Diagnostic accuracy of glycated hemoglobin for gestational diabetes mellitus: a systematic review and meta-analysis

Background We conducted a systematic review and meta-analysis to establish the overall accuracy of glycated hemoglobin (HbA1c) in the diagnosis of gestational diabetes mellitus (GDM) diagnosis. Methods We searched MEDLINE, EMBASE, SCOPUS and ClinicalTrials.gov up to October 2018, using keywords related to GDM, HbA1c and diagnosis. Studies were included that were carried out with pregnant women without previous diabetes that assessed the performance of HbA1c (index test) compared to the 75 g oral glucose tolerance test (OGTT) (reference test) for the diagnosis of GDM, that measured HbA1c by standardized methods and presented data necessary for drawing 2 × 2 tables. Results This meta-analysis included eight studies, totaling 6406 pregnant women, of those 1044 had GDM. The diagnostic accuracy of HbA1c was reported at different thresholds ranging from 5.4% (36 mmol/mol) to 6.0% (42 mmol/mol), and the area under the curve (AUC) was 0.825 (95% confidence interval [CI] 0.751-0.899), indicating a good level of overall accuracy. The pooled sensitivities and specificities were 50.3% (95% CI 24.8%-75.7%) and 83.7% (67.5%-92.7%); 24.7% (10.3%-48.5%) and 95.5% (85.7%-98.7%); 10.8% (5.7%-19.41%) and 98.7% (96.2%-99.5%); 12.9% (5.5%-27.5%) and 98.7% (97.6%-99.3%), for the cut-offs of 5.4% (36 mmol/mol), 5.7% (39 mmol/mol), 5.8% (40 mmol/mol) and 6.0% (42 mmol/mol), respectively. Conclusions We observed a high heterogeneity among the studies. The effect of ethnicities, different criteria for OGTT interpretation and the individual performance of HbA1c methods may have contributed to this heterogeneity. The HbA1c test presents high specificity but low sensitivity regardless of the threshold used to diagnose GDM. These findings point to the usefulness of HbA1c as a rule-in test. HbA1c should be used in association with other standard diagnostic tests for GDM diagnosis.