Mortality Prediction of Head Abbreviated Injury Score and Glasgow Coma Scale: Analysis of 7,764 Head Injuries Demetrios Demetriades, MD, PhD, FACS, Eric Kuncir, MD, FACS, James Murray, MD, FACS, George C Velmahos, MD, FACS, Peter Rhee, MD, FACS, Linda Chan, PhD BACKGROUND: We assessed the prognostic value and limitations of Glasgow Coma Scale (GCS) and head Abbreviated Injury Score (AIS) and correlated head AIS with GCS. STUDY DESIGN: We studied 7,764 patients with head injuries. Bivariate analysis was performed to examine the relationship of GCS, head AIS, age, gender, and mechanism of injury with mortality. Stepwise logistic regression analysis was used to identify the independent risk factors associated with mortality. RESULTS: The overall mortality in the group of head injury patients with no other major extracranial injuries and no hypotension on admission was 9.3%. Logistic regression analysis identified head AIS, GCS, age, and mechanism of injury as significant independent risk factors of death. The prognostic value of GCS and head AIS was significantly affected by the mechanism of injury and the age of the patient. Patients with similar GCS or head AIS but different mechanisms of injury or ages had significantly different outcomes. The adjusted odds ratio of death in pene- trating trauma was 5.2 (3.9, 7.0), p  0.0001, and in the age group  55 years the adjusted odds ratio was 3.4 (2.6, 4.6), p  0.0001. There was no correlation between head AIS and GCS (correlation coefficient -0.31). CONCLUSIONS: Mechanism of injury and age have a major effect in the predictive value of GCS and head AIS. There is no good correlation between GCS and head AIS. ( J Am Coll Surg 2004;199:216–222. © 2004 by the American College of Surgeons) The two most widely used methods of describing the severity of head trauma are the Glasgow Coma Scale (GCS) and the head Abbreviated Injury Score (AIS). The GCS is a physiologic scale based on clinical exami- nation and the AIS is an anatomic score based on CT scan or operative or autopsy findings. Despite the exten- sive use of these methods in clinical practice, very few and usually small studies have validated their value and the effect of other variables, such as age and mechanism of injury, on their prognostic value. Almost all existing studies include patients with major associated extracra- nial injuries, which have a major effect on patient sur- vival. No study has ever attempted to correlate head AIS with initial GCS. The present study was designed to assess the mortality prediction value of the GCS and head AIS, identify any limitations of these scores, and correlate the degree of anatomic injury with GCS. METHODS This trauma registry study included all head injuries with head AIS  1 (neck injuries were excluded) admit- ted at the Los Angeles County and University of South- ern California trauma center during a 10-year period from 1993 to 2002. Patients with head AIS = 0 were excluded irrespective of GCS. The Los Angeles County paramedics have no authority to sedate or paralyze trauma patients. The admission GCS is calculated by the trauma team. For patients with prehospital intubation, the preintubation GCS obtained by paramedics was used for the purpose of the study. The trauma registry captures 130 data fields and is maintained by seven full- No competing interests declared. Received December 30, 2003; Revised February 5, 2004; Accepted February 27, 2004. From the Division of Trauma and SICU, Department of Surgery, Keck School of Medicine, University of Southern California, Los Angeles, CA. Correspondence address: Demetrios Demetriades, MD, PhD, FACS, De- partment of Surgery, Division of Trauma and Surgical Critical Care, LAC+USCTrauma Center, 1200 N State St, Rm 1105, Los Angeles, CA 90033. 216 © 2004 by the American College of Surgeons ISSN 1072-7515/04/$30.00 Published by Elsevier Inc. doi:10.1016/j.jamcollsurg.2004.02.030