Background: Ultrasonography has been used to examine the scarred uterus in women who have had previous caesarean sections in an attempt to assess the risk of rupture of the scar during subsequent labour. The predictive value of such measurements has not been adequately assessed, however. We aimed to evaluate the usefulness of sonographic measurement of the lower uterine segment before labour in predicting the risk of intrapartum uterine rupture.
Methods: In this prospective observational study, the obstetricians were not told the ultrasonographic findings and did not use them to make decisions about type of delivery. Eligible patients were those with previous caesarean sections booked for delivery at our hospital. 642 patients underwent ultrasound examination at 36-38 weeks' gestation, and were allocated to four groups according to the thickness of the lower uterine segment. Ultrasonographic findings were compared with those of physical examination at delivery.
Findings: The overall frequency of defective scars was 4.0% (15 ruptures, 10 dehiscences). The frequency of defects rose as the thickness of the lower uterine segment decreased: there were no defects among 278 women with measurements greater than 4.5 mm, three (2%) among 177 women with values of 3.6-4.5 mm, 14 (10%) among 136 women with values of 2.6-3.5 mm, and eight (16%) among 51 women with values of 1.6-2.5 mm. With a cut-off value of 3.5 mm, the sensitivity of ultrasonographic measurement was 88.0%, the specificity 73.2%, positive predictive value 11.8%, and negative predictive value 99.3%.
Interpretation: Our results show that the risk of a defective scar is directly related to the degree of thinning of the lower uterine segment at around 37 weeks of pregnancy. The high negative predictive value of the method may encourage obstetricians in hospitals where routine repeat elective caesarean is the norm to offer a trial of labour to patients with a thickness value of 3.5 mm or greater.