The prognostic value of hemoglobin A1c in predicting fetal heart disease in diabetic pregnancies

Objective: To evaluate detailed fetal echocardiography for predicting congenital heart disease in overt diabetic pregnancies. Methods: Overt diabetic gravidas with initial hemoglobin A1c (HbA1c) at or above 8.5%, a family history of congenital heart disease, or fetal cardiac anomaly suspected on screening sonography were referred to a pediatric cardiologist for detailed fetal echocardiography. After 7 years, the results of the protocol were reviewed retrospectively. The sensitivity and specificity of the initial HbA1c in predicting congenital heart disease were assessed. Results: During the study period, 193 patients received care in the diabetes and pregnancy service. Sixty-four received fetal echocardiography, 47 because of initial HbA1c levels at or above 8.5% and 17 for other indications. Twenty fetuses had major anomalies, eight of which were cardiac (40%). Six of the cardiac anomalies had been studied by fetal echocardiography, and four were diagnosed correctly. The initial HbA1c was not different among pregnancies with major anomalies, cardiac anomalies, or no anomalies. No malformations were noted in patients with normal initial HbA1c values. The overall sensitivity of the protocol for identifying congenital heart disease was 50% (four of eight) and specificity was 54% (90 of 167). Detailed fetal echocardiography had a sensitivity of 66% (four of six) and a specificity of 100% (58 of 58) for correctly identifying fetal congenital heart disease. Conclusions: No cases of congenital heart disease were observed in patients with a normal initial HbA1c value. Among patients with abnormal HbA1c values, no critical level of glycohemoglobin was identified that provided optimal predictive power for congenital heart disease screening. We recommend detailed fetal echocardiographic imaging in all patients with initial HbA1c levels above the upper limit of normal of 6.1%.