Skip to main content
SMFM 2024 Pregnancy Meeting Main banner
Final Program


Icon Legend
This session is not in your schedule.
This session is in your schedule. Click again to remove it.
Presentation Icons
Additional registration fee required
Additional registration fee required
Faculty have requested this content not be shared on social media
Faculty have requested this content not be shared on social media
CME Credit Offered
CME Credit Offered
Poster Icons
Foundation Awardee
Foundation Awardee
No Social Media
No Social Media
Back
Q&A

Clinical Obstetrics

Poster Session 1

(180) Personalized Approach to Predict Twin Mode of Delivery

Tuesday, February 13, 2024
10:30 AM – 12:00 PM  
Location: Prince George's Exhibit Hall CDE, Lower Atrium Level

    Primary & Presenting Author(s)

  • Alexis C. Gimovsky, MD

    Associate Professor
    Women & Infants Hospital of Rhode Island, Warren Alpert Medical School of Brown University
    Providence, RI, United States

    • Coauthor(s)

  • SM

    Silas Monje, BA

    Alpert Medical School of Brown University
    Providence, RI, United States

  • JD

    Jack Dunn, PhD

    Interpretable AI
    Cambridge, MA, United States

  • JL

    Jordan Levine, BA, MA

    Dynamic Sky
    Providence, RI, United States

Objective: The reported incidence of combined twin delivery (vaginal delivery of twin A followed by cesarean of twin B) ranges between 5% and 10%. These estimates can be useful overall, but personalized predictions are more clinically valuable for patient centered counseling. We aimed to develop a model that predicts the chance of successful vaginal delivery of twins based on pre-labor patient characteristics for an individual patient using a modern machine learning approach. 

Study Design:

Secondary analysis of the Consortium for Safe Labor data set. Women with non-anomalous, diamniotic twins with a cephalic presenting twin between 32 0/7 and 38 6/7 weeks gestation were included. Only women attempting vaginal birth of twins were analyzed. We performed a retrospective cohort study, using optimal classification trees to predict twin mode of delivery (both vaginal, both cesarean or combined vaginal/cesarean).  Pre-labor patient characteristics were examined to identify variables significantly associated with the success or failure of twin vaginl birth.

Results: 830 patients met were analyzed after removing test set patients. Pre-labor characteristics associated with delivery mode included parity, pre-pregnancy body mass index, maternal age, history of prior cesarean and presentation of baby B. Using these characteristics, a predictive model was developed. A test set was withheld and then utilized to create a receiver operator curve. Accuracy was assessed with an area under the curve of AUC of 0.65.

Conclusion:

We report the successful development of a predictive calculator for delivery mode in diamniotic twin pregnancy. This tool will be useful in patient counseling regarding mode of delivery.