Abstract
Background. Recent technological advances have made HBA1c a more standardized and user-friendly test with wide availability. This study evaluated HBA1c as a screening test for gestational diabetes mellitus (GDM) in a high-risk population. Methods. A total of 442 pregnant women were assessed by HBA1c. Two thresholds were used to 'rule in or rule out' GDM, which was confirmed by the 'gold-standard' 75-g OGTT (World Health Organization criteria). Results. Eighty-four (19%) women had GDM while 358 (81%) women were normal; there was a complete overlap of the HBA1c histograms of the two populations. The area under receiver operating characteristic curve of HBA1c to detect GDM was 0·54 (95% CI 0·48-0·61). Using a value of <5·5% to rule out GDM; a sensitivity of 82·1% was achieved, with 15 (16·7%) of 90 women below the threshold being false-negatives (NPV = 83·3%). Using a threshold of HBA1c ≥ 7·5% to rule-in GDM; the specificity was 95·8% with 15 (71·4%) of 21 patients over the threshold being false-positives (PPV = 28·6%). HBA1c would eliminate the need for an OGTT in 25·1% (111/442) of whom 27% (30/111) women would be misclassified. At any HBA1c threshold with an acceptable sensitivity, the false-positive rate remained high making it necessary for too many healthy women to undergo the confirmatory OGTT. Conclusion. Despite all the progress in methodology, HBA1c remains a poor test to screen for GDM.