Advertisement

195: Accuracy of growth restriction diagnosis_implications for testing and intervention

      Objective

      The diagnosis of small for gestational age (SGA) fetal growth impacts antenatal management and decision-making. Fetuses with SGA growth undergo heightened surveillance with antenatal testing as well as planned deliveries between 37-39 weeks to avert the risk for stillbirth. The goal of this study was to examine accuracy of antenatal diagnosis of SGA fetal growth by ultrasound (US) compared to actual birth weight in term fetuses.

      Study Design

      This is a prospective cohort study of all consecutive singleton term deliveries from a single institution from 2010-2014. Estimated fetal weight (EFW) by US obtained within 4 weeks of delivery was classified into <10%ile, <5%ile, and <3%ile by Hadlock 4. Sensitivity, specificity, positive and negative predictive values for the performance of antenatal US classification as <10th, 5th, or 3rd percentiles were compared to the same classification using actual birth weight.

      Results

      A total of 1,733 fetuses had US EFW available within 4 weeks of delivery. Sensitivity, specificity, and predictive values for each category of antenatal SGA are presented in the Table. Sensitivity of US classification as <10%ile was 81% and specificity was 75%. PPV for antenatal US classification as <10%ile was only 52% corresponding to a 48% rate of misdiagnosing a fetus as SGA with birth weight that is normal (false positive). Conversely the negative predictive value of US EFW <10%ile was 92% corresponding to an 8% risk for misdiagnosing a fetus as appropriately grown antenatally with birth weight actually being less than the 10%ile (false negative). Sensitivity decreases with decreasing EFW percentiles.

      Conclusion

      Antenatal ultrasound classification of SGA <10th, <5th, and <3rd in term fetuses correctly classifies between 72-81% of fetuses with true SGA growth. The low positive predictive values among women undergoing late third trimester ultrasound suggest that US is most likely to over diagnose SGA in normal near-term and term fetuses. Increased antenatal testing and possible iatrogenic prematurity needs to be considered clinically given that almost half of the infants that carry a diagnosis of SGA at term are born AGA.