摘要
目的评估快速序贯器官衰竭评分(quick Sequential Organ Failure Assessment,qSOFA)用于成人脓毒症休克早期预测的可行性,探讨进一步提高其预测效力的可能途径。方法纳入2015年7月—2016年6月于急诊科就诊的成人脓毒症患者。根据患者72 h内是否发生休克分为休克组和非休克组。采用多因素logistic回归分析筛选脓毒症休克发生的独立危险因素,并采用受试者工作特征(receiver operating characteristic,ROC)曲线分析分别对独立危险因素、qSOFA以及其他常用评分系统[改良早期预警评分(Modified Early Warning Score,MEWS)、急诊科脓毒症相关病死率评分(Mortality in Emergency Department Sepsis Score,MEDS)、序贯器官衰竭评分(Sequential Organ Failure Assessment,SOFA)、急性生理与慢性健康(Acute Physiology and Chronic Health Evaluation,APACHE)Ⅱ评分]对脓毒症休克的评估能力进行比较。通过分析qSOFA与筛选出的危险因素的关系,探讨可能提高其预测效力的途径。结果共纳入821例患者,其中休克组108例,非休克组713例。多因素logistic回归分析结果表明,呼吸频率、收缩压、pH值、氧合指数、血清乳酸、白蛋白、格拉斯哥昏迷评分和降钙素原均为脓毒症休克的独立危险因素(P<0.05)。ROC曲线分析发现,pH值、乳酸和降钙素原的曲线下面积(area under curve,AUC)较高,分别为0.695、0.678和0.694,且乳酸的特异度(0.868)、阳性预测值(0.356)和阳性似然比(3.644)最高,而降钙素原的灵敏度(0.889)和阴性预测值(0.961)最高。将患者的MEWS、MEDS、SOFA、APACHEⅡ及qSOFA的评分情况进行ROC曲线分析,SOFA的AUC、灵敏度(0.833)和特异度(0.835)、阳性预测值(0.435)、阴性预测值(0.971)和阳性似然比最高(5.048),而MEWS的阴性似然比最高(0.581)。qSOFA评分的上述指标均非最优。结论 qSOFA评分对成人脓毒症休克的早期预测能力有限,若将pH值、乳酸和降钙素原与之联合,可能进一步提高qSOFA的预测效力。
Objective To evaluate the predicting effect of quick Sequential Organ Failure Assessment (qSOFA) on septic shock, and investigate the probability of improving the predicting effect. Methods Patients with sepsis diagnosed in Emergency Department from July 2015 to June 2016 were enrolled. They were divided into shock group and non-shock group based on whether or not they had septic shock during 72 hours after admission. The multivariate logistic regression analysis was used to find out the independent risk factors affecting the incidence of septic shock. Receiver operating characteristic (ROC) curve was used to analyze those risk factors. Modified Early Warning Score (MEWS), Mortality in Emergency Department Sepsis Score (MEDS), Sequential Organ Failure Assessment (SOFA), Acute Physiology and Chronic Health Evaluation (APACHE) Ⅱ and qSOFA were also compared with ROC curve analysis. The possibility of improvement of qSOFA predicting effect was discussed. Results A total of 821 patients were enrolled, with 108 in septic shock group and 713 in non-septic shock. The result of multivariate logistic regression analysis indicated that respiratory rate, systolic blood pressure, pH value, oxygenation index, lactate, albumin, Glasgow Coma Score and procalcitonin were the independent risk factors (P〈0.05). The result of ROC analysis showed that the area under curve (AUC) ofpH value, lactate and procalcitonin was 0.695, 0.678 and 0.694, respectively. Lactate had the highest value of specificity (0.868), positive predictive value (0.356) and positive likelihood ratio (3.644), while the sensitivity (0.889) and negative predictive value (0.961) of procalcitonin were the highest. MEWS, MEDS, SOFA, APACHE Ⅱ and qSOFA were compared with ROC. SOFA had the best predicting effect with the statistical results of AUC (0.833), sensitivity (0.835), specificity (0.435), positive predictive value (0.971), negative predictive value (0.971), and positive likelihood ratio (5.048); and MEWS had the highest negative likelihood ratio (0.581). qSOFA did not show a best predicting value. Conclusion qSOFA is not the best choice to predict the possibility of septic shock, but its predicting value might be improved when combined with pH value, lactate and procalcitonin.
出处
《华西医学》
CAS
2017年第6期812-818,共7页
West China Medical Journal
基金
国家卫生和计划生育委员会公益性行业科研专项(201302003)