1041: The association between provider and glycemic threshold for pharmacotherapy initiation for treatment of Gestational Diabetes

Abstract

Complete outcome data were available on 3831 (92.4%) women. A positive screen detected 61.8% of all FGR (n¼396 (10.3%); +LR 2.76,-LR 0.49). Test performance was not significantly different without placental biometry (+LR 3.35,-LR 0.54). Detection of preterm FGR (n¼98, 2.6%) was excellent (sens. 90.8% & NPV 99.7%), but was also not significantly improved by inclusion of placental biometry (89.8% & 99.7%). Inclusion of placental measurements as continuous data in a logistic regression model yielded minor improvements in screening performance, but did not increase the PPV over the predefined UtAD/EFW thresholds (10.7 vs 11.3%). Sensitivity analyses by FGR risk factor did not identify any significant differences in the test performance (sens. 88-92%, PPV 9.2-12.4%). CONCLUSION: This is largest study to date of 2D placental biometry measurement for the prediction of FGR. It demonstrates that placental assessment marginally improves predictive ability, but that the majority of pregnancies at risk can be identified by EFW/UtAD measurement.