What is Injury Severity Scaling?
By Elizabeth Garthe, MHS and Nicholas Mango, BS, ME
INJURY SEVERITY SCALING IS A method to classify injuries using a standardized terminology and rank them based on certain parameters, such as threat to life. There are two major users of injury severity scales and data: the medical and engineering communities. The medical community uses severity data primarily for quality improvement, accreditation and research activities. For example, trauma centers use severity data to identify patients with relatively minor injuries who had unexpectedly poor outcomes. The engineering community uses severity data to identify injuries that can be reduced or eliminated by product design changes (or safety enhancements) and to measure the effectiveness of these interventions. For example, automotive safety engineers use severity data to study the effectiveness of seat belts and airbags in reducing or eliminating severe injuries.
Types of Scales
There are two major types of injury severity scaling systems: physiological and anatomical. Physiological scales use information such as blood pressure, pulse rate or respiratory rate that may change over the duration of the injury’s treatment period. In contrast, anatomical scales generally des-cribe an injury in terms of its location and specific lesion (e.g., open fracture to the ra-dius). Physiological scales are designed for use by clinicians, whereas, anatomic scales are designed primarily for non-clinician coders.
The most widely used anatomical injury severity scaling system in the world is the Association for the Advancement of Automotive Medicine’s (AAAM) Abbrevi-ated Injury Scale (AIS).1 The most recent revision, AIS-90, is “packaged” in the form of a 74-page coding manual or “dictionary” that lists the descriptions, codes and severity levels of 1,315 injuries.1, 2 It is organized into nine chapters: Head, Neck, Face, Thorax, Abdomen, Spine, Upper Extremi-ties, Lower Extremities and External (Skin, Burns and Other Trauma).
The AIS classifies individual injuries by body region on a six-point ordinal severity scale ranging from AIS-1 (minor) to AIS-6 (maximum). The AIS does not assess the combined effects of multiply injured patients.1 The AIS generally measures the relative threat to life of an individual injury in an average, healthy person. The scale is: 1 = minor; 2 = moderate; 3 = serious; 4 = severe; 5 = critical; 6 = maximum; 9 = injury, severity unknown.
Although an injury of AIS-4 is more severe than AIS-2; it is not twice as severe. AIS-6 injuries are often, but not always, fatal. The AIS is not an outcome scale: a patient can die with a minor (AIS-1) and can survive (rarely) with a maximum (AIS-6) injury.
AIS-90 Code Structure
The “anatomy” of an AIS-90 code is a six digit “pre-dot” that uniquely identifies the injury and a one digit severity level (Note: AIS-85 code structure is different). An example of an injury code and its components follows:
AIS-90 code: 450212.1 = “nondisplaced rib fracture, closed”
4 = body region, thorax
5 = anatomic structure, skeletal
02 = specific anatomic structure or na-ture (consecutive two digit numbers)
12 = level (consecutive two digit numbers)
1 = severity level, minor
An example of a series of thorax injuries of increasing severity is provided in the accompanying table.
What Is the ISS?
The AIS scales the severity of individual injuries. What do you do when you want to compare the severity of multiply injured patients? The Injury Severity Score (ISS) was developed to represent the overall severity of patients with multiple injuries.3 The definition of the ISS is: “The sum of the squares of the highest AIS code in each of the three most severely injured ISS Body Regions.” There are six ISS body regions: Head/Neck (including cervical spine), Face, Chest (including thoracic spine and inhalation injuries), Abdominal and Pelvic Contents (including lumbar spine), Extremities and Pelvic Girdle and External (skin, burns and other trauma, except inhalation injuries). Note: the six ISS body regions are not the same as the nine AIS chapters. The ISS cannot be calculated when there is an injury of unknown severity (AIS-9). ISS values range from one to 75 (any patient with AIS-6 is automatically ISS 75).1
Who Uses the AIS?
Hospitals use AIS-85 or AIS-90 severity data for a variety of purposes, including: improving quality of care (identifying cases with unexpected–good or bad–outcomes to review); meeting state or national certification requirements; and identifying trends in severity by length of stay, cost or outcome. Severity data helps trauma centers identify priorities for community injury prevention programs and monitor their success as well as provide data for research. The AIS is required for use in the United States by the American College of Surgeons (ACS) in trauma registries.4
The National Highway Traffic Safety Administration’s (NHTSA) National Crash Sampling System, Crashworthiness Data System (NASS/CDS) used variations of the AIS-85/80 (1988-1992) and AIS-90 (1993+) to hand-code injuries from motor vehicle crashes (MVCs).2, 5, 6 More than 250,000 injuries are included in the database for this time period alone.5, 6 NHTSA also uses an AIS-based severity system for its Special Crash Investigations (SCI). One recent use of the severity data was to study patterns of severe injuries from airbags; the results from this study contributed to changes in airbag rules and regulations.7
The National Transportation Safety Board (NTSB) uses its own version of AIS to study aviation crash injuries and other transportation-related injuries. It recently investigated a sample of car crashes involving child restraints and airbags and recommended corrective actions.8 Canada, Australia, Japan, United Kingdom, Germany, Italy and Sweden all use variations of the AIS.
It is important to note that although injury severity databases exist at hospitals, NHTSA and NTSB, they use different versions of the AIS. Therefore, the current situation is that data from these different sources cannot be readily compared or shared.
Issues in Using the AIS
The AIS uses different coding rules than the ICD-9-CM system.1,9 Therefore, specialized training is needed. The AAAM sponsors the only course in the world that teaches AIS and ISS coding. Information about this two-day course can be obtained from the AAAM.
Because the AIS is periodically revised, it is important to learn the rules in effect for the latest revision. For example, there were major changes between the AIS-85 and AIS-90. Codes were added, deleted and upgraded or downgraded in severity; the total number of codes increased from 1,224 to 1,315.2
Hand Coding
The AAAM teaches “hand coding” of injury data from patient hospital records using the AIS-90. This is the most accurate way to code severity at this time. It matches the complete information from a medical record to the AIS coding manual. It requires a trained coder and can average 15 minutes per chart, but the amount of time depends on the complexity of the case and number of injuries involved. Hand-coded severity data is the “gold standard” and generally is considerably more detailed than ICD-9-CM codes. In addition to being useful in quality improvement activities, the severity data can be used to identify trauma cases that may qualify for additional reimbursement.10
The “Mapping” Methodology
Not every organization has the resources to hand code severity data. Some analyses use a “mapping” methodology from ICD-9-CM discharge diagnoses to AIS.9, 11 The mapping methodologies from ICD-9-CM to AIS-85 (and AIS-90) are not as precise as hand-coded data and generally underestimate severity.9, 11 The most common map is from ICD-9-CM codes to one-digit AIS-85 severity levels (without pre-dot). Be very careful when comparing hand-coded AIS-90 data to data that has been described to you as being “mapped.” Most of the mapped data is in AIS-85–and this is not compatible with AIS-90.
Most existing maps are “generic”–mapping all types of trauma (gunshot, stab, MVC, falls, etc.) to AIS severity levels. A new map, “crashmap,” was developed specifically for MVC injuries.12 It maps ICD-9-CM discharge diagnoses for MVC injuries to the severity code format used by a large national sample of MVC injuries.
AIS Unification Efforts
At the 1997 AAAM Annual Conference, an initiative to unify the different injury databases that use the AIS was started by the authors, Dr. John States (as well as other “founders” of the AIS) and interested members of the AAAM. A process is being planned to involve the key U.S. and international “owners” and users of these very important databases, such as NHTSA, NTSB, automotive manufacturers, hospitals, insurers, researchers, etc. The objective is to have a unified AIS system in place that would enable the pooling of data from all these sources on a worldwide basis and a defined method to interface data with mortality and reimbursement systems. With unification, AIS will provide an international language of severity to satisfy the needs of researchers, engineers, clinicians and policy makers.
References
1. Association for the Advancement of Automotive Medicine, Abbreviated Injury Scale, 1990 Revision, Des Plaines, Illinois.
2. Garthe, E., “Comparison of the AIS-85 and AIS-90 with NASS-93.” 1996 Society of Automotive Engineers (SAE) Technical Paper No. 95MJA242. Paper presented at the February 1996 Annual Conference of SAE.
3. Baker, S. P., O’Neill, B., Haddon, W. Jr., et al: The Injury Severity Score: A method for describing patients with multiple injuries and evaluating emergency care. J Trauma 1974; 14: 187-196.
4. American College of Surgeons, Committee on Trauma–Resources for Optimal Care of the Injured Patient, 1993.
5. U.S. Department of Transportation, National Highway Traffic Safety Administration, National Crash Sampling System/Crashworthiness Data System, data files: 1988-1996.
6. Garthe, E., and Mango, N. “A Method of Mapping Pre & Post NASS-93 Injury Descriptions to Enable Multi-Year Data Comparisons.” Paper presentation at the Annual Meeting of the Society of Automotive Engineers, February 1997.
7. U.S. Department of Transportation, National Highway Traffic Safety Administration, Special Crash Investigation web-site: www.nhtsa.dot.gov/people /ncsa/sci.html
8. National Transportation Safety Board: Safety Study, The Performance and Use of Child Restraint Systems, Seatbelts, and Air Bags for Children in Passenger Vehicles. US Government Printing Office, Volumes I and II: PB96-917005 and PB96-917006, Washington, DC, September 1996.
9. Garthe, E. “The Compatibility Between the Abbreviated Injury Scale (AIS-80) and the International Classification of Diseases (ICD-9-CM).” Presented at the Annual Meeting of the American Association of Automotive Medicine, San Francisco, California, 1981. Published in the Quarterly Journal of the AAAM, March 1982. Recipient of the Dr. John States Award.
10. Garthe, E. “Putting Specialized Data Bases to Work for You in More Ways Than One,” QRC Advisor, October, 1997.
11. MacKenzie, E. J., Steinwachs, M. D., and Shankar, B. “Classifying Trauma Severity Based on Hospital Discharge Diagnoses: Validation of an ICD-9-CM to AIS-85 Conversion Table,” Medical Care, 1989; 27: 412.
12. Garthe, E., and Mango, N. “A Motor Vehicle Crash Injury Specific Map: ICD-9-CM to NASS/CDS.” Paper presented at the Annual Conference of the Association for the Advancement of Automotive Medicine, Orlando, FL, November 1997.
Liz Garthe and Nick Mango are based at Garthe Associates, Marblehead, MA, and can be reached at [email protected]. Garthe Associates conducts analyses of large health care and crash injury databases. Garthe is on the national faculty of the AAAM’s Injury Scaling Course and, with Mango, teaches a course on “Current Issues in Using Crash Injury Data” for the Society of Automotive Engineers.
Top 10 Tips for Using the AIS in an Electronic Data Base
1. Be sure the severity coders received training from the AAAM.
2. Include the complete AIS code (pre-dot and severity). This permits conversion of the
“pre-dot” to other revisions of AIS, as needed for longitudinal studies.
3. Add a field to record revision of AIS used (AIS-85, AIS-90, etc.).
4. Add a field for type and version of trauma registry software used.
5. If using a “mapped” severity, add fields for type and version of map and “high” or ”low” range used.
6. Use both the ISS and New ISS for overall severity (and verify that the formulas used generate correct results–the ISS is complicated).
7. Set up a quarterly audit process to assure the quality of the severity data using a sample of cases.
8. Generate reports to share with clinicians to illustrate the value of data.
9. Coach researchers using your severity data to properly cite the revision of AIS used and how it was coded (hand-coded or mapped) for papers.
10. Use caution when comparing your data to others–many different versions of AIS are “out there.”
If you want to be on the “cutting edge” with your electronic injury database, take one more step and code the “aspect” of the injury–right, left, bilateral. This also will help you obtain proper counts of injuries from your database. Crash investigators use even more detailed descriptions of aspect; this process helps to identify specific sources of injuries from contact points inside a vehicle or aircraft (e.g., rib fractures from a steering wheel).
–By Elizabeth Garthe, MHS and Nicholas Mango, BS, ME
