Injury Severity, Vehicle Safety Features, and Intersection Crashes

Abstract

Objective: This article examines injury severity in crashes that occur within signalized intersections and identifies some factors that explain them.

Method: It uses a random effect ordered logit model and an unbalanced longitudinal panel data for 303 intersections for one medium-size city.

Results: Among its findings are 3% and 0.29% lower probabilities of sustaining evident and disabling injuries respectively when vehicle occupants wear three-point seat belts and corresponding 1.71% and 0.16% lower probabilities when they wear only shoulder belts. Also, the probabilities of sustaining evident and disabling injuries reduce by 0.87% and 0.08% when vehicles with airbags are involved in crashes. But, when front airbags deploy the probabilities of sustaining disabling, evident, and possible injuries increase by 1.3%, 11.93%, and 26.80%, respectively. Similarly, both side airbags deploying increases disabling and evident injuries by 1.8% and 15.53%, respectively. In addition, we find that the probability of sustaining severe injuries increases in head-on crashes.

Conclusion: These results emphasize the effectiveness of seat belt use, especially three-point seat belts, the inclusion of airbags in all vehicles, increased awareness of possible injuries from airbags deploying, signalized intersection designs that include pedestrian crossing signals, and additional research to identify the predominant types of injuries associated with airbags deploying.

Keywords:

• Ordered Logit
• Random Effect
• Crash Severity
• Signalized Intersections Airbag Deployment

ACKNOWLEDGEMENT

The data for this article are from a research project funded by the Urban Transit Institute at North Carolina A&T State University, Greensboro, NC. Its contents are solely those of this author.

Notes

¹ See http://www-nrd.nhtsa.dot.gov/deparments/nrd-30/ncsa/sts1/37-nc-2006.htm

² According to the North Carolina DMV-349 Crash Report Instructions Manual (NCDOT, 1999) intersection related crashes occur “within the influence area of the intersection and is caused by the operation of the intersection” (p. 77).

³ See http://www.nrd.nhtsa.dot.gov/departments/nrd-12/809-716/pages/task2.html

⁴ Although this percentage appears very high, rechecking the data did not yield reasons to be suspicious.

⁵ The NCDOT (1999) puts fell asleep, fainted, and fatigued, etc., into one group and defines it as a driver who “experienced a temporary loss of consciousness or was operating in a reduced physical and mental capacity due to weariness, medication, or other drugs” (NCDOT, 1999, p. 75).

⁶ Hereafter, disabling injuries refer to non-fatal and other types of injuries that are ambulatory.

⁷ The stepwise regression eliminated many variables from the model including fixed objects, amber time settings on the minor road, solid medians, left-turning vehicles, sidewalks, weather, and time of crash. The data source lists 10 physical conditions of drivers, 7 alcohol-and drug-suspected conditions, 32 vehicle styles, 13 types of collisions, and several categories of other variables. Only the variables that entered the stepwise regression are used in further analysis and are in Table I.

⁸ See http://www.safety-council.org/info/traffic/distract.htm, accessed February 8, 2008.

⁹ The presence of airbags in cars could reflect federal regulations of having airbags in vehicles, as well as risk-averse behaviors of vehicle owners and vehicle vintage.

¹⁰ This result is also not new. Virtually all new vehicles have stickers that warn front-seat occupants about possible injuries from airbags deploying.

¹¹ In much of the research reviewed, the negative impacts of airbag deployment are emphasized forcefully but obscured in sentences that deemphasize their importance.