Ram Nirula1, Daniel Talmor, Karen Brasel. 1. Division of Trauma/Critical Care, Department of Surgery, Medical College of Wisconsin, Milwaukee, WI, USA.
Abstract
BACKGROUND: Identification of motor vehicle crash (MVC) characteristics associated with thoracoabdominal injury would advance the development of automatic crash notification systems (ACNS) by improving triage and response times. Our objective was to determine the relationships between MVC characteristics and thoracoabdominal trauma to develop a torso injury probability model. METHODS: Drivers involved in crashes from 1993 to 2001 within the National Automotive Sampling System were reviewed. Relationships between torso injury and MVC characteristics were assessed using multivariate logistic regression. Receiver operating characteristic curves were used to compare the model to current ACNS models. RESULTS: There were a total of 56,466 drivers. Age, ejection, braking, avoidance, velocity, restraints, passenger-side impact, rollover, and vehicle weight and type were associated with injury (p < 0.05). The area under the receiver operating characteristic curve (83.9) was significantly greater than current ACNS models. CONCLUSION: We have developed a thoracoabdominal injury probability model that may improve patient triage when used with ACNS.
BACKGROUND: Identification of motor vehicle crash (MVC) characteristics associated with thoracoabdominal injury would advance the development of automatic crash notification systems (ACNS) by improving triage and response times. Our objective was to determine the relationships between MVC characteristics and thoracoabdominal trauma to develop a torso injury probability model. METHODS: Drivers involved in crashes from 1993 to 2001 within the National Automotive Sampling System were reviewed. Relationships between torso injury and MVC characteristics were assessed using multivariate logistic regression. Receiver operating characteristic curves were used to compare the model to current ACNS models. RESULTS: There were a total of 56,466 drivers. Age, ejection, braking, avoidance, velocity, restraints, passenger-side impact, rollover, and vehicle weight and type were associated with injury (p < 0.05). The area under the receiver operating characteristic curve (83.9) was significantly greater than current ACNS models. CONCLUSION: We have developed a thoracoabdominal injury probability model that may improve patient triage when used with ACNS.
Authors: Monique Brink; Jaap Deunk; Helena M Dekker; Michael J R Edwards; Digna R Kool; Arie B van Vugt; Cornelis van Kuijk; Johan G Blickman Journal: Eur Radiol Date: 2009-09-17 Impact factor: 5.315
Authors: C I Schulman; V Wilbur; B Leibowitz; L Labiste; E Perdeck; G Bahouth; K Digges; J S Augenstein Journal: Annu Proc Assoc Adv Automot Med Date: 2007