A review of ultrasound fetal weight estimation in the early prediction of low birthweight

Abstract

Estimated fetal weight is used in fetal growth monitoring and, if accurately estimated, may provide a sensitive screening tool for the small-for-gestational-age fetus, which may then proceed to further investigations. There is considerable evidence, however, that ultrasound-estimated fetal weight is inaccurate. The aim of this study was to review the literature on the efficacy of estimated fetal weight in the early prediction of low birthweight. Seven studies met the inclusion criteria. Most studies used an estimated fetal weight threshold of the 10th percentile, where sensitivity for predicting small-for-gestational-age infants in low-risk groups is low and specificity is high. The sensitivity of estimated fetal weight is higher where the prevalence of small-for-gestational-age is higher or a higher detection threshold is used. Fetal weight estimation is more sensitive and specific than other measures in detecting small-for-gestational-age, but is limited by large random errors. Random errors reduce sensitivity with less effect on specificity. High sensitivity is useful where further discriminatory tests are available; this may be the case here, where Doppler ultrasound is of proven value in high-risk groups. High specificity is required where invasive or expensive procedures will be performed on the selected group. In order to achieve sensitivity approaching 100%, a threshold 3 standard deviations of estimated fetal weight error above the 10th percentile of fetal weight is required. Smaller random errors will allow the threshold to be reduced and will increase specificity.