基于倾向性评分匹配的非计划再次手术危险因素分析及预测模型构建

Analysis of risk factors for unplanned reoperation and construction of aprediction model based on propensity score matching

目的分析非计划再次手术的危险因素,构建列线图风险预测模型筛选高危患者,为围手术期管理降低非计划再次手术率提供依据。方法收集2018-2021年该院880例非计划再次手术患者作为非计划再次手术组,运用倾向性评分匹配法按照1∶3比例匹配出2640例患者作为对照组,收集可能的风险因素指标并进行单因素分析,筛选出具有统计学意义的指标进行多因素logistic回归分析,进而构建风险预测模型,并评估该模型的预测价值。结果合并症数量3~5个(OR=1.84)、合并症数量6~10个(OR=2.94)、合并恶性肿瘤(OR=1.75)、合并终末期肾病(OR=1.92)、主要手术级别三级(OR=4.27)、主要手术级别四级(OR=7.26)、切口分组Ⅰ、Ⅱ、Ⅲ类(OR=2.18、1.97、6.85)是非计划再次手术的独立影响因素(P<0.05)。模型通过校准度检验,受试者工作特征(ROC)曲线下面积(AUC)为0.715,模型区分度良好。结论根据风险因素构建的风险预测模型,可用于识别非计划再次手术的高危人群。

Objective To analyze the risk factors of unplanned reoperation and construct a nomogram-based risk prediction model to identify high-risk patients,so as to provide a basis for perioperative management to reduce the rate of reoperation.Methods A total of 880 patients with underwent unplanned reoperation between 2018 and 2021 were included as the unplanned reoperation group.Using propensity score matching,2640 patients were matched in a 1∶3 ratio to form the control group.Potential risk factor indicators were collected and subjected to univariate analysis.Significant indicators were then selected for multivariable logistic regression analysis to construct the risk prediction model.The predictive value of the model was evaluated.Results For unplanned reoperation,The number of complications 3-5(OR=1.84),the number of complications 6-10(OR=2.94),combined with maligant tumor(OR=1.75),combined with end-stage renal disease(OR=1.92),major surgery grade 3(OR=4.27),major surgery grade 4(OR=7.26),and incision gradeⅠ,Ⅱ,Ⅲ(OR=2.18,1.97,6.85)were independent factors(P<0.05).The model passed the calibration degree test,and the area under ROC curve(AUC)was 0.715,indicating good model differentiation.Conclusion A risk prediction model based on risk factors can help identify high-risk populations for unplanned reoperation and suggest corresponding measures for prevention.