Sonographic measurement of uterine cervix at 18-22 weeks' gestation and the risk of preterm delivery.

ABSTRACT

OBJECTIVE: Premature delivery is difficult to predict and causes considerable neonatal morbidity and mortality. Despite much research, little progress has been made in timely identification of the mothers at risk. We examined the uterine cervix with ultrasonography to discover whether such a procedure would be helpful in determining which women will deliver prematurely. METHODS: We performed transvaginal ultrasound examinations in addition to routine transabdominal ultrasonography at 18 to 22 weeks' gestation in 3694 consecutive pregnant women with live singleton fetuses. We measured the length of the uterine cervix and evaluated the dilatation, if any, of the internal os. The results of cervical ultrasonography were not available to the clinicians. RESULTS: Spontaneous delivery occurred before 37 completed weeks in 88 women (2.4%) and before 35 weeks in 31 (0.8%). The relative risk of delivery before 35 weeks was 8 (95% confidence interval 3, 19) when the cervical length was 29 mm or shorter. When dilatation of the internal cervical os of 5 mm or greater was present, the relative risk of delivery before 35 weeks was 28 (95% confidence interval 12, 67). Either short cervix (29 mm or less) or dilatation of internal cervical os (5 mm or greater) was present in 3.6% of the population; this combination had a sensitivity of 29% in predicting delivery at earlier than 35 weeks. After adjusting for cervical dilatation and length by using multiple logistic regression, nulliparity also remained a risk factor for delivery before 35 weeks (odds ratio 3.6, 95% confidence interval 1.7, 7.5). CONCLUSION: Transvaginal ultrasonography performed as an addition to routine transabdominal ultrasonography at 18 to 22 weeks helps to identify many patients at significant risk for prematurity; however, low sensitivity and low positive predictive value limit its usefulness in screening low-risk obstetric populations.