A modified Kampala trauma score (KTS) effectively predicts mortality in trauma patients. Issue 1 (January 2016)

• Record Type: Journal Article
• Authors: Weeks, Sharon R.; Stevens, Kent A.; Haider, Adil H.; Efron, David T.; Haut, Elliott R.; MacKenzie, Ellen J.; Schneider, Eric B.
• Is Part Of: Digestive and liver disease. Volume 48:Issue 1(2016)
• Page Start: 125
• Page End: 129
• Publication Date: 2016-01
• Subjects: Trauma -- Injury -- Injury scores -- Scoring systems; Digestive organs -- Diseases -- Periodicals; Liver -- Diseases -- Periodicals; 616.33005
• DOI: 10.1016/j.injury.2015.07.004
• Languages: English
• ISSNs: 1590-8658

Abstract

Abstract: Background: Mortality prediction in trauma patients has relied upon injury severity scoring tools focused on anatomical injury. This study sought to examine whether an injury severity scoring system which includes physiologic data performs as well as anatomic injury scores in mortality prediction. Methods: Using data collected from 18 Level I trauma centers and 51 non-trauma center hospitals in the US, anatomy based injury severity scores (ISS), new injury severity scores (NISS) were calculated as were scores based on a modified version of the physiology-based Kampala trauma score (KTS). Because pre-hospital intubation, when required, is standard of care in the US, a modified KTS was calculated excluding respiratory rate. The predictive ability of the modified KTS for mortality was compared with the ISS and NISS using receiver operating characteristic (ROC) curves. Results: A total of 4716 individuals were eligible for study. Each of the three scores was a statistically significant predictor of mortality. In this sample, the modified KTS significantly outperformed the ISS (AUC = 0.83, 95% CI 0.81–0.84 vs. 0.77, 95% CI 0.76–0.79, respectively) and demonstrated similar predictive ability compared to the NISS (AUC = 0.83, 95% CI 0.81–0.84 vs. 0.82, 95% CI 0.80–0.83, respectively). Conclusions: The modified KTS may represent a useful tool for assessing trauma mortality risk in real time, as well as in administrative data where physiologic measures are available. Further research is warranted and these findings suggest that the collection of physiologic measures in large databases may improve outcome prediction.