Performance of local reference curve on the diagnosis of large for gestational age fetuses in diabetic pregnant women

Abstract

Objective To evaluate the performance of a local fetal weight curve based on the prediction for large gestational age (LGA) newborns in diabetic pregnant women and to compare it to reference curves established for other populations. Method A reference model for estimated fetal weight was created from a local sample of 2211 singleton low-risk pregnancies. The estimated fetal weight from 194 women with gestational diabetes mellitus was then plotted on this curve, and the results were compared to those obtained by Intergrowth 21st and Hadlock curves. Results The sensitivity of the proposed model to predict LGA fetuses was 55.6%, the specificity was 82.1%, and the accuracy was 74.7%. The sensitivity, specificity, and accuracy for the Intergrowth 21st curve were 46.3%, 87.9%, and 76.3%, respectively, and no statistically significant difference was observed compared to the proposed model. Conversely, significant differences were observed for the Hadlock curve, which presented a lower sensitivity (24.1%), higher specificity (97.1%), and similar accuracy (76.8%). Conclusion The sensitivity of the proposed model was higher compared to the Hadlock curve for the screening of LGA newborns in diabetic pregnant women. However, no significant differences were observed in comparison to the Intergrowth 21st curve.