摘要
目的 探讨英国国家早期预警评分（national early warning score, NEWS）、改良早期预警评分（modified early warning score，MEWS）和急诊脓毒症死亡风险评分（mortality in emergency department sepsis score，MEDS）对急诊感染患者预后的评估价值，并探究新的评分方法。 方法 回顾性分析2016-01～2016-08就诊于清华大学附属北京清华长庚医院急诊科的215例感染患者，记录性别、年龄、既往基础疾病、就诊时生命体征、感染部位及相关实验室检验，分别进行NEWS、MEWS和MEDS，以进入研究后28 d生存情况分为死亡组与存活组，分析两组间各评分差异。为探究新的评分方法，进一步应用Logistic回归分析评估各因素与28 d预后的关系，并最终获得拟合方程。应用受试者工作特征曲线（ROC曲线）比较各评分系统及拟合方程对28 d预后的预测能力。结果 215例急性感染患者28 d 病死率为14.88%。死亡组NEWS、MEWS和MEDS均高于存活组。单因素Logistic回归分析显示，年龄、恶性肿瘤病史、心率、呼吸频率、收缩压、血氧饱和度（SpO2）、血小板、红细胞压积、血肌酐（serum creatinine, Scr）、肾小球滤过率（estimated glomerular filtration rate, eGFR）是28 d死亡的预测因素（P＜0.05）。NEWS、MEWS、MEDS和联合多变量建立的拟合方程对28 d死亡预测的ROC曲线下面积分别为0.881、0.757、0.935和0.954。NEWS与MEDS比较差异无统计学意义（P＞0.05），与MEWS比较差异有统计学意义（P＜0.01）。联合多变量后建立的拟合方程敏感度最佳，曲线下面积最大，优于MEWS（P＜0.01）及NEWS（P＜0.05）。结论 MEDS的预测能力同NEWS能力相当，优于MEWS。联合MEDS与心率、Scr获得的拟合方程的预测能力更优于NEWS和MEWS。
Objective To investigate the value of national early warning score （NEWS）, modified early warning score （MEWS） and mortality in emergency department sepsis score （MEDS） in predicting the outcome of patients with infection in emergency department （ED）, and to explore a new scoring method. Methods A total of 215 patients with infection treated in the ED of Beijing Tsing-hua Changgung Hospital from January 2016 to August 2016 were enrolled retrospectively. Gender, age, comorbidities, vital signs, infection sites and laboratory testing results were collected. The 28-day outcome of survival or death was considered as the end point of the study. NEWS, MEWS and MEDS were compared respectively between non-survivors and survivors. Logistic regression analysis was used to determine the independent predictors for 28-day mortality, then was fitted to equation. Area under the receiver operating characteristics （ROC） curve was used to evaluate the value of these scores and equation in predicting 28-day death. Results Among these 215 patients, the mortality rate within 28 days after admission was 14.88%. It was identified by Logistic regression that age, history of malignant tumor, heart rate（HR）, respiration rate, systolic blood pressure, SpO2, platelet count, hematocrit, serum creatinine, estimated glomerular filtration rate （eGFR） were significantly independent factors in 28-day death prediction （P＜0.05）. Area under the ROC curve for 28-day mortality were 0.881, 0.757, 0.935, 0.954 for NEWS, MEWS, MEDS and equation （P＜0.05）. The comparison between NEWS and MEDS had not shown statistical significance （P〉0.05）, but the significance was illustrated when compared NEWS with MEWS （P＜0.01）. To explore a new scoring method, equation was deduced with HR and serum creatinine and was shown to have the best sensitivity, and its area under the ROC curve was greater than MEWS （P＜0.01） and NEWS （P＜0.05）. Conclusion MEDS is a good prognosis predictive tool for emergency patients with infection, its prognostic performance is as good as NEWS, but better than MEWS. Combined MEDS score with HR and serum creatinine gives the greatest prognostic power in comparison with NEWS and MEWS scores.
作者
宋琳琳
王琰
张向阳
冯莉莉
王逸群
龚晓杰
陈心培
赵心亮
陈旭岩
Song Lin-lin;Wang Yan;Zhang Xiang-yang;Feng Li-li;Wang Yi-qun;Gong Xiao-jie;Chen Xin-pei;Zhao Xin-liang;Chen Xu-yan(Department of CriticalEmergency, Beijing Tsing-hua Changgung Hospital, Beijing 102218, China)
出处
《中国急救医学》
CAS
CSCD
北大核心
2018年第6期465-470,共6页
Chinese Journal of Critical Care Medicine
