[Predictive value of combined fibronectin and ultrasound cervical assessment in twin pregnancies]

To test a sequential test with fetal fibronectin detection after ultrasound measurement of cervical length to predict preterm delivery in twin pregnancies with preterm labor.

Descriptive retrospective study on 50 women with twin pregnancy hospitalised for preterm labor between 24 and 34 weeks and 6 days of gestation. The primary outcomes were preterm delivery before 34 or 37 weeks of gestation or within 7 or 14 days. Selective use of fibronectin after cervical length measurement has been tested, with a sequential test considered positive if cervical length was less than or equal to 15mm or if cervical length was between 16 and 30mm with fetal fibronectin positive.

The sensitivity/specificity/and positive and negative predictive values of fetal fibronectin positive were 71%, 64%, 26%, et 93% for delivery within 7 days; those of cervical length less than or equal to 20mm were 89%, 51%, 31%, et 95% for delivery before 34 weeks and 6 days. The efficiency of the sequential test seemed better than each test and than for singleton pregnancies keeping an excellent negative predictive value: sensitivity of 75%, specificity of 63%, positive predictive value of 26% and negative predictive value of 93.5% for prediction of preterm delivery within 14 days. The use of this sequential test could have decreased half of fibronectin tests.

A sequential test with selective use of fetal fibronectin detection in twin pregnancies selected by ultrasound measurement of cervical length appears to be effective for predicting preterm birth if preterm labor, avoiding half of fibronectin tests.