目的 构建小儿肠套叠X线透视下整复失败的预测模型。 方法 回顾性纳入肠套叠患儿439例，空气灌肠整复成功362例，失败77例，将总样本按照7∶3比例随机拆分为训练集308例和验证集131例。收集患儿空气灌肠参数和临床信息，使用单因素logistic回归筛选变量，然后采用多因素logistic回归分析构建预测模型并绘制诺莫图，评价模型区分度、校准度及决策曲线。 结果 空气灌肠成功与失败组的除性别、肠梗阻外，其余变量组间均差异有统计学意义（P < 0.05）。多因素logistic回归分析筛选的独立危险因素为年龄、病程、肠梗阻、套头位置及空灌压力五个变量用于建模。模型预测灌肠整复失败的区分度良好，AUC = 0.854（95%CI： 0.799 ~ 0.909），截断值为0.24，敏感度、特异度分别为79%、80%；校准曲线显示模型预测风险与实际情况高度吻合，Hosmer?Lemeshow检验χ² = 8.159，P = 0.518；决策曲线显示模型具有临床适用性。 结论 空气灌肠压力值结合临床信息对灌肠复位失败风险具有较好预测能力，可为急诊肠套叠患儿临床决策提供一些参考。

Objective To establish a prediction model for failure of intussusception in children under fluoroscopy. Methods In this study， 439 children with intussusception were retrospectively included in our hospital， with 362 successful and 77 unsuccessful air enema. The total samples were randomly divided into 308 training sets and 131 verification sets according to a ratio of 7∶3. The parameters and clinical information of children with air enema were collected， the variables were screened by single factor Logistic regression analysis， and then the prediction model was constructed by multi?factor Logistic regression analysis and the Nomo diagram was drawn. The differentiation， calibration and decision curve of the model were evaluated. Results Statistical differences were found between the successful and unsuccessful groups except gender and intestinal obstruction （all P < 0.05）. The independent risk factors screened by multivariate Logistic regression analysis were age， course of disease， intestinal obstruction， head position and empty irrigation pressure for modeling. The model had good differentiation in predicting the failure of recirculation， with AUC = 0.854 （95%CI： 0.799 ~ 0.909）， truncation value of 0.24， sensitivity and specificity of 79% and 80%， respectively. The calibration curve showed that the risk predicted by the model was highly consistent with the actual situation. Hosmer Lemeshow test χ² = 8.159， P = 0.518； The decision curve shows that the model has clinical applicability. Conclusion Air enema pressure combined with clinical information has a better ability to predict the risk of failure of enema reduction， which can provide some references for clinical decision?making in children with emergency intussusception.