Assessing Hospital Based Burden for Nonfatal Motor Vehicle Crash Injuries by Race

BACKGROUND: Persistent disparities in motor vehicle crash (MVC) injuries are a major public health challenge. Almost all previous studies have been focusing on fatal injuries, partly because of data availability from the Fatality Analysis Reporting System (FARS). The lack of race information for non-fatal motor vehicle crash injuries in the United States has limited the understanding of racial disparities in motor vehicle crashes (MVCs). In this study, we described a pilot surveillance project in Nebraska that linked crash, drivers’ license, and hospital discharge data to investigate racial disparity among hospitalized non-fatal MVC injuries.

METHODS: Five years (2006-2010) of Nebraska motor vehicle crash data, drivers’ license data, and hospital discharge data were linked for this study. Only drivers' records were included in the analysis, since licensed drivers were used as at-risk population. Drivers age, gender, residence location were obtained from drivers' license data that was linked to crash data. Injury severity was assessed on the Maximum Abbreviated Injury Scale (MAIS). A log rate model was used to examine the likelihood of severe and non-severe MVC injuries by drivers’ race along the dimensions of age, sex and place of residence.

RESULTS: From 2006 to 2010, there were 1,145 Nebraskan drivers severely injured, and 18,216 non-severely injured in MVCs who were found in the hospitals discharge data. With the exception of the youngest (15-24) and oldest (65+) age groups for whites, black drivers had highest severe injury rates among the three race groups. They also had highest non-severe injury rates across all age groups. The gaps between black drivers and their two other counterparts were wider for the non-severely injured than for the severely injured. In the log rate model predicting severe injuries, only the interaction between race and residence location was significant; while in the model predicting non-severe injuries, race has significant interaction with age, gender, and residence location.    

CONCLUSIONS: In the absence of survey data, regular injury surveillance data (crash and hospital discharge data in this case) in combination with driver’s license data provide an alternative and viable solution for MVC disparity surveillance. Based on limited demographic information, we found racial disparities in severe and non-severe MVC injuries: Nebraska black drivers tended to be at higher risk, especially in metropolitan areas.