Can we improve the birth weight prediction? The effect of normal BMI using a multivariate model

Area of Science:

• Obstetrics and Gynecology
• Maternal-Fetal Medicine
• Biostatistics

Background:

• Accurate estimation of fetal weight (EFW) is crucial for optimal pregnancy management and delivery planning.
• Traditional ultrasound estimations can have significant overestimation errors, potentially leading to unnecessary interventions.
• Predictive modeling offers a potential avenue to enhance EFW accuracy.

Purpose of the Study:

• To construct and validate a predictive model to improve the accuracy of fetal weight estimation.
• To compare the predictive accuracy of the new model against standard ultrasound estimations.

Main Methods:

• A comparative, descriptive study involving 140 pregnant women categorized by pre-gestational BMI.
• Fetal weight was estimated via ultrasound at 33-35 weeks (EFW40w).
• A multivariate regression model (EFWme) was developed using newborn weight, symphysis-fundal height (SFH), EFW40w, gestational age (GA), ferritin levels, and smoking status.

Main Results:

• The multivariate model (EFWme) achieved an R2 of 0.727 (p<0.001) for estimating fetal weight.
• EFWme underestimated birth weight by only 0.07 g (0.53% mean error), while EFW40w overestimated by 300.89 g (10.12% mean error).
• Bland-Altman analysis confirmed EFWme's lower error (1.94% underestimation) compared to ultrasound (10.93% overestimation).

Conclusions:

• The developed multivariate model (EFWme) significantly enhances the accuracy of fetal weight estimation compared to standard ultrasound.
• This improved accuracy can lead to more precise clinical assessments and potentially better birth outcomes.