Accuracy of five-level adult emergency triage system on critically ill patients: a retrospective observation study in real world
Shanlin Mao,Jun Cao,Lilin Gong,Jiangwei Shao,Zhijie Xia
DOI: https://doi.org/10.3760/cma.j.issn.2095-4352.2016.09.012
2016-01-01
Abstract:Objective To observe the accuracy of the five-level adult emergency triage system (FLAETS) in determining the severity of critically ill patients,and to analyze the factors that influenced the accuracy of the triage.Methods The triage information of patients admitted to Huashan North Hospital Affiliated to Fudan University between 1 year before and 1 year (started in July 2014) after the implementation of the FLAETS were retrospective analyzed.The patients who triaged by FLAETS (from September 2014 to August 2015) were served as observational group,and those manual triaged by modified early warning score (MEWS,from July 2013 to June 2014) were set as control group.The patients with triage of Ⅰ,Ⅱ,and Ⅲ levels were enrolled (fatal,critical,urgent),and the triage results of emergency physicians-in-chief on duty were set as "gold standard".The gender,age,triage level,the proportion of over-triage,the proportion of under-triage,and mortality were compared between two groups.The reasons for the overtriage and under-triage were analyzed.The accuracy of FLAETS in determining severity of critically ill patients was calculated.Results 18449 patients were enrolled in observational group,and 720,1641 and 16088 patients were triaged as level Ⅰ,level ⅡⅡ,level Ⅲ respectively;17 378 patients were triaged as critically ill patients according to "gold standard",and level Ⅰ,level Ⅱ,level Ⅲ were 637,1476 and 15 265 patients respectively.6 352 patients were enrolled in control group,and level Ⅰ,level Ⅱ,level Ⅲ were 204,771 and 5 377 patients respectively;6002 patients were triaged as critically ill patients according to "gold standard",and level Ⅰ,level Ⅱ,level Ⅲ were 308,836 and 4 858 patients respectively.There were no statistically significant differences in gender and age between two groups (both P ＞ 0.05).Compared with the control group,the proportion of over-triage of level Ⅲ was significantly lowered [8.3％ (1 329/16088) vs.12.5％ (674/5 377),P ＜ 0.01],the proportion of under-triage of level Ⅱ and level Ⅲ was significantly lowered [level Ⅱ:0.6％ (9/1 641) vs.7.0％ (54/771),level Ⅲ:0.4％ (63/16088) vs.4.7％ (254/5 377),both P ＜ 0.01].The success rate of resuscitation in observation group was significantly higher than that of control group [80.70％ (619/767) vs.75.23％ (410/545),P =0.020],and the mortality was significantly lowered [1.11％ (193/17 378) vs.2.35％ (141/6002),P =0.037].Over-triages were mainly found in patients with chronic obstructive pulmonary disease (COPD),hypertension and the main complaint of "chest tightness,shortness of breath",and under-triages were mainly found in patients with the multiple organ dysfunction in the elder,atypical heart disease and pneumothorax.The accuracy rates of level Ⅰ,level Ⅱ and level Ⅲ in observation group (99.37％,97.42％ and 89.58％) were significantly higher than those of the control group (97.51％,92.54％,and 80.16％,all P ＜ 0.01).Conclusion The FLAETS in determining severity of critically ill patients were objective and reliable,which enable the nurse to better handle the relationship of the emergency and the ordinary,the severe and the mild,the priority and the non-priority,which made the patients received timely and effective treatment.