Combined anatomic and physiologic scoring systems for predicting in-hospital mortality in ICU patients with severe trauma: A multicenter observational cohort study

• Main Authors: Xiao-Yuan Ma, Huai-Jian Jin, Li-Xing Tian, Qian Wang, Jun-Yu Zhu, Zhi-Gao He, Wen Tao, Tao Chen, Bin Wang, Hua-Ping Liang
• Format: Article
• Language: English
• Published: Wolters Kluwer Medknow Publications 2019-01-01
• Series: Asian Pacific Journal of Tropical Medicine
• Subjects: new injury severity score; acute physiology and chronic health evaluation ii; glasgow coma scale; in-hospital mortality; predictive value; severe trauma; icu
• Online Access: http://www.apjtm.org/article.asp?issn=1995-7645;year=2019;volume=12;issue=14;spage=17;epage=24;aulast=Ma
• ISSN: 2352-4146

Objective: To evaluate the ability of new injury severity score (NISS), acute physiology and chronic health evaluation II (APACHE II), Glasgow coma scale (GCS), a combination of NISS and GCS, a combination of APACHE II and GCS, a combination of NISS and APACHE II to predict all-cause mortality of patients with severe trauma in mainland China. Methods: This was a multicenter observational cohort study conducted in the ICU of the Chonggang General Hospital, Daping Hospital of the Army Medical University and Affiliated Hospital of Zunyi Medical College from January 2012 to August 2016. The score of NISS, APACHE II, GCS, a combination of NISS and GCS, a combination of APACHE II and GCS, a combination of NISS and APACHE II were calculated based on data from the first 24 hours of ICU admission. Data were processed with Student’s t-test, chi-square test, and receiver operating characteristic (ROC) curve of six scoring systems. Calibration was assessed with the Hosmer-Lemeshow test. The primary endpoint was death from any cause during ICU stay. Results: A total of 852 and 238 patients with severe trauma were assigned to the derivation group and validation group, respectively. Area under the ROC curve (AUC) was 0.826 [95% confidence interval (CI)=0.794-0.855)] for NISS, 0.802 (95% CI=0.768-0.832) for APACHE II, 0.808 (95% CI=0.774-0.838) for NGCS, 0.859 (95% CI=0.829 -0.886) for NISS+NGCS, 0.864 (95% CI=0.835-0.890) for APACHE II +NGCS, 0.896 (95% CI=0.869-0.929) for NISS+APACHE II in the derivation cohort. Similarly, the score of NISS+APACHE II was also better than the other five scores in the validation cohort (AUC=0.782; 95% CI=0.725-0.833) and had a good calibration (P=0.41). Conclusions: Taking into account anatomical and physiological parameters completely, the combination of NISS and APACHE II performs better than NISS, APACHE II , NGCS, NISS+NGCS, APACHE II +NGCS for predicting mortality in ICU severe trauma patients. It is needful to develop models that contain various types of accessible predictors (demographic variables, injury cause/mechanism, physiological and anatomical variables, etc.) as comprehensive as possible.