Azim Honarmand1, Mohammadreza Safavi. 1. Department of Anesthesiology and Intensive Care, Isfahan University of Medical Sciences, Isfahan, Iran. honarmand@med.mui.ac.ir
Abstract
BACKGROUND: This study validates the accuracy of the Injury Severity Score (ISS) and the New Injury Severity Score (NISS) systems for prediction of need intubatin (NI), need mechanical ventilation (NMV), and duration of mechanical ventilation (DMV) in intensive care unit (ICU) trauma patient admissions. METHODS: One-hundred ten trauma patients were included in this prospective cohort study. The predictive accuracies of the ISS and the NISS were compared using Receiver Operator Characteristic (ROC) curves and Hosmer-Lemeshow (H-L) statistics for the logistic regression model of ICU admission. RESULTS: For prediction of NI, the best cut-off points were 22 for ISS and 27 for NISS. The positive prediction value was 91.6% in NISS and 87.8% in ISS. The Youden index had best cut-off points at 0.47 for NISS and 0.57 for ISS. The area under Receiver Operating Characteristic (ROC) curve was 0.79 in the ISS and 0.86 in the ISS. There were statistical differences among NISS with ISS in terms of Youden index and the area under the ROC curve (p<0.05). For the prediction of NMV, NISS yielded significantly better results in the area under the ROC curve and Youden index than those of ISS (p<0.05). CONCLUSION: For prediction of NI or NMV, the NISS has better accuracy than ISS.
BACKGROUND: This study validates the accuracy of the Injury Severity Score (ISS) and the New Injury Severity Score (NISS) systems for prediction of need intubatin (NI), need mechanical ventilation (NMV), and duration of mechanical ventilation (DMV) in intensive care unit (ICU) traumapatient admissions. METHODS: One-hundred ten traumapatients were included in this prospective cohort study. The predictive accuracies of the ISS and the NISS were compared using Receiver Operator Characteristic (ROC) curves and Hosmer-Lemeshow (H-L) statistics for the logistic regression model of ICU admission. RESULTS: For prediction of NI, the best cut-off points were 22 for ISS and 27 for NISS. The positive prediction value was 91.6% in NISS and 87.8% in ISS. The Youden index had best cut-off points at 0.47 for NISS and 0.57 for ISS. The area under Receiver Operating Characteristic (ROC) curve was 0.79 in the ISS and 0.86 in the ISS. There were statistical differences among NISS with ISS in terms of Youden index and the area under the ROC curve (p<0.05). For the prediction of NMV, NISS yielded significantly better results in the area under the ROC curve and Youden index than those of ISS (p<0.05). CONCLUSION: For prediction of NI or NMV, the NISS has better accuracy than ISS.
Authors: Lilia de Souza Nogueira; Cristiane de Alencar Domingues; Renato Sérgio Poggetti; Regina Marcia Cardoso de Sousa Journal: PLoS One Date: 2014-11-06 Impact factor: 3.240