976 The utility of third trimester fetal biometry in predicting SGA infants in a zambian population

Abstract

Small for gestational age (SGA) is associated with adverse neonatal and childhood outcomes in low- and middle-income countries (LMIC). Higher rates of SGA have been measured among infants with fetal biometry demonstrating fetal growth restriction (FGR). The correlation between third trimester fetal biometry and SGA in LMIC is not well-studied. We sought to determine if third trimester fetal biometry is predictive of SGA in a Zambian cohort. The Zambian Preterm Birth Prevention Study (ZAPPS) is an ongoing prospective cohort of pregnant women at a single tertiary hospital in Lusaka. Women who were ≥18 years old with a viable, singleton, intrauterine pregnancy on ultrasound (US) at < 24 weeks gestation were eligible. Fetal growth centile at 32 weeks was estimated by fetal biometry using two-parameter INTERGROWTH-21st norms. FGR was defined as an estimated fetal weight (EFW) < 10%ile. The primary outcome was SGA, defined as birthweight (BW) < 10%ile. Data were analyzed by chi-square and logistic regression as appropriate. 1450 women were enrolled in ZAPPS between 2015-2017. A 32-week US (mean 32.1 ± 0.8 weeks) and infant birthweight were available for 708 mother-infant pairs (49%). 39/708 women (6%) had FGR at the 32-week US. 125/708 (18%) women delivered SGA infants. Of these infants, 20 (19%) had FGR at 32 weeks. Of the 39 women with FGR at 32 weeks, 20 (51%) had SGA infants (OR 5.7, 95% CI 2.9-11.0). A Bayesian logistic regression model was used to calculate the probability of SGA by EFW at 32 weeks (Figure 1). An EFW < 1260g by INTERGROWTH-21 at 32 weeks was associated with >50% risk of SGA. Based on ROC curve analysis (Figure 2), the specificity of using FGR at the 32-week US to predict SGA infants was very strong (97%), but sensitivity was poor (16%). Third trimester fetal biometry is a useful tool to identify pregnancies at high risk for SGA in Zambia. However, the low sensitivity of FGR as a predictor for SGA indicates other factors beyond 32 weeks may influence this outcome.View Large Image Figure ViewerDownload Hi-res image Download (PPT)