Opportunities for crash and injury reduction: A multiharm approach for crash data analysis

ABSTRACT

Objective: A multiharm approach for analyzing crash and injury data was developed for the ultimate purpose of getting a richer picture of motor vehicle crash outcomes for identifying research opportunities in crash safety.

Methods: Methods were illustrated using a retrospective analysis of 69,597 occupant cases from NASS CDS from 2005 to 2015. Occupant cases were analyzed by frequency and severity of outcome: fatality, injury by Abbreviated Injury Scale (AIS), number of cases, attributable fatality, disability, and injury costs. Comparative analysis variables included precrash scenario, impact type, and injured body region.

Results: Crash and injury prevention opportunities vary depending on the search parameters. For example, occupants in rear-end crash scenarios were more frequent than in any other precrash configuration, yet there were significantly more fatalities and serious injury cases in control loss, road departure, and opposite direction crashes. Fatality is most frequently associated with head and thorax injury, and disability is primarily associated with extremity injury. Costs attributed to specific body regions are more evenly distributed, dominated by injuries to the head, thorax, and extremities but with contributions from all body regions. Though AIS 3+ can be used as a single measure of harm, an analysis based on multiple measures of harm gives a much more detailed picture of the risk presented by a particular injury or set of crash conditions.

Conclusions: The developed methods represent a new approach to crash data mining that is expected to be useful for the identification of research priorities and opportunities for reduction of crashes and injuries. As the pace of crash safety improvement accelerates with innovations in both active and passive safety, these techniques for combining outcome measures for insights beyond fatality and serious injury will be increasingly valuable.

KEYWORDS:

• Crash data mining
• epidemiology
• NASS CDS
• disability
• cost
• attributable fatality

Acknowledgments

The authors are grateful to Dr. Peter Martin (NHTSA) and Dr. HyunJung Kwon (TRC Inc.) for advice and assistance in the application of the attributable fatality method.

Funding

Funding from the National Highway Traffic Safety Administration is gratefully acknowledged.