Developing a Predictive Model for Minimal or Mild Endometriosis as a Clinical Screening Tool in Infertile Women:
Jie Zhang1, Jing Wang1, Jingyi Zhang1
1State Key Laboratory of Reproductive Medicine, Clinical Center of Reproductive Medicine, The First Affiliated Hospital of Nanjing Medical University (Ms. Jie Zhang, Ms. Jingyi Zhang, Ms. Xu, Ms. Zhu, Mr. Dai, and Drs. Wang, Shu, Jinyong Liu, Hou, Diao, Jiayin Liu, and Mao).
Study Objective:
To develop a noninvasive predictive model based on patients with infertility for identifying minimal or mild endometriosis.
Design:
A retrospective cohort study.
Setting:
This study was conducted at a tertiary referral center.
Patients:
A total of consecutive 1365 patients with infertility who underwent laparoscopy between January 2013 and August 2020 were divided into a training set (n = 910) for developing the predictive model and a validation set (n = 455) to confirm the model's prediction efficiency. The patients were randomly assigned in a 2:1 ratio.
Interventions:
Sensitivities, specificities, area under the curve, the Hosmer-Lemeshow goodness of fit test, Net Reclassification Improvement index, and Integrated Discrimination Improvement index were evaluated in the training set to select the optimum model. In the validation set, the model's discriminations, calibrations, and clinical use were tested for validation.
Measurements And Main Results:
In the training set, there were 587 patients with minimal or mild endometriosis and 323 patients without endometriosis. The combination of clinical parameters in the model was evaluated for both statistical and clinical significance. The best-performing model ultimately included body mass index, dysmenorrhea, dyspareunia, uterosacral tenderness, and serum cancer antigen 125 (CA-125). The nomogram based on this model demonstrated sensitivities of 87.7% and 93.3%, specificities of 68.6% and 66.4%, and area under the curve of 0.84 (95% confidence interval 0.81-0.87) and 0.85 (95% confidence interval 0.80-0.89) for the training and validation sets, respectively. Calibration curves and decision curve analyses also indicated that the model had good calibration and clinical value. Uterosacral tenderness emerged as the most valuable predictor.
Conclusion:
This study successfully developed a predictive model with high accuracy in identifying infertile women with minimal or mild endometriosis based on clinical characteristics, signs, and cost-effective blood tests. This model would assist clinicians in screening infertile women for minimal or mild endometriosis, thereby facilitating early diagnosis and treatment.
