Objectives: Early estimation of the mortality risk of severely injured patients is mandatory. To estimate the seriousness of the condition of patients with trauma, we developed the emergency trauma score (EMTRAS) for ease of use, with simple parameters that are available within 30 minutes.
Design: Prospective analysis of the German Trauma Registry of multitrauma patients.
Setting: EMTRAS was derived from data from 1993 through 2003. Potential parameters that were prognostic for mortality in univariate analysis were evaluated by multivariate binary logistic regression. Selected parameters were then assigned a subscore that varied from 0 to 3. The EMTRAS score was a simple addition of these subscores. EMTRAS was compared with other scores’ receiver operating characteristic curves. After completion, EMTRAS was validated in patients from 2004 and 2005.
Patients: A total of 11,533 patients were to be used for developing the score and 3314 patients for validating it.
Main Results: The strongest predictors of mortality were age, prehospital Glasgow Coma Scale, base excess (mmol/L), and prothrombin time (% of reference). These parameters were categorized in subscores of 0 through 3. Age: <40, 40 through 60, 61 through 75, and >75 scored 0, 1, 2, and 3, respectively. Glasgow Coma Scale: 13 through 15, 10 through 12, 6 through 9, and 3 through 5 scored 0, 1, 2, and 3, respectively. Base excess: >−1, −5 through −1, −10 through −5.1, and <−10 scored 0, 1, 2, and 3, respectively. Prothrombin time: <80%, 80% through 50%, 49% through 20%, and >20% received a score of 0, 1, 2, and 3, respectively. In the validation dataset, the area under the receiver operating characteristic curve for EMTRAS was 0.828.
Conclusions: EMTRAS combines four early parameters from the emergency room and accurately predicts mortality. Knowledge of the anatomical injuries is not necessary. The determination of the EMTRAS will inform caregivers of the seriousness of patients with trauma at an early stage.
From the Department of Traumatology (MRR, MWNN, MV, FS, HJtD), University Medical Center Groningen, University of Groningen, The Netherlands; Institutes for Research in Operative Medicine (RF, EAMN), University of Witten/Herdecke, Germany; and Department of Trauma and Orthopaedic Surgery (BB, DR), University of Witten/Herdecke at the Cologne Merheim Medical Center, Germany.
Supported by the Deutsche Forschungsgemeinschaft (NE 385/5–1 and 2, Sa 862/1–1) and Novo Nordisk, Copenhagen, Denmark.
The authors do not have any potential conflicts of interest to disclose.
For information regarding this article, E-mail: firstname.lastname@example.org