脓毒症相关凝血-炎症评分的构建及其对脓毒症预后评估效能分析

Construction of Sepsis-Related Coagulo-Inflammatory Score and it's prognostic effects on sepsis

目的基于脓毒症凝血、炎症指标创建脓毒症相关凝血-炎症评分(Sepsis-Related CoaguloInflammatory Score,SRCIS),并评估其对脓毒症28 d病死率的预测价值。方法采用前瞻性巢式病例对照研究,按脓毒症Sepsis 3.0标准纳入于2017年9月—2018年1月在四川大学华西医院急诊科就诊的脓毒症患者,通过多因素分析筛选影响脓毒症28 d病死率的凝血、炎症标志物,基于凝血、炎症的独立危险因素创建SRCIS,验证该SRCIS评分对脓毒症预后的评估效能。结果纳入合格病例123例,其中28 d死亡17例,病死率为13.8%。生存组和死亡组的人口学特征和基础疾病等基线差异无统计学意义(P>0.05)。多因素分析显示,活化部分凝血活酶时间(activated partial thromboplastin time,APTT)[比值比(odds ratio,OR)=1.015,95%置信区间(confidence interval,CI)(1.017,1.189),P=0.017]和C反应蛋白(C-reactive protein,CRP)[OR=1.100,95%CI(1.006,1.025),P=0.002]为影响脓毒症28 d病死率的独立凝血-炎症危险因素。APTT和CRP预测脓毒症28 d病死率的最佳截断值分别为39.25 s和198.05 mg/L,曲线下面积(area under the curve,AUC)分别为0.618和0.671。APTT<39.25 s时病死率为8.79%,APTT≥39.25 s时病死率达28.13%,差异有统计学意义(P<0.05);CRP<198.05 mg/L时病死率为8.89%,CRP≥198.05 mg/L病死率为27.27%,差异也有统计学意义(P<0.05)。整合APTT和CRP,基于其各自的最佳截断值综合赋分,创建SRCIS评分。SRCIS评分预测脓毒症28 d病死率的AUC为0.707,优于传统的序贯器官衰竭评分(Sequential Organ Failure Assessment,SOFA)(AUC=0.681)和快速序贯器官衰竭评分(quick Sequential Organ Failure Assessment,qSOFA)(AUC=0.695)等脓毒症评估工具。当SRCIS评分为0、1、2时,对应的脓毒症28 d病死率分别为6.94%、16.22%和42.86%,有很高的区分度。结论 APTT和CRP分别为影响脓毒症28 d病死率的独立凝血和炎症因素,相较传统的SOFA、qSOFA等工具,基于APTT和CRP创建的SRCIS能更好地预测脓毒症28 d病死率。

Objective To build a score with the coagulation, inflammation indexes of sepsis patients, named Sepsis-Related Coagulo-Inflammatory Score （SRCIS）, and then evaluate the prognostic capability of it in predicting the 28-day mortality of septic patients after the diagnosis. Methods In this prospective nested case-control study, we recruited septic patients according to the Sepsis 3.0 standards, who visited the Emergency Department, West China Hospital of Sichuan University from September 2017 to January 2018. Multiple factor analysis was conducted to confirm which coagulation or inflammation biomarkers were independent risk factors related to the 28-day mortality after their diagnosis. After that, the SRCIS was built based on those independent risk factors. Finally, receiver operating characteristic curve （ROC） analysis was conducted to verify its prognostic capability for the 28-day mortality of septic patients. Results A total of 123 cases were included. Among them, 17 patients died within 28 days, and the mortality rate was 13.8%. There were no significant differences in the demographic characteristics or comorbidities between the survival group and dead group （P〉0.05）. Multivariate logistic analysis showed that both activated partial thromboplastin time （APTT） [odds ratio （OR）=1.015, 95% confidence interval （CI） （1.017, 1.189）, P=0.017] and C-reactive protein （CRP） [OR=1.100, 95%CI （1.006, 1.025）, P=0.002] were independent risk factors for predicting the 28-day mortality of septic patients. ROC analysis indicated that the cut-off values of APTT and CRP predicting the 28-day mortality rate of sepsis were 39.25 seconds and 198.05 mg/L, respectively, and the areas under the curve （AUC） of them were 0.618 and 0.671, respectively. The results indicated that the mortality increased from 8.79% to 28.13%, when APTT prolonged to no less than 39.25 seconds （P〈0.05）. The mortality also increased from 8.89% to 27.27% when CRP elevated to no less than 198.05 mg/L （P〈0.05）. The AUC of SRCIS in predicting the 28-day mortality of patients with sepsis was 0.707, which was better than that of Sequential Organ Failure Assessment （SOFA） （AUC=0.681） and quick Sequential Organ Failure Assessment （qSOFA） （AUC=0.695）. The corresponding 28-day mortality rates for patients with sepsis were 6.94%, 16.22%, and 42.86% （P〈0.05）, respectively, when the SRCIS score were 0, 1, and 2. Conclusions APTT and CRP are independent risk factors in predicting the 28-day mortality of patients with sepsis. Compared with traditional scoring systems such as SOFA and qSOFA, SRCIS performances better in predicting the 28-day mortality for patients with sepsis.