The objective of this randomized clinical trial is to evaluate the clinical efficacy of routine prenatal ultrasound examination. The primary hypothesis to be tested is that routine prenatal ultrasound will result in decreased intervention in pregnancies for postdate induction. The secondary hypotheses are that routine ultrasound will result in: a) an improvement in the accuracy of the prenatal diagnosis of intrauterine growth retardation, b) a greater proportion of twin gestations diagnosed before 26 weeks gestation. Pregnant patients who have no defined indication for an ultrasound scan will be randomly assigned at the first prenatal visit to receive either a routine prenatal ultrasound examination between ten and eighteen weeks gestation, or to receive usual prenatal care, including ultrasound examination if indications develop. Patients will be enrolled from the practices of family physicians and obstetricians in rural and urban, university and private practice settings. Balanced block prospective stratification will occur by provider site to help control for potential biases involved in using multiple sites. All ultrasounds will be performed by qualified ultrasonographers under the supervision of maternal fetal specialists who are expert in obstetrical ultrasonography. Major endpoints will include the incidence of interventions for postdate pregnancy: antepartum testing, induction and caesarean section. In addition, sensitivity and specificity of the diagnosis of intrauterine growth retardation and gestational age at diagnosis of twin gestation will be examined. A pilot trial in which 536 patients have been enrolled has demonstrated the feasibility of a large trial at this site. A sample size of 4,183 patients is estimated to be adequate to test the primary hypothesis. The study sample is being generated from a variety of practice sites. The system by which ultrasound examination is performed by technicians under the supervision of maternal fetal specialists has practical applicability. The results of this trial should have generalizibility to the pregnant population of the United States.
