急性胃肠损伤分级联合序贯器官衰竭评估对严重脓毒症预后评估的价值

Value of Acute Gastrointestinal Injury Grading Combined with Sequential Organ Failure Assessment Scoring Systems in Prognosis Assessment of Severe Sepsis

目的探讨急性胃肠损伤(AGI)分级联合序贯器官衰竭评估(SOFA)的新评分系统对严重脓毒症预后的评估价值。方法选择2012年7月—2013年7月浙江省舟山医院重症医学科收治的严重脓毒症患者46例,根据其入院4周内存活或死亡分为存活组(n=26)和死亡组(n=20)。患者入ICU后第1周采用AGI诊断标准每日进行AGI评估,并赋予相应分值,同时进行SOFA评分、急性生理学与慢性健康状况评分系统Ⅱ(APACHEⅡ)评分和多器官功能障碍综合征(MODS)(2004)评分,计算改良SOFA评分(AGI评分与SOFA评分相加);采用受试者工作特征(ROC)曲线和Hosmer-Lemeshow拟合优度检验分析改良SOFA评分系统和其他4种评分系统对严重脓毒症预后评估的价值,并采用多因素Logistic回归分析5种评分系统预测预后的价值。结果存活组入ICU后第2、3、5、7天以及最高改良SOFA评分、AGI评分、SOFA评分、MODS(2004)评分均低于死亡组(P<0.05);存活组入ICU后第1、2、3、5、7天以及最高APACHEⅡ评分均低于死亡组(P<0.05)。第1、2、3、5、7天以及最高改良SOFA评分预测预后的ROC曲线下面积(AUC)分别为0.60、0.84、0.92、0.93、0.97、0.86,AGI评分AUC分别为0.56、0.84、0.95、0.96、0.99、0.85,SOFA评分AUC分别为0.59、0.82、0.87、0.86、0.92、0.83,APACHEⅡ评分AUC分别为0.74、0.88、0.92、0.96、0.97、0.83,MODS(2004)评分AUC分别为0.61、0.85、0.87、0.87、0.92、0.80。Hosmer-Lemeshow拟合优度检验结果显示,5种评分系统的预测与实际结果间比较,差异均无统计学意义(P>0.05),拟合度较好。多因素Logistic回归分析结果显示,改良SOFA评分≥8.1分(OR=3.167)、AGI评分≥1.3分(OR=6.000)、SOFA评分≥7.7分(OR=3.600)、APACHEⅡ评分≥11.8分(OR=6.333)、MODS(2004)评分≥8.6分(OR=5.000)是严重脓毒症患者预后不良的危险因素(P<0.05)。结论改良SOFA评分可以提高SOFA评分对严重脓毒症预后评估的价值。改良SOFA评分≥8.1分对严重脓毒症患者死亡风险有一定的预测作用。

Objective To evaluate the value of acute gastrointestinal injury （AGi） grading combined with sequential organ failure assessment （SOFA） scoring systems in prognosis assessment of severe sepsis. Methods Forty - six patients with severe sepsis admitted in the Department of Intensive Care Unit, Zhoushan Hospital from July 2012 to July 2013 were selected as the subjects and divided into survival group （n = 26） and death group （n = 20） based on they survived or died within 4 weeks after admissionl In the 1 st week after admission, the AGI of enrolled patients was evaluated daily according to the diagnostic criteria of AGI and scored. At the same time, SOFA scoring, Acute Physiology and Chronic Health Evaluation Ⅱ （APACHE Ⅱ ） scoring and MODS （2004） scoring and modified SOFA scoring （adding AGI and SOFA scores together） were performed. Receiveroperating characteristic （ROC） curve and Hosmer- Lemeshow goodness -of- fit test were used to analyze the resolution and calibration of modified SOFA scoring system and other 4 scoring systems in prognosis assessment of severe sepsis. Multivariate Logistic regression model was used to evaluate the values of these systems in prognosis. Results Compared with the death group, the survival group had lower modified SOFA, AGI, SOFA and MODS （2004） scores on the 2nd, 3rd, 5th and 7th days after admission （P 〈 0. 05）, as well as lower highest scores of modified SOFA, AGI, SOFA and MODS （2004） during this period （P 〈0. 05）. The daily and highest APACHEⅡscores on the 1st, 2nd, 3rd, 5th and 7th days after admission in the survival group were lower than those in the death group, respectively （P 〈 0. 05）. The AUC values of modified SOFA scoring in daily scoring and the highest scoring for prognosis assessment on the 1st, 2nd, 3rd, 5th and 7th days after admission were 0. 60, 0. 84, 0. 92, 0. 93, 0. 97, 0. 86. AUC values of AGI scoring were 0. 56, 0.84, 0. 95,0O. 96, 0. 99, 0. 85. AUC values of SOFA scoring were 0. 59, 0. 82, 0. 87, 0. 86, 0. 92, 0. 83, respectively. AUC values of APACHE Ⅱ scoring were 0. 74, 0. 88, 0. 92, 0. 96, 0. 97, 0. 83, respectively. AUC values of MODS （2004） scoring were 0. 61 , 0. 85, 0.87, 0. 87, 0. 92, 0. 80, respectively. Hosmer - Lemeshow goodness - of - fit test showed that the results of prognosis assessment of the 5 scoring systems did not differ significantly from the clinical outcomes of the patients （ all P 〉 0.05） , which demonstrated that the goodness - of - fit of these 5 scoring systems was relatively good. Results of the analysis with a multivariate Logistic regression model indicated that modified SOFA scoring ≥ 8. 1 （ OR = 3. 167 ）, AGI seoring ≥ 1.3 （ OR = 6. 000）, SOFA scoring ≥7.7 （OR =3.600）, APACHE Ⅱ scoring≥11.8 （OR=6.333）, MODS （2004） scoring≥ 8.6 （OR = 5. 000） were independent predictors of poor outcome of patients with severe sepsis （ P 〈 0. 05 ）. Conclusiom The resolution and calibration of SOFA scoring in the prognosis assessment of severe sepsis can be improved through modified SOFA scoring. Modified SOFA scoring≥8.1 is a predictor for death in patients with severe sepsis.