c c c c c c c c c c c c
this consideration. women attending both the public hospital References
The results herein reported are from a and the offices of private obstetricians. 1. National Institutes of Health Consensus
post hoc analysis of a prospective study Women attending privately were older Development Conference. Final Statement
carried out in 2010 (3) to determine the prev- and had a much higher prevalence of GDM, on Diagnosing Gestational Diabetes Melli-
alence of GDM with the proposed IADPSG presumably because of their increased tus. Bethesda, MD, National Institutes of
(4) criteria compared with the prevalence age. Determination of the true prevalence Health, 10 May 2013
using the then existing Australian Diabetes of GDM in a community or country must 2. American Diabetes Association. Standards
in Pregnancy Society (ADIPS) criteria. consider all of the health care delivery of medical care in diabetes—2011. Diabetes
Briefly, this study was carried out in a options. Using an OR of 2.0 will reduce Care 2011;34(Suppl. 1):S11–S61
city where all pregnant women are re- the prevalence of GDM, compared with 3. Moses RG, Morris GJ, Petocz P, San Gil F,
quested to have a diagnostic 75-g glucose using an OR of 1.75, and shift the di- Garg D. The impact of potential new diag-
tolerance test, in this case with samples agnostic emphasis away from the fasting nostic criteria on the prevalence of gesta-
fasting and at 1 and 2 h, at the end of the tional diabetes mellitus in Australia. Med
glucose value. In our community, use of an
J Aust 2011;194:338–340
second trimester. There is a high compli- OR of 2.0 would give the same approxi- 4. Metzger BE, Gabbe SG, Persson B, et al.;
ance with this request. Samples were mate prevalence of GDM as would the International Association of Diabetes &
collected for the first 6 months of 2010 older ADIPS criteria. Pregnancy Study Groups (IADPSG) Con-
from women attending the antenatal clinics There is a continuum of risk for sensus Panel Writing Group and the Hyper-
at the only public obstetric hospital (n 5 increasing glucose levels in pregnancy glycemia & Adverse Pregnancy Outcome
571) and from women attending a major and a variety of adverse pregnancy out- (HAPO) Study Steering Committee. The di-
private pathology provider (n 5 704). Ap- comes. There is also an increasing prev- agnosis of gestational diabetes mellitus: new
proximately one-half of the women attend- alence of diabetes and lesser degrees of paradigms or status quo? J Matern Fetal Neo-
ing the private pathology provider were glucose intolerance in the community that natal Med 2012;25:2564–2569

care.diabetesjournals.org DIABETES CARE, VOLUME 36, NOVEMBER 2013 e187