Validation of a Clinical Model to Predict Starting Initial Gonadotropin Dose in Women Undergoing IVF

Background & Objective: The selection of the initial gonadotropin dose in assisted reproductive techniques, particularly in In-Vitro Fertilization (IVF), can enhance patient outcomes and minimize the risks associated with over- or under-stimulation during IVF cycles. The initial gonadotropin dose can be predicted through variables such as age, Anti-Müllerian Hormone (AMH), and Follicle-Stimulating Hormone (FSH). The present study was conducted with aim to validate the La Marca nomogram for predicting the initial gonadotropin dose in women undergoing IVF. Materials & Methods: This cross-sectional observational study was conducted on 187 women undergoing IVF at the infertility center of Shariati Hospital, affiliated with Iran University of Medical Sciences in Tehran, Iran, in 2024. Data regarding age, Body Mass Index (BMI), Antral Follicle Count (AFC), AMH, FSH, oocyte count, oocyte quality, and initial gonadotropin dose were extracted from patients' medical records. The validity and performance of the La Marca nomogram were assessed through cross-validation. The multivariate linear regression was used for analysis of the data. Results: The mean age of women was 34.37±4.78 years. Multivariate analysis across all datasets revealed that age (β=0.04), AMH (-0.16), and FSH (0.09) were significant determinants of the initial gonadotropin dose (P-value<0.05). A 10-fold cross-validation showed similar performance of classifiers in both training and test datasets, indicating minimal overfitting. In the test data, regression coefficients for age, AMH, and FSH were 0.05, -0.16, and 0.10, respectively Conclusion: The present study provides evidence regarding the appropriate performance of the La Marca nomogram for predicting the initial gonadotropin dose in women undergoing IVF.