宫颈息肉宫腔镜下切除术后复发的风险预测可视化模型建立及效能验证  

Establishment and Efficacy Verification of a Visual Model for Predicting the Risk of Recurrence of Cervical Polyps After Hysteroscopic Resection

作　　者：张静蕾[1] 徐流立[1] 马媛媛 杨霄[1] ZHANG Jinglei;XU Liuli;MA Yuanyuan(Department of Obstetrics and Gynecology,Anyang Regional Hospital,Anyang City,He’nan Province 455000;不详)

机构地区：[1]濮阳市安阳地区医院妇产科,河南省安阳市455000 [2]安阳市中医院妇产科

出　　处：《医学理论与实践》2024年第18期3081-3085,共5页The Journal of Medical Theory and Practice

摘　　要：目的:建立宫颈息肉宫腔镜下切除术后复发的风险预测可视化模型,并验证其效能。方法:回顾性收集2019年1月-2022年7月于本院行宫腔镜下切除术的420例患者的临床资料,经计算机产生随机数表以2∶1将其分为训练集(280例)、验证集(140例)。所有患者均接受随访,将复发患者纳入复发组,其余纳入无复发组。比较训练集复发组、无复发组一般资料;用logistic回归模型分析宫颈息肉宫腔镜下切除术后复发的影响因素,并建立回归方程;用受试者工作特征曲线(ROC)、Calibration曲线、决策曲线评价模型的预测效能、校准能力及临床净获益。结果:随访时间12~36个月,中位随访时间18个月。训练集患者宫颈息肉宫腔镜下切除术后复发率11.79%;logistic回归模型显示,年龄、月经不规则、人乳头瘤病毒(HPV)感染、产次、术后未治疗是宫颈息肉宫腔镜下切除术后复发的影响因素(P<0.05);建立logistic回归方程:Y=0.563 X_(1)+0.752 X_(2)+0.854 X_(3)+0.621 X_(4)+0.806 X_(5);基于训练集logistic回归分析结构建立宫颈息肉宫腔镜下切除术后复发的风险预测列线图模型,该模型预测训练集、验证集术后复发的灵敏度分别为97.14%、94.29%,特异度分别为91.74%、80.73%,曲线下面积(AUC)分别为0.956、0.911;训练集、验证集Calibration曲线经Hosmer-Lemeshow检验,差异无统计学意义(P>0.05);Bootstrap法内部验证结果显示,训练集、验证集的C-index指数分别为0.911(95%CI:0.832~0.990)、0.905(95%CI:0.841~0.969);训练集、验证集分别在风险阈值0~0.83、0~0.61范围内获取临床净收益。结论:年龄、月经不规则、HPV感染、产次、术后未治疗是宫颈息肉宫腔镜下切除术后复发的影响因素,据此建立的风险预测列线图模型效能良好。Objective:To establish a visualized model for predicting the risk of recurrence after hysteroscopic resection of cervical polyps,and to verify its efficacy.Methods:The clinical data of 420 patients who underwent hysteroscopic resection of cervical polyps in our hospital from January 2019 to July 2022 were retrospectively collected,and they were randomly divided into a training set(280 cases)and a validation set(140 cases)by a computer-generated random number table at a ratio of 2∶1.All patients were followed up,and those who had recurrence were included in the recurrence group,while the rest were included in the non-recurrence group.The general data of the recurrence group and the non-recurrence group in the training set were compared.Logistic regression model was used to analyze the influencing factors of recurrence after hysteroscopic resection of cervical polyps,and a regression equation was established.The receiver operating characteristic curve(ROC),Calibration curve,and decision curve were used to evaluate the predictive efficacy,calibration ability,and clinical net benefit of the model.Results:The follow-up time was 12~36 months,with a median follow-up time of 18 months.The recurrence rate after hysteroscopic resection of cervical polyps in the training set was 11.79%.Logistic regression model showed that age,irregular menstruation,human papillomavirus(HPV)infection,parity,and untreated postoperative were influencing factors of recurrence after hysteroscopic resection of cervical polyps(P<0.05).Logistic regression equation was established:Y=0.563 X_(1)+0.752 X_(2)+0.854 X_(3)+0.621 X_(4)+0.806 X_(5).Based on the structure of logistic regression analysis in the training set,a nomogram model for predicting the risk of recurrence after hysteroscopic resection of cervical polyps was established,which predicted the sensitivity of recurrence after surgery in the training set and validation set were 97.14%and 94.29%,respectively,and specificity were 91.74%and 80.73%,respectively,and the area under the curve(AUC

关 键 词：宫颈息肉 宫腔镜下切除术 复发 列线图模型 

分 类 号：R711.74[医药卫生—妇产科学]