Understanding the association between the Abbreviated Injury Scale score for the head region and outcomes following traumatic brain injury, Colorado 1998-2000
Date
2007
Journal Title
Journal ISSN
Volume Title
Abstract
Annually, an estimated 1.4 million Americans are affected by traumatic brain injury (TBI). It is the leading cause of morbidity and mortality among trauma individuals. While advances in medicine have helped to decrease mortality from TBI, less is known about the long-term outcomes following TBI. The goal of this research was to further ourunderstanding of long-term outcomes following TBI by identifying associations betweenone anatomical coding system and one-year outcomes following TBI. The Abbreviated Injury Severity Score is an anatomical scoring system used by injury researchers throughout the United States and the world to characterize individual injury. One attribute of this measure is the ability to independently characterize injury severity for one body region, for this study, specifically, the head region. Although the Abbreviated Injury Severity Score for the Head region (HAIS) is ananatomic measure of TBI severity, no study has assessed the inter- and intra-rater agreement of HAIS between trauma registrars at hospitals and trained state coders. Further, no studies have specifically assessed the association between HAIS and longterm outcomes following TBI using population-based data. The objectives of this study were to utilize data from two Colorado population-based studies containing HAIS scores to investigate the reliability of HAIS and its ability to predict long-term physical and cognitive outcomes after traumatic brain injury. The purposes of this study were: 1) to describe the intra-rater reliability of HAIS scores by having a trained coder employed by the Colorado Department of Public Health and Environment recode HAIS scores for traumatic brain injured cases from the Colorado Traumatic Brain Injury Surveillance system for the years 1999-2000; 2) to describe the inter-rater reliability of HAIS scores by comparing HAIS scores from cases in the Colorado Traumatic Brain Injury Surveillance system for 2000 to HAIS scores from trauma registrars at hospitals throughout Colorado; and, 3) to use HAIS scores to evaluate functional outcomes of traumatic brain injured individuals in Colorado who were enrolled between 1998 and 1999 in the Colorado Traumatic Brain Injury Registry and Follow-up System. Cases were defined using the International Classification of Diseases, 9th Revision, Clinical Modification diagnostic codes for TBI (800 - 801.9, 803 - 804.9, or 850 854.1, and 959.01) and included Colorado residents who were either admitted to hospitals or died prior to reaching the hospital from a TBI. To assess inter- and intra-rater agreement, data was selected from the Colorado Traumatic Brain Injury Surveillance system for years 1999 - 2000. A sample of 250 cases was randomly selected to assess intra-rater agreement. Approximately 624 cases were selected to assess inter-rater agreement. Weighted and non-weighted kappa statistics were used to assess inter- and intra-rater agreement, respectively. Landis and Koch (1977) cut points were used to interpret agreement findings. To identify long-term outcomes following TBI, 1,802 cases were used from the Colorado Traumatic Brain Injury Registry and Follow-up System (CTBIRFS), 1998 - 1999. Outcomes selected for this study were based on the conceptual model of function and disability developed by the World Health Organization. Logistic regression models were used to determine the association between TBI severity categories (HAIS) and one-year activity and societal participation outcomes. Logistic regression was used to determine the association between HAIS and cognitive outcomes one-year following TBI. All statistical analyses were conducted using SAS 9.1©. Results of this study found intra-rater agreement of HAIS to be “almost perfect” while inter-rater agreement between the trained state coder and the hospital trauma registrars was “substantial.” This finding was surprising given that individuals performing the coding often have varying levels of education and training, experience,and use and knowledge of database systems. Factors that potentially affect agreement thatwere not tested include injury factors such as impact forces, multi-system trauma, pharmaceutical drug usage, and use of personal protective equipment, such as helmets. Future studies should be conducted to identify the role of these factors when coding HAIS. In order to accurately assess function and disability following TBI, the severity of the TBI must be taken into account. Using HAIS categories mild, moderate, and severe TBI, individuals with moderate TBI (5.04 [95% confidence interval (1.67, 15.6)]) and severe TBI (4.08 [95% confidence interval (1.29, 12.7)]), were five times as likely to report needing help with Activities of Daily Living throughout the study period as compared to those with mild TBI, after adjusting for identified potential confounders. Similarly, subjects with moderate and severe TBI were more than 60% as likely to report needing help with Instrumental Activities of Daily Living throughout the study period as compared to those with mild TBI. The odds ratios for moderate and severe TBI were 1.90 [95% confidence interval (1.01, 3.57)] and 1.62 [95% confidence interval (0.81, 3.26)], respectively. Adjusting for identified potential confounders, subjects with moderate and severe TBI were more than 50% as likely to report diminished societal participation throughout the study period as compared to those with mild TBI. The odds ratios for moderate and severe TBI were 1.72 [95% confidence interval (1.18, 2.51)] and 1.58 [95% confidence interval (1.01, 2.47)], respectively. However, moderate and severe TBI were not associated with cognitive dysfunction. The results of this study indicate that HAIS is a good predictor of function and disability at the individual and societal levels, as measured by the activities and participation domains. The study failed to find an association between HAIS and cognitive disability. The results of this study support the need for individuals with a moderate and severe TBI (HAIS score greater than three) to participate in some form of rehabilitation to increase function and reduce disability following TBI. The objective of this study was to use data from the CTBIRFS and the CO TBI Surveillance system to expand upon the literature regarding outcomes following TBI. Specifically, the purpose of this study was to increase understanding of the Abbreviated Injury Scale for the head (HAIS) - an anatomical scoring system that potentially could be a predictor for long-term outcomes following TBI. As medicine advances and more individuals survive TBI, demands on rehabilitation resources will rise. The results of this study indicate that HAIS is a reliable scoring system that is associated with one-year outcomes following TBI. Using HAIS to assess severity of TBI will allow clinicians to identify and target rehabilitative efforts for TBI individuals and help individuals receive the rehabilitation services they need. Future research is needed to expand upon these findings to identify barriers to rehabilitation, such as cost and access to care, and to assess the role of rehabilitation on quality of life following TBI.
Description
Rights Access
Subject
medicine
epidemiology