Most ultrasound estimated fetal weight (EFW) formulas incorporate abdominal circumference, which may overstimate growth restriction in fetal gastroschisis. The aim of this study was to determine the optimal ultrasound formula for prediction of birthweight and fetal growth restriction (FGR) in gastroschisis.
We conducted a retrospective cohort analysis of singleton fetuses with gastroschisis. Percentage of error between ultrasound EFW (performed within 2 weeks of delivery) and birthweight was calculated. Agreement between EFW by ultrasound formulas and birthweight was determined by Bland-Altman limits of agreement; concordance between ultrasound and birthweight diagnosis of FGR was evaluated with McNemar’s test.
Birthweight was best predicted by the formulas of Shepard et al and Siemer et al. Only these formulas demonstrated significant agreement with birthweight for prediction of FGR at the 5th and 10th percentiles.
The formulas of Shepard et al and Siemer et al best estimate birthweight, and their use has the potential to reduce rates of overdiagnosis of FGR.
Gastroschisis is a congenital abdominal wall defect that occurs in 1-5 per 10,000 live births, with an increasing incidence over the past 2 decades. The major predictors of morbidity for gastroschisis include prematurity, fetal growth restriction (FGR), and bowel complications. Traditionally, high rates of FGR are noted with gastroschisis. Because FGR and gastroschisis are both risk factors for perinatal morbidity that includes intrauterine fetal death, elective preterm delivery is often undertaken based on ultrasound prediction of FGR. Consequently, accurate ultrasonographic estimation of fetal weight and prenatal detection of FGR for fetuses with gastroschisis is an essential tool in obstetric management.
Ultrasound estimation of birthweight is most predictive of actual fetal weight with the use of multiple biometric parameters, which include the abdominal circumference (AC). Currently, there are a number of formulas that are available for the prediction of birthweight, including the formulas of Hadlock et al (formulas 1 and 2), Siemer et al, Warsof et al, and Shepard et al. The commonly used formula of Hadlock et al for ultrasound estimation of fetal weight was derived from structurally normal fetuses in whom the AC resembled a circle in the correct ultrasound plane. The extrusion of abdominal contents in gastroschisis results in a small AC, which leads to the underestimation of estimated fetal weight (EFW) and an increased rate of diagnosis of FGR, which potentially contributes to iatrogenic prematurity. Recently, Siemer et al proposed a novel formula for the prediction of birthweight without the use of the AC in fetal abdominal wall defects. Our objective was to identify the optimal EFW formula for ultrasound prediction of birthweight and FGR in fetal gastroschisis.
Materials and Methods
A retrospective cohort analysis of all singleton deliveries that were complicated by fetal gastroschisis at the University of North Carolina (UNC) Hospitals between January 2000 and June 2009 was completed. We included all singletons delivered at >30 weeks’ gestation and limited our analysis to cases with an ultrasound examination within 2 weeks of delivery. Gestational age >30 weeks was selected in this analysis because the goal of this study was to determine the optimal ultrasound formula for the prediction of EFW; selection of as robust as possible population (later gestational age with delivery within 2 weeks) was appropriate to address the specific aims of the study. The study was approved by the UNC School of Medicine Institutional Review Board.
All ultrasound examinations were performed in the UNC Prenatal Diagnosis Center using ATL Ultramark 5000 (National Ultrasound, Duluth, GA) and GE Voluson Expert or GE Voluson E8 (GE Healthcare, Milwaukee, WI) with 4- to 8-mHz transabdominal probes. Images and reports were stored in the R4 ultrasound reporting system (R4 Acert version 4.10; R4, LLC, Strongsville, OH).
Ultrasound reports were abstracted for fetal biometry that included biparietal diameter, head circumference, AC, and femur length. In our ultrasound unit, AC is performed by measurement of transverse abdominal diameter and anterior-posterior diameter of the abdomen at the level of the liver and portal veins; the AC is calculated by the ultrasound software. The gastroschisis was not included in the measurement. All measurements were the original values that were obtained at the time of the procedure; no repeat measurements were preformed for this study. Ultrasound EFW was calculated post hoc for the last ultrasound evaluation before delivery with the equations of Hadlock et al, Siemer et al, Warsof et al, and Shepard et al ( Table 1 ). The percentage of error between ultrasound EFW and birthweight ([EFW – birthweight]/birthweight × 100) was calculated for each EFW formula. The median and interquartile ranges of the percentage of error for each formula were compared with those that were derived by the formula of Siemer et al with the use of the Wilcoxon signed rank test.
|Hadlock et al (1)||Log 10 EFW = 1.3596 + 0.0064(HC) + 0.0424(AC) + 0.174 (FL) + 0.00061(BPD × AC) – 0.00386 (AC × FL)|
|Hadlock et al (2)||Log 10 EFW = 1.335 – 0.0034 (AC × FL) + 0.0316(BPD) + 0.0457(AC) + 0.1623(FL)|
|Warsof et al||Log 10 EFW = 1.599 + 0.144(BPD) + 0.032(AC) – 0.000111(BPD) (AC)|
|Shepard et al||Log 10 EFW = –1.7492 + 0.166(BPD) + 0.046(AC) – 0.002546(AC)|
|Siemer et al||EFW = −145.577 + 23.724 × FL 2 + 1.255 × BPD 3 + 0.001 × e OFD − 0.0000406 × 10 FL + 1.03 × e FL|
The Bland-Altman analysis was used to assess agreement among analytic methods of measurement by calculating a mean difference between the ultrasound EFW by each formula and birthweight and to demonstrate the magnitude and direction (over- or underestimation) of the calculated value (mean gram estimate) and the actual birthweight. Bland-Altman is an accepted method for the determination of the 95% limits of agreement, which represent the expected difference between calculated and measured birthweight. An a priori determination of mean difference of >200 g between ultrasound EFW and birthweight as being clinically significant was used to determine agreement.
Ultrasound EFW formulas were tested for the ability to predict FGR as defined by birthweight of <5th and <10th percentile for gestational age. McNemar’s test was used to determine agreement between the diagnosis of FGR at time of ultrasound scan and birthweight measure ( P < .05 demonstrates lack of agreement). Overall test accuracy for detection of birthweight of <10th percentile was calculated for each formula by the method
accuracy = ( sensitivity ) ( prevalence ) + ( specificity ) ( 1 − prevalence ) ,