摘要目的:评估乳管内窥镜风险分级在乳头溢液诊疗中的价值。方法:回顾性分析2017—2023年因乳头溢液行乳管内窥镜检查后手术治疗的687例患者的临床资料，通过手术病理验证及对部分患者的长期随访验证，评估乳管内窥镜风险分级对溢液乳管评估的准确性。同时评估乳管内窥镜风险分级对乳腺癌诊断的敏感性、特异性、阳性预测值、阴性预测值及准确度，并采用受试者工作特征（ROC）曲线对其进行验证。结果:687例乳头溢液患者中101例为乳腺癌患者，其中3类16例（恶性率2.9%）、4类47例（恶性率为61.0%）、5类37例（恶性率为88.1%）、6类1例（恶性率为100%）。乳管内窥镜风险分级对乳腺癌诊断的敏感性为84.2%、特异性为94%、阳性预测值为70.8%、阴性预测值为97.2%、准确度为92.6%。ROC曲线下面积为0.826（ P<0.001）。 结论:乳管内窥镜风险分级对乳头溢液患者乳腺癌发生风险的评估是有效的，对乳头溢液患者的诊疗具有较高的指导价值。

abstractsObjective:To evaluate the value of mammary ductoscopy risk classification in the diagnosis and treatment of nipple discharge.Methods:The clinical data of 687 patients who underwent mammary ductoscopy for nipple discharge from 2017 to 2023 were analyzed retrospectively. The accuracy of mammary ductoscopy risk classification in the assessment of leaking mammary ducts was evaluated through surgical pathologic verification and long-term follow-up verification of some patients. The sensitivity, specificity, positive predictive value, negative predictive value and accuracy of mammary ductoscopy risk classification in the diagnosis of breast cancer were evaluated and verified by receiver operating characteristic (ROC) curve.Results:101 of the 687 patients with nipple discharge had breast cancer, including 16 in category 3 (malignant rate 2.9%), 47 in category 4 (malignant rate 61.0%), 37 in category 5 (malignant rate 88.1%), and 1 in category 6 (malignant rate 100%). The sensitivity, specificity, positive predictive value, negative predictive value, and accuracy of mammary ductoscopy risk classification for the diagnosis of breast cancer were 84.2%, 94%, 70.8%, 97.2%, and 92.6%, respectively. The area under ROC curve was 0.826 ( P<0.001), which was statistically significant. Conclusion:The mammary ductoscopy risk classification is effective in evaluating the risk of breast cancer in patients with nipple discharge, and has a high guiding value for the diagnosis and treatment of these patients.