Prediction of Spontaneous Preterm Birth by Cervical Elastogram

Background: Preterm birth is the leading cause of perinatal morbidity and mortality in multiple pregnancies. The medical and educational expenditure and lost productivity associated with preterm birth is very high. Sonoelastography is a new ultrasound technology which has enabled the measurement of tissue stiffness and it has been widely applied in assessment of breast lump and liver cirrhosis. Sonoelastography is potentially useful for the objective assessment of cervical consistency which could be related to risk of preterm birth. Objective: To assess the cervical consistency by shear wave elastography in prediction of preterm birth in multiple pregnancies. Setting: This is a prospective non-interventional observational study. Subjects: Chinese women carrying multiple pregnancies attending antenatal visit are invited to join the study. Method: Demographic data and antenatal history will be obtained. Transvaginal scan for the cervix will be performed by a trained researcher or obstetrician for the assessment of cervical consistency, cervical length, posterior cervical angle and fetal viability. The measurements will be repeated during 5 antenatal visits or separate study visits at 11-15, 16-19, 20-23, 24-27 and 28-32 weeks of gestation. At the end of pregnancy, the delivery data and neonatal outcome will be collected. Sample Size: The risk of preterm birth \<34 weeks in twin pregnancies is approximately 20%. A sample size of 120 was determined as being sufficient to test the diagnostic performance of cervical elasticity in mid trimester assuming that the risk of preterm delivery \<34 weeks vs \>34 weeks is 1:5 and that cervical elasticity area under curve (AUC) of the receiver operating characteristics (ROC) curve was 0.7, with 80% power and alpha of 5%. Data Analysis: Changes in the cervical elastography will be compared using the Student paired two-tailed t test, and in the case of unequal standard deviation (SD), using a Wilcoxon signed-rank paired two-tailed test. For comparison of frequencies, Chi-square test, or a two-sided Fisher exact test will be used where appropriate. Linear regression will be used to assess the correlations. The level of statistical significance is set at p\<0.05 (two-sided).