A prediction model of live birth in patients undergoing frozen⁃thawed embryo transfer after TCRA / 安徽医科大学学报

ABSTRACT

Objective @#To develop a prediction model for live birth based on the basic characters of patients after transcervical resection of adhesions (TCRA) undergoing frozen⁃thawed embryo transfer (F⁃ET) .@*Methods @# A total of 491 cycles who went F ⁃ET after TCRA were included. The cycles were randomly divided into training (n = 347) and validation (n = 144) cohorts at 7 ∶ 3 ratio. Relevant factors of live birth were selected via logistic regression analysis based on the data from patients in the training cohort. Multivariate logistic regression analysis was used to establish a nomogram , and then this predict model was calibrated and verified. @*Results @#Age , body mass index (BMI) , miscarriages and endometrial thickness were significantly related to the live birth. The statistical nomogram was built with the area under the ROC cu2rve for the training cohort was 0. 713 (95% CI: 0. 519 - 0. 815) , Hosmer⁃Lemeshow goodness of fit test were χ = 7. 062 , P = 0. 530 , and the slope of calibration curve was close to 1. AUC of validation cohort was 0. 609 (95% CI: 0. 660 - 0. 561) , presenting a well⁃pleasing goodness⁃of⁃fit and stability in this model.@*Conclusion @#Age , BMI , miscarriages and endometrial thickness can better predict the probability of live birth undergoing F ⁃ET after TCRA.