This paper discusses a methodology to map crash injuries from the format used in individual or state-wide hospital discharge databases (ICD-9-CM (International Classification of Diseases, ninth revision, Clinical Modification)) to the format used for the crash injury database maintained as part of the National Highway Traffic Safety Administration's (NHTSA) National Accident Sampling System Crashworthiness Data System (NASS/CDS). A unique attribute of this map (referred to as "crashmap") is that it is specifically "tuned" to provide the best fidelity for the injuries sustained in motor vehicle crashes. By converting hospital discharge databases, crashmap can greatly increase the number of severe cases for analysis beyond that available in NASS/CDS. The crashmap can also be used with linked crash data, such as data produced from NHTSA's Crash Outcome Data Evaluation System (CODES) project. Results from analysis of 8000+ patients and 50,000 injuries with current mapping and scoring methods illustrate the issues involved in obtaining the correct severity scores, Injury Severity Score (ISS) values and counts for the NASS/CDS injury system with these tools. Examples of crashmap results are shown for the conversion of the pre-dots, severity, and "numerical consistency" of the injuries between the systems. Examples for crash injuries in NASS/CDS and hospital data show how numerical consistency helps in comparing counts of injuries between systems. (A)
Abstract