A New Sonographic Weight Estimation Formula for Fetuses with Congenital Diaphragmatic Hernia

Ultraschall Med. 2015 Jun;36(3):284-9. doi: 10.1055/s-0034-1366517. Epub 2014 Jun 13.

Abstract

Purpose: The accuracy of the sonographic weight estimation (WE) of fetuses with congenital diaphragmatic hernia (CDH) is significantly lower than that of fetuses without any malformations. The objective of this study was to develop and evaluate the first specific sonographic weight formula for fetuses with CDH.

Materials and methods: In a retrospective, multicenter, cohort study, a statistical estimation technique known as "multivariable fractional polynomial regression" was applied to a group of 146 fetuses with CDH. Each fetus underwent an ultrasound examination with complete biometric parameters within 7 days of delivery. A new formula was derived using the obtained data and was then compared with other commonly used equations. The accuracy of the different formulas was compared using means of signed percentage errors (SPE), medians of absolute percentage errors (MAPE), and fractions of estimates within prespecified error bounds.

Results: The new derived formula is: EFW = 10^(4.6729 107 371 + 0.2365 011 768 * HC + 0.2228 897 682 * FL^2 - 0.0129 895 773 * FL^3 - 1.0470 039 072 * (FL * HC)^0.5 + 0.0004 314 661 * (AC * HC) - [in case of liver herniation] 0.0062 112 122), where EFW is the estimated fetal weight, HC is the head circumference, AC is the abdominal circumference, and FL is the femur length. The new formula proved to be superior to other established equations, showing both the lowest median absolute percentage error (MAE: 6.97) and mean signed percentage error (SPE: 0.40), and the best distribution of absolute percentage errors within prespecified error bounds.

Conclusion: This new formula significantly improves weight estimation in fetuses with CDH.

Publication types

  • Multicenter Study

MeSH terms

  • Cohort Studies
  • Female
  • Fetal Weight*
  • Hernias, Diaphragmatic, Congenital / diagnostic imaging*
  • Humans
  • Infant, Newborn
  • Male
  • Models, Statistical
  • Pregnancy
  • Regression Analysis
  • Retrospective Studies
  • Ultrasonography, Prenatal / methods*