Predicting the risk of endometriosis in Chinese infertile women: development and assessment of a predictive nomogram

Purpose This study aimed to establish a risk prediction model of endometriosis in infertile women and verify the model. Methods A retrospective study was made of 140 infertile women hospitalized at Henan Provincial People's Hospital between January 2018 and May 2024. They were divided into the Endometriosis group (EMs) and the No Endometriosis group (No-EMs). The baseline characteristics of the two groups were compared. The least absolute shrinkage and selection operator (LASSO) regression model was utilized to optimize feature selection. Subsequently, logistic regression (LR) analysis was utilized to formulate a predictive model that integrated the selected features. The discrimination and calibration of the predictive model were evaluated using the C-index and calibration plot. Internal validation was conducted using bootstrapping methods. Results The LASSO regression model identified five feature selections: menstrual pattern, menstrual cycle length, severity of dysmenorrhea, duration of infertility, and type of infertility. LR analysis revealed that the severity of dysmenorrhea (OR = 10.278, 95% CI = 2.372–73.400, p = 0.005) and the type of infertility (OR = 2.604, 95% CI = 1.247–5.563, p = 0.012) emerged as independent risk factors for EMs in infertile women. The model displayed good discrimination with a C-index of 0.743 (95% CI = 0.660–0.826)and good calibration. Internal validation through the Bootstrap method confirmed a high C-index value of 0.709. Conclusion The development of Nomogram prediction models offers significant clinical predictive utility in evaluating the risk of EMs among infertile women. It equips clinicians with rational treatment strategies and novel perspectives for managing infertile women.