Journal of Southeast University (Medical Science Edition), Volume. 44, Issue 3, 355(2025)

Constructing a predictive model for early postpartum pelvic floor dysfunction diseases in postpartum women based on Lasso-Logistic regression

SONG Xiaocui1, GUO Hongyan2, and CAO Lin3
Author Affiliations
  • 1Department of Obstetrics, the First Affiliated Hospital of Xingtai Medical College (the First Hospital of Xingtai), Xingtai 054001, China
  • 2Department of Obstetrics, the Fourth People's Hospital of Hengshui, Hengshui 053000, China
  • 3Physical Examination Center, Xingtai Central Hospital, Xingtai 054001, China
  • show less

    ObjectiveTo explore the influencing factors of early postpartum pelvic floor dysfunction diseases (PFD) in parturients and establish a column chart model for predicting the risk of postpartum PFD based on Lasso-Logistic regression.MethodsA total of 373 women who delivered successfully in the First Hospital of Xingtai from January 2023 to July 2024 were selected as the study objects and divided into modeling cohort group (n=261) and validation cohort group (n=112). All parturients underwent PFD screening at 6-8 weeks postpartum, and a column chart model was constructed in R4.3.1 based on the influencing factors of postpartum PFD screened by Lasso-Logistic regression analysis. Receiver operating characteristic (ROC) curve, Hosmer-Lemeshow test, and calibration curve were used to evaluate the effectiveness of the model in predicting the risk of postpartum PFD. The clinical application value of model was evaluated by decision curve analysis (DCA).Results72 cases of PFD occurred in the modeling cohort group, 34 cases of PFD occurred in the validation cohort group, and the total incidence of PFD in 373 postpartum women was 28.4%. Compared with the non-PFD group, the average maternal age and average neonatal body mass in the PFD group were higher (P<0.05), and there were statistically significant differences in PFD family history, delivery times, delivery mode, perineal laceration and the second stage of labor between the two groups (P<0.05). Lasso-Logistic regression analysis showed that age of parturients, delivery times, delivery mode, perineal laceration, the second stage of labor, and neonatal body mass were all influencing factors of postpartum PFD (P<0.05). ROC curve analysis showed that the area under the curve (AUC) of the modeling cohort was 0.827 (95% CI: 0.772-0.882), and the AUC of the validation cohort was 0.838 (95% CI: 0.762-0.915). In the Hosmer-Lemeshow test, the modeling cohort showed χ2=7.556, P=0.478; the validation cohort showed χ2=4.680, P=0.791. DCA showed that the model had high value in clinical application.ConclusionThe age of parturients, delivery times, delivery mode, perineal laceration, the second stage of labor, and neonatal body mass are the influencing factors for postpartum PFD. The column chart model established based on these six factors has better predictive performance and it can help guide clinical prevention of postpartum PFD.

    Tools

    Get Citation

    Copy Citation Text

    SONG Xiaocui, GUO Hongyan, CAO Lin. Constructing a predictive model for early postpartum pelvic floor dysfunction diseases in postpartum women based on Lasso-Logistic regression[J]. Journal of Southeast University (Medical Science Edition), 2025, 44(3): 355

    Download Citation

    EndNote(RIS)BibTexPlain Text
    Save article for my favorites
    Paper Information

    Category:

    Received: Feb. 11, 2025

    Accepted: Aug. 26, 2025

    Published Online: Aug. 26, 2025

    The Author Email:

    DOI:10.3969/j.issn.1671-6264.2025.03.002

    Topics