14 th World Congress on Ultrasound in Obstetrics and Gynecology Oral communication abstracts OC014 Clinical significance of sonographic findings in placenta previa with prior Cesarean section J. I. Yang 1 , Y. K. Lim 1 , S. J. Chang 1 , H. S. Kim 1 , M. R. Kim 1 , J. P. Lee 1 , K. J. Hwang 1 , K. H. Chang 1 , H. S. Ryu 1 , K. S. Oh 1 , S. S. Ju 2 1 Ajou University School of Medicine, Republic of Korea, 2 College of Medicine Pochon CHA University Bundang, Republic of Korea Objective: To evaluate the efficacy of intraplacental lacunae by transvaginal sonography in diagnosing the placenta accreta and increta and the prediction of clinical outcome. Methods: Forty-seven patients delivered in our hospital with history of cesarean section, diagnosis of placenta previa totalis by transvaginal sonography, and postpartum placental pathology were included in the study. Intraplacental lacunae were classified into 4 different grades from 0 to 3 by the degree (HJ Finberg, 1992). Placental pathology was done to all the patients delivered, and in case of hysterectomy, pathologic examination of removed uterus was also performed.Then they were compared with antepartum sonographic findings. Results: The lacunae was found in 8 cases with grade 1, 11 cases with grade 2, 5 cases with grade 3, and they were not showed in 23 cases. Sensitivity, specificity, positive predictive value, and negative predictive value of diagnosing placenta accreta above grade 1 lacunae were 79.2%, 87.0%, 86.4%, and 80%, respectively. When the pathologic diagnosis of placenta and removed uterus were placenta increta or percreta, all sonographic findings were above grade 2 lacunae. Sensitivity, specificity, positive predictive value, and negative predictive value of diagnosing placenta increta with lacunae above grade 2 were 93.8%, 100%, 100%, and 96.9%. Hysterectomy was performed to 17 cases, among them, 1 case showed grade 1 lacuna, 11 cases showed grade 2 lacunae, and 5 cases showed grade 3 lacunae. Without lacunae, no hysterectomy was done. Comparing the absence of lacunae or grade 1 lacunae, the relative risks of cesarean hysterectomy was 17 folds in the above grade 2 lacunae. Conclusions: Lacunar findings of transvaginal ultrasonography in patients with placenta previa totalis and history of cesarean section are very usefulness for diagnosis of placenta accreta and increta. These might be used as a valuable clinical prognostic marker. OC015 Head engagement can be determined by translabial ultrasound and is predictive of normal vaginal delivery H. P. Dietz , V. Lanzarone Royal Prince Alfred Hospital, Sydney, Australia Objective: Engagement of the fetal head at term has long been used as a predictor of labour outcome. Surprisingly, no attempts seem to have been made to define engagement with modern imaging techniques. A recent pilot study suggested that head engagement can be determined by translabial ultrasound and is strongly associated with delivery mode. In this paper the authors tested parameters of head engagement for their predictive value. Methods: 202 nulliparous women in their first ongoing pregnancy were seen between 36 and 40 weeks’ gestation in a prospective observational study. The assessment included an interview, abdominal palpation and (in a subset of 154 women) vaginal examination for a Bishop Score, as well as translabial ultrasound. Antenatal and delivery information was obtained from patient notes and the institutional obstetric database. Head engagement was determined in the midsagittal translabial plane, supine and after voiding, with a line vertical to the central axis of the symphysis pubis used as plane of reference. Results: Of 202 women, 1 was excluded from analysis as she was found to have a previously undiagnosed breech presentation. The average age was 30, the average gestational age was 38 weeks. Head engagement was determined abdominally in 200 women, vaginally in 154 and sonographically in 201 women. The three methods were strongly intercorrelated (all p < 0.001). All three methods were associated with delivery mode, with the sonographic method showing by far the strongest association (p < 0.001 with both normal vaginal delivery and vaginal delivery. Conclusions: Head engagement in the late third trimester is a predictor of vaginal delivery in nulliparous women. It can be assessed by translabial ultrasound, and the imaging assessment clearly is more predictive than abdominal or vaginal palpation of the fetal head. Used in conjunction with other parameters, it may allow construction of a predictive model for use in a clinical intervention trial. OC016 Predicting delivery mode H. P. Dietz , V. Lanzarone Royal Prince Alfred Hospital, Sydney, Australia Objective: Emergency operative delivery is often a traumatic experience for the mother and associated with excess morbidity, resource use and medicolegal risk. Prediction would therefore hold considerable promise. In this study we created a predictive model from known and newly described predictors of vaginal delivery and tested this model for its predictive value. Methods: 202 nulliparous women were seen between 36 and 40 weeks’ gestation in a prospective observational study. A subset of 126 women had an assessment including an interview, vaginal examination for a Bishop Score, as well as translabial ultrasound. The latter was performed supine and after voiding to determine cervical length, pelvic organ descent on Valsalva and head engagement. Results: Maternal age, current body mass index, Bishop score and a history of C/S in mother or sisters were predictive of vaginal delivery on univariate analysis, as were the following measurements obtained on translabial ultrasound: position of the bladder neck on Valsalva, cervical length and head engagement relative to the symphysis pubis. Head engagement and age were the strongest single predictors. The best regression model for predicting vaginal delivery contained all of the above except cervical length on US and yielded a corrected Nagelkerke’s R2 of 26.8%. The likelihood of caesarean section in the deciles of predictive scoring was as follows (from the lowest decile to the highest): 8%, 0%, 0%, 8%, 0%, 16%, 42%, 33%, 42%, 85%. Conclusions: Prediction of delivery mode in nulliparae is feasible and moderately powerful. A combination of clinical and ultrasound parameters yielded a predictive model that may become useful as an entry criterion for intervention trials. Intervention may become possible both at the low- risk (e.g. selecting women for low- tech birthing units) and the high- risk end of the spectrum- eg. offering elective caesarean section to women with a high likelihood of emergency operative delivery. OC017 Prostaglandin-induced cervical ripening monitored by ultrasound Z. T ´ oth , O. T ¨ or ¨ ok, A. Jakab, T. Kov´ acs University of Debrecen, Hungary Objective: To determine the changes in the uterine cervix visible at transvaginal ultrasound examination during prostaglandin- induced cervical ripening. Methods: Length and width of the cervix and the opening of the inner cervical os were measured. Percentage funneling and cervical index were calculated before dinoproston gel instillation for cervical ripening in 45 women with singleton pregnancies at 37–42 weeks 220 Ultrasound in Obstetrics & Gynecology 2004; 24: 217–268