BACKGROUND: This study was performed to compare the efficacies of Injury Severity Score (ISS) and New Injury Severity Score (NISS), and to investigate whether replacing ISS with NISS in Trauma and Injury Severity Score (TRISS) changes the predictive power for mortality. METHODS: We retrospectively analyzed 550 patients aged over 16 years seen in our center over a period of four years who had injuries in at least two organ systems. The ISS, NISS and TRISS were calculated for each patient. TRISS was calculated in two ways: TRISS 1 with age, Revised Trauma Score (RTS) and ISS, and TRISS 2 with age, RTS and NISS. The cut-off values for TRISS 1, TRISS 2, ISS and NISS were determined by ROC analysis. RESULTS: The ideal cut-off values for the prediction of mortality were 21.0 (AUC: 0.907) for ISS and 25.0 (AUC: 0.914) for NISS. There was no statistically significant difference between the ROC curves of ISS and NISS. The ideal cut-off values for predicting mortality were 90 (AUC: 0.934) for TRISS 1 and 86 (AUC: 0.935) for TRISS 2. There was no statistically significant difference between the ROC curves of TRISS 1 and TRISS 2. CONCLUSION: Use of NISS instead of ISS in the TRISS model demonstrated no significant difference, and it can thus be recommended for use.
BACKGROUND: This study was performed to compare the efficacies of Injury Severity Score (ISS) and New Injury Severity Score (NISS), and to investigate whether replacing ISS with NISS in Trauma and Injury Severity Score (TRISS) changes the predictive power for mortality. METHODS: We retrospectively analyzed 550 patients aged over 16 years seen in our center over a period of four years who had injuries in at least two organ systems. The ISS, NISS and TRISS were calculated for each patient. TRISS was calculated in two ways: TRISS 1 with age, Revised Trauma Score (RTS) and ISS, and TRISS 2 with age, RTS and NISS. The cut-off values for TRISS 1, TRISS 2, ISS and NISS were determined by ROC analysis. RESULTS: The ideal cut-off values for the prediction of mortality were 21.0 (AUC: 0.907) for ISS and 25.0 (AUC: 0.914) for NISS. There was no statistically significant difference between the ROC curves of ISS and NISS. The ideal cut-off values for predicting mortality were 90 (AUC: 0.934) for TRISS 1 and 86 (AUC: 0.935) for TRISS 2. There was no statistically significant difference between the ROC curves of TRISS 1 and TRISS 2. CONCLUSION: Use of NISS instead of ISS in the TRISS model demonstrated no significant difference, and it can thus be recommended for use.
Authors: Hideo Tohira; Ian Jacobs; David Mountain; Nick Gibson; Allen Yeo Journal: Scand J Trauma Resusc Emerg Med Date: 2012-09-10 Impact factor: 2.953