基于宫颈多点活检与宫颈管搔刮术构建并验证宫颈阳性病变的诊断模型  

The diagnosis model of positive cervical lesions established and verified based on cervical directed biopsy and endocervical curettage

作　　者：蒋彬彬 王丽利 金建峰 JIANG Binbin;WANG Lili;JIN Jianfeng(Department of Pathology,Hangzhou Fuyang Women and Children Hospital,Zhejang Province,Hangzhou 311400,Zhejiang,China;Cervical Diagnosis and Treatment Center,Hangzhou Fuyang Women and Children Hospital,Zhejiang Province,Hangzhou 311400,Zhejiang,China)

机构地区：[1]浙江省杭州市富阳区妇幼保健院病理科,杭州311400 [2]浙江省杭州市富阳区妇幼保健院宫颈诊治中心,杭州311400

出　　处：《中国性科学》2024年第7期83-87,共5页Chinese Journal of Human Sexuality

基　　金：杭州市医药卫生科技项目(B20231070)。

摘　　要：目的基于宫颈多点活检(CDB)、宫颈管搔刮术(ECC)构建并验证宫颈阳性病变的诊断模型。方法选取2020年1月至2021年4月浙江省杭州市富阳区妇幼保健院收治的798例行阴道镜下CDB和ECC的宫颈疾病患者作为研究对象,分为训练集(n=559)和测试集(n=239)。根据CDB、ECC活检最差结果,将训练集患者分为CDB+ECC阴性组(n=401)和CDB+ECC阳性组(n=158),分析宫颈阳性病变影响因素并构建验证诊断模型。结果对于训练集患者,Logistic回归结果显示:人乳头瘤病毒(HPV)状态(OR=2.385,95%CI:1.578~3.604)、液基薄层细胞学检查(TCT)结果(OR=2.644,95%CI:2.080~3.361)、宫颈萎缩(OR=0.351,95%CI:0.218~0.566)、阴道镜检查(OR=2.388,95%CI:1.606~3.548)是宫颈阳性病变的独立影响因素(P<0.05)。诊断模型公式=-4.670+0.869×HPV状态+0.972×TCT结果-1.046×宫颈萎缩+0.870×阴道镜检查结果。模型诊断宫颈阳性的受试者工作特征(ROC)曲线的曲线下面积(AUC)为0.804。测试集模型验证ROC曲线的AUC为0.793,模型诊断与模型测试的AUC比较,差异无统计学意义(P>0.05)。结论HPV状态、TCT结果、宫颈萎缩、阴道镜检查结果是宫颈阳性病变的相关影响因素。以CDB和ECC结果为基础的宫颈阳性病变诊断模型具有一定的临床参考价值。Objective To establish and verify the diagnostic model of positive cervical lesions based on cervical directed biopsy(CDB)and endocervical curettage(ECC).Methods A total of 798 patients with cervical diseases undergoing colposcopy with CDB and ECC admitted to Hangzhou Fuyang Women and Children Hospital,Zhejiang Province from January 2020 to April 2021 were selected as the study objects,and were divided into training set(n=559)and test set(n=239).According to the worst results of CDB and ECC biopsy,the patients in the training set were divided into CDB+ECC negative group(n=401)and CDB+ECC positive group(n=158),and the influencing factors of positive cervical lesions were analyzed and a validation diagnosis model was established.Results For the patients in the training set,Logistic regression showed human papilloma virus(HPV)status(OR=2.385,95%CI:1.578-3.604),and thinprep cytologic test(TCT)results(OR=2.644,95%CI:2.080-3.361),cervical atrophy(OR=0.351,95%CI:0.218-0.566),colposcopy(OR=2.388,95%CI:1.606-3.548)were independent influencing factors for positive cervical lesions(P<0.05).Diagnostic model formula=-4.670+0.869×HPV status+0.972×TCT results-1.046×cervical atrophy+0.870×colposcopy results.The area under the curve(AUC)of the receiver operating characteristic(ROC)curve for model diagnosis of cervical positivity was 0.804.The AUC of the ROC curve verified by the test set model was 0.793,and there was no significant difference between the AUC of model diagnosis and model test(P>0.05).Conclusions HPV status,TCT results,cervical atrophy and colposcopy results were related influencing factors for positive cervical lesions.The diagnosis model of positive cervical lesions based on CDB and ECC results has certain clinical reference value.

关 键 词：宫颈多点活检技术 宫颈管搔刮术 液基薄层细胞学检查 模型 诊断价值 

分 类 号：R711[医药卫生—妇产科学]