A visualized clinical model predicting good quality blastocyst development in the first IVF/ICSI cycle

Area of Science:

• Reproductive Medicine
• In Vitro Fertilization (IVF)
• Embryology

Background:

• Predicting good quality blastocyst (GQB) formation is crucial for successful in vitro fertilization (IVF) and intracytoplascial sperm injection (ICSI) cycles.
• Accurate prediction models can optimize patient counseling and treatment strategies.

Purpose of the Study:

• To develop and validate a visualized clinical model for predicting good quality blastocyst (GQB) formation in patients undergoing their first IVF/ICSI cycle.
• To identify key predictors of GQB development.

Main Methods:

• Retrospective analysis of 4783 patients from their first IVF/ICSI cycle (January 2015 - December 2019).
• Development of a nomogram model using LASSO regression on a training set (n=3826) to identify critical predictors.
• Validation of the model's predictive accuracy and discriminative ability using receiver operating characteristic (ROC) and calibration curves on a separate testing set (n=957).

Main Results:

• Maternal age, maternal serum anti-Müllerian hormone (MsAMH) concentration, and the number of oocytes retrieved were identified as significant predictors of GQB formation.
• The nomogram model demonstrated strong predictive ability with Area Under the Curve (AUC) values of 0.831, 0.734, and 0.748 for predicting ≥1, ≥3, and ≥5 GQB in the training set, respectively.
• Validation in the testing set showed similar predictive performance (AUCs of 0.805, 0.695, and 0.707 for ≥1, ≥3, and ≥5 GQB).

Conclusions:

• The developed visualized nomogram model offers significant predictive value for GQB development in the context of initial IVF/ICSI cycles.
• This tool can enhance clinical counseling for patients undergoing their first IVF/ICSI treatment.