Salivary Estriol: A New Detector of Preterm Labor?

The rate of preterm births in the United States has changed very little in the last few decades. However, the ability to recognize maternal and fetal complications which require a preterm delivery has improved and physicians are better able to deliver the baby prior to maternal compromise or intrauterine death. In addition, neonatal intensive care units are now available to care for these sometimes very fragile neonates (newborn infants). Thus, while the preterm birth rate has remained stable, neonatal outcomes for both indicated and spontaneous preterm births have improved.

Physicians have little clinical ability to predict which pregnancy will result in preterm labor and birth. Analysis of risk factors have been advocated by some clinicians and published widely in the medical literature. Some physicians utilize the assessments by Creasy and Holbrook to identify patients which require more frequent prenatal examinations. More recently, laboratory and ultrasound testing have been proposed. Ultrasound measurement of cervical length, as well as biochemical markers for fetal fibronectin and estriol have been explored as tests that may improve our ability to predict preterm labor. Cervicovaginal fetal fibronectin was the first marketed assessment of preterm birth. It is joined now by salivary estriol as an assessment of risk for delivery. Both have excellent negative predictive value which means if the test is negative, the chance of preterm birth is very low. However, the positive predictive value with either of these tests is not as good. Less than 50% of patients with a positive test, whether at a low or high risk of premature delivery, will deliver prior to 35 weeks.

All of these tests to detect premature births increase the cost of health care in the pregnant women. Interestingly, salivary estriol has been widely advertised as a single test. However, a close look at the literature shows that serial or multiple, testing is required for accuracy. Therefore, patients may benefit from a more traditional assessment of their risk of delivering early before further testing is considered. Testing in multiple gestations is not possible at present with the salivary estriol because of the high levels associated with the larger twin placenta.

As is frequently true with highly emotional issues, practitioners and patients need to look at these new diagnostic modalities critically before instituting new and potentially expensive testing for the pregnant woman.

Date last updated:  January 6, 2003