Skip to main content
SMFM 2026 Pregnancy Meeting Main banner
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 outside of the session
Faculty have requested this content not be shared outside of the session
CME Credit Offered
CME Credit Offered
Abstract Award
Abstract Award
Recording available 2/16-5/2
Recording available 2/16-5/2
AIUM Credit
AIUM Credit
Foundation Awardee
Foundation Awardee
Poster Icons
Abstract Award
Abstract Award
Foundation Awardee
Foundation Awardee

Poster Session 1

Category: Hypertension

Poster Session 1

(174) Comparing Preeclampsia characteristics in Singleton and Twin Pregnancies using Machine Learning

Wednesday, February 11, 2026
10:30 AM - 12:00 PM  

    Submitting Author and Presenting Author(s)

  • Daniel Gabbai, MD, MPH

    Lis Hospital for Women’s Health, Tel Aviv Sourasky Medical Center
    Lis Hospital for Women’s Health, Tel Aviv Sourasky Medical Center, Tel Aviv, Israel

    • Coauthor(s)

  • Itamar Gilboa, MD

    Lis Hospital for Women’s Health, Tel Aviv Sourasky Medical Center
    Lis Hospital for Women’s Health, Tel Aviv Sourasky Medical Center, Tel Aviv, Israel

  • Lee Reicher, MD, PhD (she/her/hers)

    Resident
    Lis Hospital for Women’s Health, Tel Aviv Sourasky Medical Center
    Tel Aviv, HaMerkaz, Israel

  • YY

    Yariv Yogev, MD

    Lis Hospital for Women’s Health, Tel Aviv Sourasky Medical Center Gray Faculty of Medicine, Tel Aviv University, Israel
    Lis Hospital for Women’s Health, Tel Aviv Sourasky Medical Center, Tel Aviv, Israel

  • Emmanuel Attali, MD (he/him/his)

    Head of Maternal Wards
    Lis Hospital for Women’s Health, Tel Aviv Sourasky Medical Center
    Lis Hospital for Women’s Health, Tel Aviv Sourasky Medical Center, Tel Aviv, Israel

Objective:
To compare characteristics of preeclampsia (PET) in singleton and twin pregnancies using machine learning, and to determine differences in their relative importance.

Study Design:
A retrospective cohort study at a university-affiliated tertiary medical center (2012–2024), including all singleton and twin pregnancies. The primary outcome was PET, defined according to ACOG criteria. Characteristics included demographic, clinical, obstetric, and routine laboratory data available prior to delivery. Missing values were imputed using the median. Data were randomly split into training (70%) and testing (30%) sets. Separate XGBoost models were developed for each group, Feature importance was compared between groups using both model gain and SHAP (SHapley Additive exPlanations) value.

Results:
1. A total of 141,425 singleton pregnancies and 5,574 twin pregnancies were included. PET occurred in 2,453 (1.7%) singleton pregnancies and 522 (9.4%) twin pregnancies.
2. The models demonstrated high discrimination in both groups (AUC = 0.92 for singletons, 0.95 for twins). 
3. Among twins, the most powerful predictors of PET (by model gain) were albumin/creatinine ratio (31%), pre-pregnancy BMI (8%), platelet count (8%), neutrophil-to-lymphocyte ratio (7%), maternal age (7%), white blood cell count (6%), hemoglobin (5%), AST (5%), gestational weight gain (5%), and gestational age at delivery (4%).
4. While the same predictors were relevant in singletons, SHAP analysis revealed distinct differences in their effect distributions between groups.



Conclusion:
Although singleton and twin pregnancies share similar predictors for PET, the magnitude and direction of their influence differ. These findings support the development of group-specific risk assessment tools to enhance preeclampsia prediction and prevention strategies