Research on Establishing a Prediction Model for Pregnancy Outcomes Based on Retrospective Analysis of 1,131 Cases of Frozen-Thawed Single Embryo Transfer

Abstract

Objective: This study is based on preliminary clinical data from a public welfare project funded by the Jinhua Science and Technology Bureau. A total of 1,131 cases of frozen-thawed single embryo transfer were retrospectively enrolled to investigate clinical indicators such as patient age, embryo developmental stage, and endometrial type. Key factors influencing pregnancy outcomes were identified, and a predictive model was constructed. This provides a prerequisite for analyzing the correlation between the immune functional status of Treg lymphocytes and pregnancy outcomes following embryo transfer, and also serves as a reference for clinical immune assessment and individualized embryo transfer strategies in assisted reproduction. This research intends to set up a pregnancy outcome prediction model relying on 1131 frozen-thawed single embryo transfer instances. It takes into account elements like patients age, d3 embryo transfer, d4 fused embryo, d5 or d6 blastocyst, endometrial thickness and type, etc. Methods: Via systematic data gathering, which encompasses fundamental patient details, data associated with embryo development, and endometrial data, and by utilizing statistical instruments like SPSS and R software for univariate analysis and multivariate logistic regression analysis, independent factors influencing pregnancy results were screened. Based on this situation, a prediction model for pregnancy outcomes was built by using a nomogram. Results: Validated through ROC curves and AUC values, the model showed good discriminatory ability and calibration, efficiently forecasting the pregnancy result of single embryo transfer and offering a significant reference for clinical decision-making. Conclusion: It is feasible to set up a pregnancy outcome prediction model for single embryo transfer, and it has high precision.