重症急性胰腺炎并发腹腔高压预测模型的建立

Establishment of prediction model for severe acute pancreatitis complicated with abdominal hypertension

目的构建SAP并发腹腔内高压(IAH)的预测模型,并对模型的预测效能进行评价。方法回顾性分析2017年1月至2021年12月间河北省沧州中西医结合医院急诊科收治的322例SAP患者的临床资料,根据是否并发IAH分为IAH组(153例)和对照组(169例),比较两组患者的临床特征、实验室检查结果。采用多因素logistic逐步回归法分析SAP患者合并IAH的危险因素,采用R软件建立预测SAP并发IAH的Nomogram模型,绘制模型的受试者工作特征曲线(ROC),计算曲线下面积(AUC),评估其预测效能;运用校准图、Hosmer-Lemesshow检验及决策曲线分析分别评估模型的预测准确性和临床应用价值。应用Bootstrap方法对模型进行内部验证。结果IAH组体重指数、CRP、降钙素原(PCT)、WBC、急性生理和慢性健康评估Ⅱ(APACHEⅡ)评分、改良的CT严重指数评分(MCTSI)、并发症(腹腔积液、腹腔感染、胃肠功能障碍、休克、多器官功能障碍综合征)发生率、机械通气及大容量液体复苏(24 h≥4 L)例数均高于对照组,血清白蛋白、血清钙低于对照组,差异均有统计学意义(P值均<0.05)。多因素logistic回归分析结果显示,血清白蛋白(OR=0.815,95%CI 0.710~0.937)、CRP(OR=1.005,95%CI 1.002~1.008)、MCTSI(OR=2.043,95%CI 1.695~2.463)、合并胃肠道功能障碍(OR=4.179,95%CI 2.170~8.049)、大容量液体复苏(OR=4.265,95%CI 2.269~8.015)为SAP并发IAH的独立危险因素。以上述5项因素为参数建立Nomogram预测模型,预测并发IAH的AUC值为0.886,Hosmer-Lemesshow检验结果显示模型预测结果和实际临床观察结果一致性较高(P=0.189)。决策曲线分析结果显示,当模型的预测概率为10%~85%时给患者带来更多的获益。结论成功建立了SAP并发IAH的早期预测模型,该模型能较好地预测SAP并发IAH的风险。

Objective To construct the prediction model of SAP complicated with intra-abdominal hypertension(IAH),and evaluate the prediction efficiency of the model.Methods The clinical data of 322 SAP patients admitted to the emergency department of Cangzhou Hospital of Integrated Chinese and Western Medicine in Hebei Province from January 2017 to December 2021 were retrospectively analyzed.They were divided into IAH group(n=153)and control group(n=169)according to whether they had IAH complications or not.The clinical characteristics and laboratory test results of the two groups were compared.Multifactor logistic step-up regression was used to analyze the risk factors of SAP patients complicated with IAH.A nomogram model for predicting SAP complicated with IAH was established by using R software.The receiver operating characteristic curve(ROC)of the model was plotted,and the area under the curve(AUC)was calculated to evaluate its prediction efficiency.Calibration chart,Hosmer-Lemesshow test and decision curve analysis were used to evaluate the prediction accuracy and clinical application value of the model.The Bootstrap method was applied to verify the model internally.Results In IAH group,cases with body mass index,CRP,procalcitonin(PCT),WBC,acute physiological and chronic health assessmentⅡ(APACHEⅡ)score,modified CT Severity Index score(MCTSI),incidence of complications(abdominal effusion,abdominal infection,gastrointestinal dysfunction,shock,multiple organ dysfunction syndrome),mechanical ventilation,the number of high-volume fluid reactivation(24 h≥4 L)were more than those in control group;serum albumin and serum calcium in IAH group were lower than those in control group,and the differences were statistically significant(all P value<0.05).Multivariate logistic regression analysis showed that serum albumin(OR=0.815,95%CI 0.710-0.937),CRP(OR=1.005,95%CI 1.002-1.008),MCTSI(OR=2.043,95%CI 1.695-2.463),complication of gastrointestinal dysfunction(OR=4.179,95%CI 2.170-8.049),and high-volume fluid resuscitation(OR=4.265,95%CI 2.269-8.015)were independent risk factors for IAH in SAP.The Nomogram prediction model was established using the five factors above as parameters,and the AUC value for predicting IAH complication was 0.886.The Hosmer-Lemesshow test showed a high consistency between the prediction results and the actual clinical observation results(P=0.189).The results of decision curve analysis showed that the prediction probability of the model was between 10%and 85%,which could bring more benefits to patients.Conclusions The early prediction model of SAP with concurrent IAH is successfully established,which can better predict the risk of SAP with concurrent IAH.