Evaluation of fasting blood glucose as an initial diagnostic test for gestational diabetes mellitus (GDM). (#135)
Aim: Can fasting blood glucose (FBG) level accurately identified patients at low risk of elevated 1-hr and 2-hr (post-challenge) BGLs on an oral glucose tolerance test (OGTT), negating the need for an OGTT.
Methods: We analysed the OGTTs of 68,170 pregnant women tested between 1 July 2013 to 30 June 2015 as well as determining the rate of screening 50 gm glucose challenge tests (GCTs). We defined a subgroup of 26,219 having OGTT in the final 6 months with a low rate of prior GCT. We analysed the relationship between FBG and post-challenge BGLs. We performed a receiver operating characteristic (ROC) analysis to best determine the FBG level at which GDM was unlikely. We used published data to determine the excess cases of macrosomia potentially missed. Medical benefits schedule was used to assess costs.
Results: In our cohort, 15.6% of women had GDM, 8.6% having an elevated FBG>5.0mmol/l. At FBG=5.1mmol/l, the mean 1-hr BGL=8.3mmol/l and the 2-hr BGL=6.7mmol/l, suggesting that a lower FBG may identify women with high probability of having a normal OGTT. The ROC analysis identifed an FBG>4.6mmol/l having a specificity of 55% and sensitivity of 71% for diagnosis. At FBG=4.6mmol/l, 8.6% of women had elevated post-challenge BGLs. Women having an FBG<4.7mmol/l comprised 74% of cohort with cumulative rate of GDM in these women being only 5.5%. Based on the HAPO data, the number of excess cases of macrosomia missed was calculated to be only 0.24% (CI 0.17-0.31).
Conclusion: For every 100 women screened using FBG, 74 do not need an OGTT having a FBG<4.7mmol/l and 8.6 are diagnosed with GDM having FBG>5.0mmol/l. Only 18 women need an OGTT having FBG between 4.7-5.0mmol/l. Less than 1 extra case of macrosomia is missed. These data suggest that FBG can be used to cost-effectively and simply screen for GDM