Abstract
Transvaginal sonography (TVS) of the cervix has been shown to be superior to the digital examination in detecting patients at risk for sponatenous preterm birth. The cervical length, width, length or area of the internal os (funneling) and, to some extent, the position and structure of the cervix, can be determined from ultrasound images. We have established longitudinal sonographic criteria separate for twin and singleton pregnancies in a vertical and horizontal position of the patient. Shortening of the cervical length and funneling occur earlier in an upright compared to a supine maternal position and in twin compared to singleton pregnancies. For dividing high-risk and low-risk groups, we considered the cut-off values specific for gestational age, position and singleton or twin pregnancy. Multiple logistic regression revealed that, between 20 and 24 + 6 weeks and between 25 and 29 + 6 weeks, funnel width in an upright position and the detection of fibronectin were the most significant independent variables to diagnose subsequent preterm birth. Receiver-operator curve analysis demonstrated that measurements of the cervical length and funnel width in an upright position had higher detection rates of patients at risk for spontaneous preterm birth than examinations in a supine position. There seems to be some evidence from the use of historic controls that the rate of spontaneous preterm birth may decrease with the application of early longitudinal TVS, indicating lifestyle changes, and the use of a vaginal pessary designed to support the incompetent cervix. Although TVS has its place in the follow-up of twin pregnancies with advanced dilatation and prolonged interval between deliveries, it is recommended to perform TVS early enough to recognize symptoms in a reversible instead of an irreversible phase. In the future, TVS may help to indicate and control interventions aimed at prevention or therapy of cervical incompetence.