128例严重脓毒症患儿院内死亡预后评估方程的建立与评价

目的 建立儿童严重脓毒症死亡预后评估方程,为预后评估和制定救护方案提供科学指导.方法 前瞻性纳入儿童重症监护病房(PICU)收治的严重脓毒症患儿128例,采集性别、年龄、生命体征、实验室指标、小儿危重病例评分(PICS),根据院内生存状态将其分为病死组(56例)与生存组(72例);采用Cox比例风险回归建立死亡预后评估方程,采用受试者工作特征曲线(ROC)评价方程的诊断性能.结果 单因素分析病死组与生存组,在纳入的指标中,年龄、血清清蛋白、血小板计数、白细胞计数、病原菌、机械通气时间、小儿危重病例评分(PCIS)、脓毒性休克及多器官功能障碍综合征(MODS)比较,两组差异有统计学意义.入选多元

Derivation and evaluation of prognostic equation to predict in-hospital mortality of 128 children with severe sepsis

Objective To derive a prognostic equation to predict mortality of children with severe sepsis,and provide scientific guidance for prognostic assessment and treatment scheme formulation.Methods 128 children with severe sepsis admitted to pediatric intensive care unit(PICU) were included perspectively,and their gender,age,vital signs,loboratory indicators and pediatric critical illness score(PICS) were collected.They were divided into death group(56 cases) and survival group(72 cases) according to their in-hospital survival condition.Cox proportional-hazards regression was employed to derive the prognostic equation to predict mortality and the receiver operator characteristic(ROC) curve was used to evaluate its diagnostic performance.Results Diffe-rences of indicators included such as age,serum albumin,platelet count,white blood cell count,pathogens,duration of mechanical ventilation,pediatric critical illness score(PCIS),septic shock and multiple organ dysfunction syndrome(MODS) demonstrated statistical significance when univariate analysis was conducted in death group and survival group.Covariates selected for multivariate Cox regression equation were age(HR=0.281,95%CI:0.122~0.643),PCIS(HR=0.222,95%CI:0.099~0.499) and MODS(HR=3.110,95%CI:1.408~6.869).Prognostic equation to predict mortality was:h(t|X)/h0(t)=Exp(-1.271 age-1.506PCIS+1.135MODS).The areas under the ROC curve of prediction in-hospital mortality of children with severe sepsis according to equation and three factors of equation in descending order were equation,PCIS,MODS and age.Conclusion Prognostic equation to predict mortality based upon comprehensive indexes including age,PCIS and MODS has a certain reference value for evaluation of prognosis of children with severe sepsis.