# Construction and Validation of a Clinical Pregnancy Outcome Prediction Model for Infertility Treatment Using IVF/ICSI: A Retrospective Study Based on 11,449 Cases

## TL;DR

This study developed a model to predict pregnancy outcomes after IVF/ICSI using data from over 11,000 patients, helping clinicians make better treatment decisions.

## Contribution

A validated predictive model for IVF/ICSI pregnancy outcomes using LASSO regression and a nomogram for clinical decision support.

## Key findings

- Eight key predictors of pregnancy outcomes were identified, including male age, AFC, FSH, and embryo quality.
- The model showed strong performance with AUC values of 0.839 in training and 0.827 in validation.
- Calibration and decision curve analysis confirmed the model's accuracy and clinical utility.

## Abstract

Infertility is a prevalent global reproductive health issue. In vitro fertilization (IVF) and intracytoplasmic sperm injection (ICSI), as pivotal assisted reproductive technologies, are widely implemented in clinical practice. However, clinical pregnancy outcomes following IVF/ICSI are influenced by various factors, making accurate prediction essential for optimizing treatment strategies.

To develop and validate a predictive model for clinical pregnancy outcomes following IVF/ICSI treatment.

A retrospective analysis was conducted on clinical data from 154,307 patients who underwent assisted reproductive treatment due to infertility at the First People’s Hospital of Yunnan Province. Based on inclusion and exclusion criteria, 11,449 patients who underwent IVF/ICSI were included. Key predictors were identified using LASSO regression. A Nomogram scoring system was developed for an intuitive visualization of individualized prediction results. Model performance was evaluated using the area under the receiver operating characteristic (ROC) curve, calibration curves, decision curve analysis (DCA), and clinical impact curves.

LASSO regression identified eight critical predictors influencing clinical pregnancy outcomes: male age, antral follicle count (AFC), Day 3 follicle-stimulating hormone (FSH) level, endometrial thickness, female age, number of usable embryos, number of high-quality blastocysts, and number of embryos transferred. The predictive model demonstrated excellent performance in both the training and validation cohorts, with AUC values of 0.839 [95% CI (0.825, 0.852)] and 0.827 [95% CI (0.817, 0.835)], respectively, indicating strong discriminatory ability. Calibration curves confirmed a high degree of consistency between predicted probabilities and actual outcomes. Decision curve analysis revealed substantial net clinical benefit across various risk thresholds, while clinical impact curves further validated the model’s practical applicability in clinical settings.

This study identified key factors influencing clinical pregnancy outcomes following IVF/ICSI treatment, including male age, antral follicle count (AFC), Day 3 follicle-stimulating hormone (FSH) level, endometrial thickness, female age, number of usable embryos, number of high-quality blastocysts, and number of embryos transferred. This model serves as a scientifically sound decision-support tool for clinicians in the management of infertility treatment with IVF/ICSI.