Early diagnosis of traumatic brain injury (TBI) and reliable prediction of outcome are essential for determining treatment strategies and allocating resources. This study introduces the Eppendorf-Cologne Scale (ECS) and evaluated its predictive accuracy for outcome and TBI presence compared with those of the Glasgow Coma Scale (GCS).METHODS
A retrospective cohort analysis of severely injured trauma patients registered in the Trauma Registry of the German Society for Trauma Surgery from 1993 to 2010 was conducted. Only directly admitted patients alive on admission and with complete data on GCS, pupil reactivity, and size were included. The ECS was modeled using pupil reactivity, size, and a modified GCS motor component. The unadjusted predictive role of each component was evaluated using multivariable regression analysis. The predictive power regarding the presence of TBI and outcome of the ECS and the GCS was modeled using area under the receiver operating characteristic (AUROC) curve analyses.RESULTS
A total of 28,305 patients fulfilled the study inclusion criteria. The ECS outmatched the predictive accuracy of the GCS for outcome (AUROC, 0.824; 95% confidence interval [95% CI], 0.817–0.831; and AUROC, 0.811; 95% CI, 0.804–0.818, respectively; rs = 0.887, p < 0.001) and TBI presence (AUROC, 0.813; 95% CI, 0.805–0.822; and AUROC, 0.777; 95% CI, 0.768–0.786, respectively; rs = 0.889, p < 0.001). Patients with TBI were five times more often unconscious at the scene and showed a 3.5-fold increased in-hospital mortality. An ECS score of 8 was associated with a 20-fold higher mortality compared with an ECS score of 0. The ECS differentiates patients with a fourfold higher mortality within the GCS 3 collective.CONCLUSION
The ECS shows a significantly higher accuracy for prediction of outcome and TBI presence compared with the GCS and provides a simple, yet reliable, stratification tool for early decision making.LEVEL OF EVIDENCE
Prognostic study, level III.