Comparison of SAPS 3-PIRO score with other four scoring systems for assessing sepsis / 中华急诊医学杂志

ABSTRACT

Objective To investigate the feasibility of SAPS 3-PIRO score for prediction of outcomes in severe sepsis patients in the intensive care units (ICU).The optimal scoring system was also evaluated in this study.Methods Data of 677 patients with sepsis,treated in ICU of the emergency department at the Beijing Chaoyang Hospital between January 2008 and December 2011,were analyzed.The inclusion criteria were:age ≥18 years and the diagnostic criteria of severe sepsis and organ damage recommended by the International Sepsis Definitions Conference in Washington DC,USA in 2001,as well as the criteria developed by the Conference as a reference.The patients,who did not meet the criteria were excluded from the study.Age,gender,underlying disease,main infection site,and duration of ventilator support of all cases were recorded.Vital signs,hematological findings (gases,biochemistry,coagulation) and causative pathogens and organ damage on day-1 in the ICU were also recorded.According to the data of the most unfavorable findings in the ICU on day-1,scores got from APACHE Ⅱ,SOFA (sequential organ failure assessment),SAPS Ⅲ(simplified acute physiological score Ⅲ),SAPS 3-PIRO (predisposition,infection,response,organ failure/dysfunction) and MEDS (mortality in emergency department sepsis score) scoring systems were calculated.All patients were followed up for 28 days.The outcome of survival and death at 28 days were the endpoints of this study.SPSS V13.0 (SPSS,Chicago,IL,USA) was used to analyze data.Continuous variables with normal distribution were expressed as mean ± standard deviation.The independentsample t test was applied when normality (and homogeneity of variance) assumptions were fulfilled otherwise the t' test would be used.The distribution of categorical variables in two groups was tested using the x2 test.Binary logistic regression models were also used according to the survival status.The values of scoring system in predicting outcomes and the determination of dividing value in predicting death were analyzed using receiver operating characteristic (ROC) curves.The area under the receiver operating characteristic (AUROC) curves was compared using Z test.P values less than 0.05 were considered statistically significant.Results The total 28-day mortality was 41.9％.APACHE Ⅱ,SOFA,SAPS Ⅲ,SAPS 3-PIRO and MEDS scoring systems were used to find the independent predictive factors of death within 28 hours in patients with severe sepsis.By using ROC curve and AUROC curve to compare the validity of these scoring systems,SAPS 3-PIRO score had comparable predictive capability in comparison with APACHE Ⅱ,SOFA and SAPS Ⅲ scores,and MEDS was superior to SAPS 3-PIRO,APACHE Ⅱ,SOFA and SAPS Ⅲ score.Conclusions For patients with severe sepsis admitted in ICU,MEDS was superior to SAPS 3-PIRO,APACHE Ⅱ,SOFA and SAPS Ⅲ score in predicting prognosis.MEDS has favorable capability in predicting death within 28 days.MEDS ≥ 11 was the sign of increased mortality.SAPS 3-PIRO score has comparable predictive capability in comparison with APACHE Ⅱ,SOFA and SAPS Ⅲ score and it may be suggested for clinical practice.