OBJECTIVE: Our purpose was to assess the utility of ultrasonographic measurement of cervical length for predicting preterm delivery in pregnant women with twin pregnancy and preterm labor. PATIENTS AND METHOD: Prospective study in 192 singleton and 66 twin pregnancies with preterm labor before 34 weeks. Vaginal sonography for measurement of cervical length, and screening for fetal fibronectin in vaginal smears were performed at admission. The incidence of delivery before 34 weeks was compared with cervical length and the presence of fetal fibronectin. RESULTS: In singleton pregnancies, delivery before 34 weeks was significantly more frequent in women with a cervical length <30 mm (26.6 vs. 2.0%; P = 0.0004), or presence of fetal fibronectin (27.5 vs. 6.0%; P = 0.001). The corresponding negative predictive values (NPV) were 98% and 94%, respectively. In twin pregnancies, receiver characteristic curve analysis showed that a cervical length of 20 mm had the highest diagnostic performance in predicting preterm delivery. The corresponding NPV was 63%. In this group, delivery before 34 weeks showed no significant difference in women with a cervical length < 30 mm (43.7 vs. 38.8%), cervical length < 20 mm (52.0 vs. 36.6%, P = 0.219), or presence of fetal fibronectin (47.6 vs. 21.7%, P = 0.0705). DISCUSSION AND CONCLUSION: Ultrasonographic measurement of cervical length does not predict preterm delivery in twin pregnancies with preterm labor.
OBJECTIVE: Our purpose was to assess the utility of ultrasonographic measurement of cervical length for predicting preterm delivery in pregnant women with twin pregnancy and preterm labor. PATIENTS AND METHOD: Prospective study in 192 singleton and 66 twin pregnancies with preterm labor before 34 weeks. Vaginal sonography for measurement of cervical length, and screening for fetal fibronectin in vaginal smears were performed at admission. The incidence of delivery before 34 weeks was compared with cervical length and the presence of fetal fibronectin. RESULTS: In singleton pregnancies, delivery before 34 weeks was significantly more frequent in women with a cervical length <30 mm (26.6 vs. 2.0%; P = 0.0004), or presence of fetal fibronectin (27.5 vs. 6.0%; P = 0.001). The corresponding negative predictive values (NPV) were 98% and 94%, respectively. In twin pregnancies, receiver characteristic curve analysis showed that a cervical length of 20 mm had the highest diagnostic performance in predicting preterm delivery. The corresponding NPV was 63%. In this group, delivery before 34 weeks showed no significant difference in women with a cervical length < 30 mm (43.7 vs. 38.8%), cervical length < 20 mm (52.0 vs. 36.6%, P = 0.219), or presence of fetal fibronectin (47.6 vs. 21.7%, P = 0.0705). DISCUSSION AND CONCLUSION: Ultrasonographic measurement of cervical length does not predict preterm delivery in twin pregnancies with preterm labor.