orals Australasian Diabetes in Pregnancy Annual Scientific Meeting 2013

PREDICTION OF POST-PARTUM ABNORMAL GLUCOSE TOLERANCE IN WOMEN WITH GESTATIONAL DIABETES MELLITUS. (#11)

Jeff R Flack 1 2 , Bin B Jalaludin 3 4 , Glynis P Ross 1
  1. Bankstown-Lidcombe Hospital, Bankstown, NSW, Australia
  2. University of NSW, Sydney, NSW, Australia
  3. Epidemiology, Healthy People and Places Unit, South Western Sydney Local Health District, Liverpool, NSW, Australia
  4. School of Public Health and Community Medicine, University of NSW, Sydney, NSW, Australia

Background: UK NICE Guidelines suggest follow-up assessment with fasting BGL (FBGL) but not a Glucose Tolerance Test (oGTT)(1) in women with gestational diabetes (GDM). We previously published data indicating this would miss a substantial number with post-partum glucose tolerance abnormalities(2). A recent UK abstract reported good sensitivity and specificity for FBGL≥4.7mmol/L and HbA1c≥5.7% cut-offs in predicting post-partum dysglycaemia(3).
Aims: To assess prediction of 6-8 weeks post-partum dysglycaemia based on findings of post-partum oGTT results and concomitant HbA1c levels, in a large GDM cohort in South-Western Sydney.
Methods: Retrospective analysis of prospectively collected data, (1993-2012), from our GDM database. Selected women had attended a post-partum oGTT with concomitant HbA1c collection. Glycaemic status was classified according to oGTT results, (not HbA1c), with IFG being FBGL≥6.1mmol/L. Receiver operating characteristic (ROC) curves of sensitivity plotted against 1-specificity were constructed for postnatal fasting glucose and HbA1c to detect post-partum dysglycaemia.
Results: There were 2024 women with available data, with oGTT (meanᆂSD) 9.7ᆂ2.6 weeks after delivery. Post-partum dysglycaemia (27.7%) was: 5.6% IFG, 15.0% IGT, 2.9% both IFG and IGT, and 4.2% Type2DM. The area under the ROC curve (AUC) for any abnormality of glucose tolerance was higher for FBGL (AUC 0.732, 95% CI 0.703–0.762) than HbA1c (AUC 0.620, 95% CI 0.591–0.649)(Fig 1). This pattern was the same amongst the four major ethnic groups represented [South-East Asian (41.3%), Middle Eastern (24.1%) European (22.8%), and Indian/Pakistani (8.2%)], although AUC was greater for FBGL for Middle Eastern women (AUC 0.809, 95% CI 0.749–0.868)(Fig 2). Despite these findings, relying on both FBGL plus HbA1c would have missed 4 of 85 women with Type2DM and 23 of 303 with IGT, and 1357/1464 with Normal Glucose Tolerance would still have required an oGTT.

1121-PredictionROCFigs1&24-7-13.jpg

Conclusions: Even using low cut-offs of FBGL≥4.7mmol/L and HbA1c≥5.7%, data from this cohort of GDM women still support the use of oGTT to detect abnormalities of glucose tolerance post-partum.

Acknowledgment: All of the Diabetes Educators who have collected data and maintained the database.

  1. Diabetes in Pregnancy CG63 Diabetes in pregnancy: management of diabetes and its complications from pre-conception to the postnatal period. NICE Clinical guidelines, CG63 - Issued: March 2008.
  2. Post-Partum Glucose Tolerance Assessment In Women Diagnosed With Gestational Diabetes: Evidence Supporting The Need To Undertake An OGTT. J. R. Flack, T. J. Payne, G. P. Ross Diabet. Med. [Letter] 27, 242–248 (2010).
  3. In women with gestational diabetes, can HbA1c or fasting glucose at 3 months postpartum be used to identify persistent dysglycaemia? KF Hunt et al. Abstract 123 Proceedings 7th International DIP Symposium on Diabetes, Hypertension, Metabolic Syndrome, and Pregnancy [DIP 2013] Meeting, Florence Italy 14-16 March 2013.