个体化预测体外循环心脏术后严重高乳酸血症风险列线图模型的构建

Establishment of Nomogram model for individualized prediction of the risk of severe hyperlactatemia after cardiopulmonary bypass heart surgery

目的 构建个体化预测体外循环心脏术后严重高乳酸血症风险列线图模型。方法 选取2020年1月至2022年12月于江苏省人民医院行体外循环心脏手术的患者作为为训练集,按照相同标准选取2023年1月至2023年6月的患者作为验证集。收集训练集患者的临床资料,根据严重高乳酸血症发生情况将患者分为严重高乳酸血症组和非严重高乳酸血症组。使用单因素和多因素Logistic回归分析体外循环心脏术后发生严重高乳酸血症的危险因素,并建立相关列线图预测模型。采用一致性指数(C-index)、校准曲线、受试者工作特征(ROC)曲线和决策曲线评估模型的预测效能。结果 训练集入组的469例体外循环心脏手术患者中,严重高乳酸血症组61例,非严重高乳酸血症组408例。两组ASA分级、NYHA心功能分级、体外循环时间、手术时间、急性肾损伤和呼吸机使用时间比较差异有统计学意义(P<0.05)。多因素Logistic回归分析显示,ASA分级为Ⅳ~Ⅴ级、NYHA心功能分级为Ⅲ~Ⅳ级、体外循环时间≥120 min、手术时间≥5 h、急性肾损伤和呼吸机使用时间≥24 h是体外循环心脏术后发生严重高乳酸血症的独立危险因素(P<0.05)。训练集和验证集的C-index分别为0.807(95%CI0.779~0.835)和0.792(95%CI0.768~0.816),Hosmer-Lemeshoe拟合优度检验显示训练集和验证集结果分别为(χ^(2)=6.418,P=0.506)和(χ^(2)=5.767,P=0.457),训练集和验证集ROC曲线下面积(AUC)分别为0.801(95%CI0.772~0.831)和0.787(95%CI0.761~0.813),在1%~80%的列线图可预测范围内,模型净获益。结论 ASA分级为Ⅳ~Ⅴ级、NYHA心功能分级为Ⅲ~Ⅳ级、体外循环时间≥120 min、手术时间≥5 h、急性肾损伤和呼吸机使用时间≥24 h是体外循环心脏术后发生严重高乳酸血症的危险因素,基于上述因素构建的列线图模型对体外循环心脏术后严重高乳酸血症的发生具有一定的预测价值。

Objective To establish a Nomogram model for personalized prediction of the risk of severe hyperlactatemia after cardiopulmonary bypass heart surgery.Methods Patients who underwent cardiopulmonary bypass cardiac surgery in Jiangsu Provincial People′s Hospital from January 2020 to December 2022 were selected as the training set,and patients from January 2023 to June 2023 were selected as the validation set according to the same criteria.Clinical data of the patients in the training set were collected,and they were divided into a severe hyperlactatemia group and a non-severe hyperlactatemia group based on the occurrence of severe hyperlactatemia.Univariate and multivariate Logistic regression were used to analyze the risk factors for severe hyperlactatemia after cardiopulmonary bypass heart surgery,and relevant Nomogram prediction models were established.The concordance index(C-index),the calibration curve,the receiver operating characteristic(ROC)curve and decision curve analysis(DCA)were used to evaluate the prediction efficiency of the model.Results Among the 469 patients with cardiopulmonary bypass cardiac surgery enrolled in the training set,61 patients were in the severe hyperlactatemia group and 408 patients were in the non-severe hyperlactatemia group.There was a statistically significant difference in American Society of Anesthesiologists(ASA)grading,New York Heart Association(NYHA)cardiac function grading,cardiopulmonary bypass time,surgery time,acute kidney injury,and ventilator usage time between the two group(P<0.05).Multivariate Logistic regression analysis showed that ASA grading ofⅣ-Ⅴ,NYHA cardiac function grading ofⅢ-Ⅳ,cadiopulmonary bypass time≥120 min,surgery time≥5 h,acute kidney injury,and ventilator usage time≥24 h were independent risk factors for severe hyperlactatemia after cardiopulmonary bypass heart surgery(P<0.05).The C-index for the training set and the validation set was 0.807(95%CI 0.779-0.835)and 0.792(95%CI 0.768-0.816),respectively.The Hosmer-Lemeshoe goodness-of-fit test showed the results of the training set and the validation set wereχ^(2)=6.418,P=0.506 andχ^(2)=5.767,P=0.457.The AUC of the ROC curve for the training set and the validation set was 0.801(95%CI 0.772-0.831)and 0.787(95%CI 0.761-0.813),respectively.Within the predictable range of 1%to 80%of the column chart,the model has a net benefit.Conclusions ASA grading ofⅣ-Ⅴ,NYHA cardiac function grading ofⅢ-Ⅳ,cardiopulmonary bypass time≥120 min,surgery time≥5 h,acute kidney injury,and ventilator usage time≥24 h are risk factors for severe hyperlactatemia after cardiopulmonary bypass heart surgery.The Nomogram model constructed based on the above factors has certain predictive value for the occurrence of severe hyperlactatemia after cardiopulmonary bypass heart surgery.