P06.08: Prediction of emergency Caesarean sections: which variables are important?

Abstract

To explore the relationship between intrapartum variables and emergency Caesarean section (EMCS) rates. Term nulliparous women recruited in early labour, an admission scan performed to include transabdominally umbilical and middle cerebral artery, fetal head position and on transperineal scan- head perineum distance, caput and moulding. The scan was repeated with the next vaginal examination (VE), ideally 4 hours apart. This scan included an abdominal scan for fetal head position and transperineal scan for head perineum distance, caput and moulding. Cervical dilatation was documented from the corresponding digital VE. 1st scan predictors of EMCS were maternal age, BMI, gestational age, umbilical-cerebral ratio, moulding, HPD, rotation, cervical dilatation and caput. 2nd scan predictors were HPD decrease/hr, rotation improvement/hr, dilatation increase/hr and caput increase/hr. A univariable analysis was performed; each predictor was related to EMCS in a simple logistic regression analysis. Next, a multivariable analysis was performed to investigate which variables were most predictive. One model was constructed on all patients, using baseline and first scan measurements. Another model was constructed on patients with two scans, including second scan measurements as predictors. April 2015-Jan 2018, 270 patients were recruited of which 1 withdrew consent. Of 269 patients, 50(19%) had only a single scan. EMCS was required in 79/269 patients (29%) overall: 69/219 patients with two scans (32%) and 10/50 patients with a single scan (20%). Cervical dilatation change was the most important variable for predictive ability(p<0.0001) and was deduced from 2 scans, followed by a ‘spot’ cervical dilatation, HPD, and caput(p<0.0001) from one scan. A model with these 4 variables had an apparent AUC=0.82. This is the first study to comprehensively quantify the extent to which repeat maternal and intrapartum variables are related to EMCS deliveries longitudinally. When validated externally, these data will inform the development of a more robust prediction model for EMCS. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.