Purpose: Birth weight is an important predictive parameter for neonatal morbidity and mortality in the small fetus. Accurate estimation of fetal weight is therefore a valuable tool for determining the further obstetric management. The majority of studies presenting new formulas have included relatively small samples with a narrow range of birth weights, mostly term fetuses. In a previous study, we evaluated several weight formulas over the whole range of birth weights and in defined subgroups. We were able to show that some regression formulas appeared to be favorable within these subgroups. Notably, the highest levels of inaccuracy were found in the group of infants with a birth weight of less than 2500 g. This led us to hypothesize that a formula based on the lower birth weight group might increase the accuracy of weight estimation. The aim of the present study was therefore to develop a new specific formula for estimating weight in fetuses less than or equal to 2500 g and to compare the new regression formula with commonly used weight equations.
Materials and methods: This study included 260 pregnancies. The inclusion criteria were a singleton pregnancy; birth weight equal to or less than 2500 g; an ultrasound examination with complete biometric parameters within 7 days prior to delivery; and an absence of structural or chromosomal malformations. The data for the first 130 newborns were used to develop a new formula. The remaining 130 infants were used to evaluate the new regression formula and to compare it to commonly used weight equations. Stepwise regression analysis was carried out with the birth weight as the dependent variable and biometric parameters as independent parameters to obtain the best-fit formula.
Results: The mean absolute percentage error for the new formula was 7.71 %. Compared to the other formulas, it generated the highest intraclass correlation coefficient. By the limits of agreement, the new formula demonstrated only a slight tendency towards underestimating fetal weight, and it provided the smallest range of all weight equations.
Conclusion: With proper application, our new formula can improve the accuracy of fetal weight estimation.