Savitz et al., 2002b; Yang et al., 2002c). These studies have also noted that the population mean gestational age at birth, as estimated from prenatal ultrasound, was approximately 1 week lower than the mean estimated from LMP. The use of ultrasound to estimate gestational age resulted in the birth of many fewer infants at what was considered postterm and a small increase in the numbers of infants delivered at what was considered preterm.

Despite its accuracy in estimating gestational age, the routine use of prenatal ultrasounds to estimate duration of pregnancy is limited by access to health care issues. Early prenatal ultrasounds require early prenatal care. In the United States, only 84 percent of pregnant women receive prenatal care during the first trimester, and 3.5 percent did not access prenatal care until the third trimester or had no prenatal care (CDC, 2004d). Access to prenatal care is a more serious issue for many women who have the highest risks of preterm delivery (i.e., young, poor, and immigrant women) and there are racial disparities in prenatal care (6.0 percent of non Hispanic black mothers and 5.3 percent of Hispanic mothers had no or late prenatal care, as compared with 2.1 percent of non-Hispanic white mothers (Goldenberg et al., 1992; CDC, 2004d). The United States has no national standard for the routine use of prenatal ultrasound. Although more pregnant women in the United States are receiving ultrasounds than in the past (68 percent in 2002 versus 48 percent in 1989), many may be performed too late in pregnancy or the quality of the ultrasound may not be sufficient for the accurate and reliable estimation of duration of pregnancy (CDC, 2004d).

In the absence of a known date of conception, the more liberal use of early prenatal ultrasounds enhances the best obstetric dating of a pregnancy (Neilson, 1998). Although an ultrasound in the first trimester is most accurate, as mentioned above (Kalish et al., 2004; Salvedt et al., 2004), unless an ultrasound is clinically indicated earlier in the pregnancy, the American College of Obstetricians and Gynecologists practice guidelines note that a single ultrasound at 16 to 20 weeks of gestation can also screen for fetal anomalies (ACOG, 2004). In a meta-analysis of nine trials of routine vs selective use of prenatal ultrasound examinations, routine ultrasounds were associated with earlier detection of multiple pregnancies, reduced rates of induction of labor for post-term pregnancies and increased terminations of pregnancy for fetal congenital anomalies (Neilson, 1998). There were no differences in perinatal mortality, but the studies would have to have had much larger sample sizes to be able to detect a difference.

Although it may be difficult to demonstrate other benefits, there is no doubt that better obstetric dating has clinical benefits for both the mother and the child. Better obstetric dating assists the clinician with making many important decisions, including those related to the timing and the mode of delivery, intrauterine treatments, the inhibition of labor, or the administra-