Development of a cross-validated model for predicting emergency cesarean for intrapartum fetal compromise at term

Christopher Flatley, Kristen Stacey Gibbons, Cameron Hurst, Sailesh Kumar

Research output: Contribution to journalArticlepeer-review


Objectives: To develop a model for predicting emergency cesarean for fetal distress (ECFD) at term using a combination of maternal and late pregnancy ultrasound parameters measured at more than 36 gestational weeks. 

Methods: A study of prospectively collected data, including ultrasound scans at 36–38 weeks, for singleton non-anomalous deliveries at Mater Mother's Hospital, Brisbane, Australia, between January 2010 and April 2017. Univariable and multivariable mixed-effects generalized linear models were generated. The final model was validated by the K-fold cross validation technique. 

Results: Overall, 5439 women met the inclusion criteria; of these, 230 (4.2%) underwent ECFD. There were more nulliparous women and women with induction of labor (IOL) in the ECFD cohort (both P < 0.001). ECFD neonates had lower z-scores for estimated fetal weight (EFW), cerebroplacental ratio (CPR), and middle cerebral artery pulsatility index; and higher scores for umbilical artery pulsatility index. Ethnicity, nulliparity, IOL, EFW z-score, and CPR z-score were included in the final prediction model, which showed high accuracy with an area under the receiver operator characteristic curve of 0.77. 

Conclusion: The study shows that a prediction model combining the continuous standardized measures of CPR and EFW and several maternal factors was able to identify ECFD with improved accuracy.

Original languageEnglish
Pages (from-to)41-47
Number of pages7
JournalInternational Journal of Gynecology and Obstetrics
Issue number1
Publication statusPublished - 1 Jan 2020
Externally publishedYes


Dive into the research topics of 'Development of a cross-validated model for predicting emergency cesarean for intrapartum fetal compromise at term'. Together they form a unique fingerprint.

Cite this