Application of the mortality in emergency department sepsis score for emergency department patients with sepsisi / 中华急诊医学杂志

ABSTRACT

ObjectiveTo assess the sepsis score used for detecting the mortality of patients with sepsis in emergency department, and to compare with APACHE Ⅱ score, simplified acute physiology Ⅱ score ( SAPS Ⅱ ) and modified early warning score (MEWS) in terms of 28-day mortality of patients. Methods A total of 613 patients with sepsis were enrolled from the emergency department for a prospective study from September 2009 to September 2010. The sepsis score, APACHE Ⅱ score, SAPS Ⅱ score and MEWS score all were recorded and compared. The patients with sepsis were followed up for 28 days. Based on the sepsis score, patients with sepsis were stratified into 5 mortality risk groups, namely very low risk group (0～4 points), low risk group (5 ～7 paints), moderate risk group (8 ～ 12 points), high risk group ( 13 ～ 15points) and very high risk group (more than 15 points). The actual mortality rates were compared among all 5 groups by using Chi square test. Then, comparison between survivors and non-survivors carried out with logistic regression analysis to determine the independent risk factors of mortality.Receiver operating characteristic curve (ROC curve) was used to compare the sepsis score with APACHE Ⅱ score, SAPS Ⅱscore and MEWS in respect of the prognosis validity. ResultsTen patients were out of the follow-up and the data of 603 patients followed up were completely documented. The actual mortality rates of 5 risk groups were 0％, 7.7％, 18.5％, 46. 7％ and 63％, respectively. There were significant differences in age and four scoring systems between survivors ( n = 440) and non-survivors ( n = 163 ) ( P ＜ 0. 01 ). Sepsis score,APACHE Ⅱ score, SAPS Ⅱ and MEWS all were valid and eligible for detecting the risk of mortality in patients with sepsis. The ROC areas under the curve (AUC) of these 4 scoring systems were 0. 767, 0. 743,0. 741 and 0. 636, respectively. ConclusionsThe sepsis score can be used to stratify patients with sepsis according to mortality risk with better sensitivity to predict 28-day mortality. It is rational for evaluation in prediction of patients with sepsis in Emergency Department.