Combining demographic data and transvaginal ultrasonography: a predictive model for endometrial carcinoma in postmenopausal patients

• Published in: BMC women's health Vol. 24; no. 1; pp. 539 - 7
• Main Authors: Li, Xueru, Wang, Haiyan, Wang, Tong, Cui, Haiou, Wu, Lixian, Wang, Wen, Wang, Fuxia
• Format: Journal Article
• Language: English
• Published: England BioMed Central Ltd 27.09.2024; BioMed Central; BMC
• Subjects: Aged; Biopsy; Body Mass Index; Cancer; Cancer therapies; Carcinoma; Care and treatment; Demography; Diabetes; Endometrial cancer; Endometrial Neoplasms - diagnosis; Endometrial Neoplasms - diagnostic imaging; Endometrial Neoplasms - pathology; Endometrium - diagnostic imaging; Endometrium - pathology; Female; Health aspects; Hormone replacement therapy; Hospitals; Humans; Hyperplasia; Hypertension; Logistic Models; Menopause; Methods; Middle Aged; Morphology; Oncology, Experimental; Ovaries; Performance evaluation; Polycystic ovary syndrome; Postmenopausal women; Postmenopause; Practice guidelines (Medicine); Prediction models; Predictive Value of Tests; Regression analysis; Retrospective Studies; Risk assessment; Risk Assessment - methods; Risk Factors; ROC Curve; Sensitivity and specificity; Statistical models; Strategic planning (Business); Transvaginal ultrasonography; Tumors; Ultrasonic imaging; Ultrasonography - methods; Ultrasound imaging; Uterine Hemorrhage - etiology; Uterus; China; Sensitivity and specificity; Demography; Endometrial cancer; Transvaginal ultrasonography
• ISSN: 1472-6874; 1472-6874
• DOI: 10.1186/s12905-024-03374-8

Although clinical guidelines exist for diagnosing abnormal uterine bleeding, there is a significant lack of agreement on the best management strategies for women presenting with symptom, particularly in diagnosing endometrial cancer. This study aimed to develop a preoperative risk model that utilizes demographic factors and transvaginal ultrasonography of the endometrium to assess and predict the risk of malignancy in females with endometrial cancer. In this retrospective study, a logistic regression model was developed to predict endometrial carcinoma using data from 356 postmenopausal women with endometrial lesions and an endometrial thickness (ET) of 5 mm or more. These patients had undergone transvaginal ultrasonography prior to surgery, with findings including 247 benign and 109 malignant cases. The model's predictive performance was evaluated using receiver operating characteristic (ROC) curve analysis and compared with post-surgical pathological diagnoses. Our model incorporates several predictors for endometrial carcinoma, including age, history of hypertension, history of diabetes, body mass index (BMI), duration of vaginal bleeding, endometrial thickness, completeness of the endometrial line, and endometrial vascularization. It demonstrated a strong prediction with an area under the curve (AUC) of 0.905 (95% CI, 0.865-0.945). At the optimal risk threshold of 0.33, the model achieved a sensitivity of 82.18% and a specificity of 92.80%. The established model, which integrates ultrasound evaluations with demographic data, provides a specific and sensitive method for assessing and predicting endometrial carcinoma.