Objective: The long-term prediction of delivery date with obstetric ultrasound.
Method: The cervical length (CL) and the fetal vertebral, middle cerebral and umbilical resistance indices were measured in 317 pregnancies between 30 and 37 weeks. Subsequently, multivariate analysis was applied to calculate the best model for the prediction of delivery date.
Results: The best model included the examinations of the CL and the middle cerebral artery resistance index. In comparison with estimated delivery date (EDD) based on the last menstrual period, a small improvement was detected, particularly in earlier deliveries, however 34% of predictions still had an error of more than 5 days.
Conclusions: Long-term sonographic prediction of delivery date depends mainly on the CL and improves the prediction of the EDD. However, it is still too inaccurate for clinical use.