反复宫腔内人工授精失败患者IVF助孕受精失败的高危因素及预测模型研究

Study on high risk factors and establishment of a prediction model for IVF fertilization failure after repeated intrauterine insemination failure

目的:分析因反复宫腔内人工授精(IUI)助孕失败而行体外受精(IVF)助孕患者的受精失败危险因素,建立logistic回归预测模型进行受精失败风险的预测。方法:回顾分析2020年1月~2021年11月在广西壮族自治区妇幼保健院生殖医学中心因经3次IUI助孕未孕改行常规IVF治疗的201对患者夫妇的病例资料(IVF第1周期)。以受精率30%为临界值,将患者分为受精正常组(受精率≥30%)和受精失败组(受精率<30%),筛选存在组间差异的风险因素。把P<0.25的风险因素作为二元logistic回归的候选预测变量,进行logistic逐步回归分析(前向法:LR法),建立预测受精失败的模型。通过Hosmer-Lemeshow拟合优度检验和受试者工作特征曲线(ROC)下面积(AUC)对模型的拟合度、预测效果进行评价,并计算敏感度、特异度和总准确度,评价模型的准确性。结果:受精正常组和受精失败组的女方平均年龄、女方不孕年限、男方不育类型、男方不育年限、洗涤前精子前向运动精子百分率和精子正常形态百分率比较,差异有统计学意义(P<0.05)。预测模型通过Hosmer-Lemeshow拟合优度检验显示模型拟合程度良好(χ^(2)=3.823,df=8,P=0.873)。logistic回归模型的ROC曲线下面积为0.780(95%CI为0.691~0.868)。以0.186作为预测概率分界值,灵敏度为74.2%,特异度为75.3%,总准确度为75.1%。结论:女方不孕年限和处理前精子形态正常率是反复IUI失败患者行IVF受精失败的高风险因素。本研究建立的IVF受精失败预测模型预测效能良好,能为胚胎实验室预判受精结局提供参考。

Objective:To explore the risk factors of IVF fertilization failure after repeated intrauterine insemination(IUI) failure, and to establish a logistic regression prediction model for fertilization failure.Methods:A retrospective analysis was conducted on the case data of 201 patients who converted to routine IVF treatment(IVF cycle 1) due to 3 times failure of IUI from January 2020 to November 2021 in Reproductive Medicine Center, Maternal and Child Health Hospital of Guangxi Zhuang Autonomous Region.With the fertilization rate equal 30% as critical value, all patients were divided into normal fertilization group(≥30%) and fertilization failure group(<30%).The risk factors with significant differences were screened between groups.The risk factors with P<0.25 were taken as candidate predictive variables of binary logistic regression, and stepwise logistic regression analysis(Forward :LR) was conducted to establish a model for predicting fertilization failure.The Hosmer-Lemeshow goodness of fit test and area under receiver operating characteristic curve(AUC) were used to evaluate the fitting degree and prediction effect of the model.Sensitivity, specificity and total accuracy were calculated to evaluate the accuracy of the model.Results:There were significant differences between the normal fertilization group and the failed fertilization group in terms of the average female age, the female infertility years, the male infertility type, the male infertility years, the percentage of progressive motility and normal sperm morphology before washing(P<0.05).The Hosmer-Lemeshow goodness of fit test showed that the model fit well(χ^(2)=3.823,df=8,P=0.873).The area under the ROC curve of the logistic regression model was 0.780(95%CI:0.691 ~ 0.868).The sensitivity, specificity and total accuracy of 0.186 were 74.2%,75.3% and 75.1%,respectively.Conclusion:The high risk factors for IVF fertilization failure after repeated IUI failures include: the years of female infertility, and the percentage of sperm with normal sperm morphology before washing.The prediction model of IVF fertilization failure established has good prediction efficiency and can provide a reference for embryo laboratory to predict the fertilization outcomes.