摘要目的：比较单用或联用解剖学评分、生理学评分对重症加强治疗病房（ICU）严重创伤患者死亡的预测价值。方法采用回顾性研究方法，选择2011年1月至2014年12月第三军医大学大坪医院、遵义医学院重症医学科收治的严重创伤患者，纳入年龄≥16岁、伤后24 h内入院、ICU住院时间≥48 h、损伤严重程度评分（ISS）≥16分的创伤患者。按预后分为存活组和死亡组。计算患者解剖学评分〔ISS、新的损伤严重程度评分（NISS）〕和生理学评分〔急性生理学与慢性健康状况评分系统Ⅱ（APACHEⅡ）评分〕，评价各评分系统单用或联用对严重创伤患者死亡的预测能力。结果共纳入614例严重创伤患者，死亡153例，病死率为24.9%。死亡组ISS、NISS、APACHEⅡ、ISS+ APACHEⅡ、NISS+ APACHEⅡ评分均明显高于存活组〔ISS （分）：29.15±7.75比24.31±6.50，NISS（分）：41.96±12.01比29.64±8.19，APACHEⅡ（分）：23.71±6.58比17.02±5.49，ISS+ APACHEⅡ（分）：52.86±10.00比41.33±8.70，NISS+ APACHEⅡ（分）：65.67±13.46比46.66±10.43，均P＜0.01〕。ISS、NISS、APACHEⅡ、 ISS+ APACHEⅡ、NISS+ APACHEⅡ预测创伤后死亡的受试者工作特征曲线下面积（AUC）分别为0.687、0.792、0.782、0.809和0.860。ISS+ APACHEⅡ、NISS+APACHEⅡ较单独应用ISS、 NISS或APACHEⅡ有较大的AUC，且NISS+ APACHEⅡ的AUC明显大于ISS+APACHEⅡ的AUC（均P＜0.05）。NISS+ APACHEⅡ预测严重创伤患者死亡的AUC最大，最佳诊断阈值为56分时，敏感度为75.2%，特异度为82.0%，阳性预测值为58.1%，阴性预测值为90.9%，阳性似然比为4.17，阴性似然比为0.30，约登指数为0.572。结论解剖学评分联合生理学评分较其单独使用对严重创伤患者死亡具有更好的预测价值，为临床筛选高危创伤患者并进行早期干预提供了新的方法。

abstractsObjectiveTo compare the predictive value of anatomic scoring system, physiological scoring system, and the combination of two systems in death prediction of patients with severe trauma in intensive care unit (ICU). Methods A retrospective analysis of patients with severe trauma admitted to department of critical care medicine of Daping Hospital, the Third Military Medical University, and Zunyi Medical University from January 2011 to December 2014 was conducted. The patients meeting the following criteria were enrolled: over 16 years old, admitted to hospital shorter than 24 hours after trauma, length of ICU stay≥48 hours, and injury severity score (ISS)≥16. Patients were divided into two groups: survivors and non-survivors. The data of anatomic scoring system, including ISS and new injury severity score (NISS), and physiological scoring system, including acute physiology and chronic health evaluationⅡ(APACHEⅡ) score were collected. The predictive power for death of the scoring system alone or combination in patients with severe trauma was evaluated.Results A total of 614 patients with severe trauma were enrolled, and there were 153 deaths with a mortality rate of 24.9%. ISS, NISS, APACHEⅡ, ISS+ APACHEⅡ, NISS+ APACHEⅡ of non-survivors were significantly higher than those of survivors (ISS: 29.15±7.75 vs. 24.31±6.50, NISS: 41.96±12.01&nbsp;vs. 29.64±8.19, APACHEⅡ: 23.71±6.58 vs. 17.02±5.49, ISS+ APACHEⅡ: 52.86±10.00 vs. 41.33±8.70, NISS+ APACHEⅡ: 65.67±13.46 vs. 46.66±10.43, allP< 0.01). The area under receiver operating characteristic curve (AUC) of ISS, NISS, APACHEⅡ, ISS+ APACHEⅡ, NISS+ APACHEⅡ was 0.687, 0.792, 0.782, 0.809, and 0.860, respectively. Both of ISS+ APACHEⅡ and NISS+ APACHEⅡ had higher AUC than that of ISS, NISS or APACHEⅡ alone; and the AUC of NISS+ APACHEⅡ was significantly larger than that of ISS+ APACHEⅡ(allP< 0.05). NISS+ APACHEⅡ showed the largest AUC in death prediction of severe trauma patients. The cut-off value, sensitivity, specificity, positive predict value (+PV), negative predict value (-PV), positive likelihood ratio (+LR), negative likelihood ratio (-LR), and Youden index of NISS+ APACHEⅡ, which had the greatest AUC, were 56, 75.2%, 82.0%, 58.1%, 90.9%, 4.17, 0.30, and 0.572, respectively.Conclusion The combination of anatomic scoring system and physiological scoring system is better than single scoring system for death prediction in patients with severe trauma in ICU, and it may be considered to be a new method for early identification of death risk in patients with severe trauma.