摘要
目的 探讨子宫内膜不典型增生术后病理升级的高危因素并构建预测模型,旨在为早期识别高危人群及制定针对性管理方案提供更多借鉴.方法 回顾性纳入2019年1月至2022年1月在中国人民解放军海军第九七一医院行诊断性刮宫确诊子宫内膜不典型增生的患者共110例,根据全子宫切除术后是否升级为恶性病变分为恶变组(37例)和非恶变组(73例),采用单因素和多因素法评估刮宫术后病理组织学检查升级为恶性病变的独立危险因素,构建基于上述独立危险因素列线图模型,描绘受试者工作特征(ROC)曲线分析列线图模型用于刮宫术后病理组织学检查升级为恶性病变风险的预测临床效能.结果 单因素分析结果显示,绝经状态和超声评估＞Ⅰ级血流信号均可能与刮宫术后病理组织学检查升级为恶性病变有关(P＜0.05);Logistic回归模型多因素分析结果显示,已绝经和超声评估＞Ⅰ级血流信号均是刮宫术后病理组织学检查升级为恶性病变的独立危险因素(P＜0.05).基于上述独立危险因素构建列线图,其中已绝经和超声评估＞Ⅰ级血流信号分别赋值为23分、100分;Bootstrap软件内部验证结果显示上述列线图模型一致性指数为0.85,提示模型预测区分度良好.上述列线图模型预测曲线下面积为0.87(95％CI:0.81～0.95),最佳截断值为90分,模型预测灵敏度和特异度分别为82.48％、90.73％.结论 子宫内膜不典型增生刮宫术后病理组织学检查升级为恶性病变与绝经状态和超声评估血流信号分级有关;基于上述危险因素构建列线图模型用于预测刮宫术后病理组织学检查升级为恶性病变具有良好效能.
Abstract
Objective To investigate the risk factors of histopathological examination upgrade to malignant lesions after curettage of patients with endometrial dysplasia and construct the predictive model to provide more reference for postoperative management and individualized intervention plan formulation.Methods We retrospectively selected 110 patients with endometrial dysplasia who underwent diagnostic curettage between January 2019 and January 2022 in our hospital.Patients were divided into the malignant lesion group(37 cases)and the non-malignant lesion group(73 cases)according to whether lesion upgrade to malignant lesion after total hysterectomy.Univariate and multivariate methods were used to evaluate the independent risk factors for malignant lesions upgraded by histopathologic examination after curettage and the Nomogram model based on the above independent risk factors was constructed.The clinical efficacy of receiver Operating characteristic(ROC)curve analysis and Nomogram model were used to evaluate the clinical effectiveness of predicting of histopathological examination upgrade to malignant lesions after curettage.Results Univariate analysis showed that menopausal status and ultrasound assessment of blood flow signal greater than grade Ⅰ may be associated with histopathologic upgrade to malignant lesions after curettage(P＜0.05).Logistic regression model multivariate analysis results showed that postmenopausal and ultrasound assessments with blood flow signals greater than grade Ⅰ were independent risk factors for upgrading to malignant lesions by histopathologic examination after curettage(P＜0.05).The constructed Nomogram assigned 23 points for postmenopausal status and 100 points for ultrasound assessments with blood flow signals greater than grade Ⅰ.The internal verification result of Bootstrap software showed that the model consistency index of the above column chart was 0.85,indicating that the model predicted has satisfactory differentiation.The area under the curve predicted by the Nomogram model was 0.87(95％CI:0.81-0.95),and the optimal cut-off value was 90 points and the prediction sensitivity and specificity of the Nomogram model were 82.48％and 90.73％,respectively.Conclusions The histopathologic upgrade to malignant lesion after curettage of endometrial atypical hyperplasia was related to menopausal status and the grading of blood flow signal evaluated by ultrasound.Based on the above risk factors,it is effective to construct the Nanogram model to predict the histopathologic upgrade to malignant lesions after curettage.
NETL
NSTL
维普
万方数据