Introduction: To evaluate the influence of the time interval between examination and delivery on the accuracy of sonographic fetal weight estimation (WE).
Materials and methods: 8723 singleton pregnancies were included in this retrospective cohort study. Fetuses were divided into eight groups with regard to the time interval between estimation and delivery (group 1: 0 days; group 2: 1-3 days; group 3: 4-7 days; group 4: 8-14 days; group 5: 15-21 days; group 6: 22-28 days; group 7: 29-35 days; group 8: 36-42 days). The accuracy of WE was compared between the different time interval groups and five commonly used formulas using means of percentage errors (MPE), medians of absolute percentage errors, and proportions of estimates within 10 % of actual birth weight.
Results: In group one, the Hadlock I and Warsof formula showed a systematic underestimation of fetal weight (negative MPEs). No systematic error was found with the Hadlock II formula and the equations of Merz and Shepard showed a systematic overestimation (positive MPEs). MPE values of the Hadlock I, II and Warsof formulas were closest to zero in WEs of group two. From group three to six, MPE values decreased continuously. With the Merz and Shepard equations MPEs were closest to zero in group four.
Discussion: The best accuracy of sonographic WE with most of the commonly used equations is achieved within a scan-to-delivery interval of 1 week.
Keywords: Birth weight; Fetal weight estimation; Time interval; Ultrasound.