A validation study of national early warning score in evaluation of death risk in elderly patients with critical illness

Abstract

To verify the validity and feasibility of national early warning score (NEWS) in evaluation of death risk in elderly patients with critical illness,in order to find out which scoring method is more suitable for elderly critical illness patients. A prospective case-control study was conducted. The critical illness patients aged over 60 years old with the length of hospital stay over 24 hours, and admitted to Department of Emergency of Qingdao Municipal Hospital from January to December 2015 were enrolled. The clinical data including in emergency and the actual outcome of patients were collected, and the patients were divided into death group and survival group according to 30-day outcome. Patients in the two groups were assessed by using NEWS and risk classification according to the first results of vital signs monitoring. Multivariate logistic regression model was used to analyze the relationship between the NEWS classification and the risk of death in elderly critical ill patients. 1 950 emergency elderly patients with critical illness were enrolled, with 78 cases (4.0％) dead within 30 days and 1 872 survived (96.0％).Compared with the survival group, patients in death group were older (years:79.8 ± 10.8 vs.75.3 ± 8.9,t =4.335,P ＜0.001),and had higher acute physiology and chronic health evaluation Ⅱ (APACHE Ⅱ) score (22.9± 4.6 vs.18.2 ± 4.8,t =8.487,P ＜ 0.001),lower Glasgow coma scale (GCS) score (12.2 ± 4.5 vs.13.4 ± 5.2,t =-2.007,P =0.045),higher incidence of respiratory system diseases (29.5％ vs.17.9％,x 2 =12.742,P =0.013),higher NEWS score (11.2 ± 5.5 vs.3.9 ± 2.7,t =22.063,P ＜ 0.001),as well as higher proportion of patients with NEWS classification of high risk and very high risk (65.4％ vs.15.8％,x 2 =263.125,P ＜ 0.001).With the increase of NEWS risk classification, mortality rate was also increased, and the mortality rate in the patients with low, medium, high and very high risk were 0.81％ (9/1 108),3.63％ (18/496),5.83％ (13/223),30.89％ (38/123),respectively, with statistically significant difference (x 2 =179.741,P ＜ 0.001).It was showed by logistic regression analysis that the NEWS score of elderly patients with critical illness were positively correlated with 30-day death. The 30-day death risk of patients with middle risk, high risk and very high risk was 4.600,9.052 and 54.598 folds of the patients with low risk respectively. NEWS score can be used to assess the risk of death in emergency elderly patients with critical illness. NEWS risk classification can quantify and classify the risk of death in the elderly patients with critical illness.