Clinical Report: Predictive Model for Inadequate Ovarian Response in ART
Overview
This study developed and validated a predictive model for poor ovarian response (POR) in patients undergoing Assisted Reproductive Technology (ART). The final model identified key predictors.
Background
Assisted Reproductive Technology (ART) is essential for infertility management, yet poor ovarian response (POR) remains a significant challenge. Current diagnostic criteria do not adequately predict POR.
Data Highlights
Predictor
Model Type
Significance
AMH
Stepwise Logistic Regression
Significant
Basal FSH
Stepwise Logistic Regression
Significant
BMI
Stepwise Logistic Regression
Significant
Antral Follicle Count
Stepwise Logistic Regression
Significant
Key Findings
The study included 1,789 patients undergoing IVF/ICSI with a GnRH antagonist protocol.
Four prediction models were developed and compared, with stepwise logistic regression being the optimal model.
Key predictors of POR identified were AMH, basal FSH, BMI, and antral follicle count.
The model demonstrated high discriminatory ability and good calibration across various datasets.
A nomogram was developed.
Clinical Implications
The predictive model can assist clinicians in identifying patients at risk for poor ovarian response before treatment begins, allowing for tailored interventions. The nomogram provides an intuitive tool for risk assessment in clinical practice.
Conclusion
This study presents a robust predictive model for poor ovarian response in ART, enhancing decision-making and potentially improving patient outcomes through personalized treatment strategies.
