A risk prediction nomogram of endometrial carcinoma and precancerous lesions in postmenopausal women: A retrospective study.

This study aimed to develop a risk prediction nomogram for endometrial carcinoma and precancerous lesions in postmenopausal women to provide postmenopausal patients with more information on disease probability, work out personalized medical plans, and reduce unnecessary invasive clinical examinations. We enrolled 340 patients who underwent hysteroscopy at Beijing Maternity Hospital between March 2016 and July 2018. The patients were divided into the low-risk (275 patients) and high-risk (65 patients) groups, according to the results of the pathological examinations. Binary logistic analysis was performed to evaluate the 20 potential risk factors for endometrial cancer and precancerous lesions in postmenopausal women and to screen for certain risk factors using the Statistical Package for the Social Sciences version 26.0. Using R 4.0.3, we built a prediction nomogram that incorporated the selected factors. The discrimination, calibration, and clinical usefulness of the prediction model were assessed using the concordance (C)-index, calibration plot, and decision curve analysis. Internal validation was assessed using bootstrapping validation. Predictors included in the prediction nomogram included obesity, vaginal bleeding, family history of gynecological malignancies, endometrial thickness ≥ 1.15 cm, and color Doppler flow imaging blood flow. The model displayed good discrimination, with a C-index of 0.853, and good calibration. Decision curve analysis showed that the model was clinically useful, with a benefit range of 2% to 93%. A high C-index value of 0.844 could still be reached in the interval validation. Obesity, vaginal bleeding, family history of gynecological malignancies, endometrial thickness ≥ 1.15 cm, and color Doppler flow imaging blood flow were independent risk factors for endometrial cancer and precancerous lesions. Thus, the prediction nomogram can be conveniently used to facilitate individual risk prediction in patients with endometrial cancer and precancerous lesions.