Citation

Braver ER, Zador PL, Thum D, Mitter EL, Baum HM, Vilardo FJ. Accid. Anal. Prev. 1997; 29(1): 79-96.

Affiliation

Insurance Institute for Highway Safety, Arlington, VA 22201, USA.

Copyright

(Copyright © 1997, Elsevier Publishing)

DOI

unavailable

PMID

9110043

Abstract

Studies of the crash experience of tractors pulling multiple trailers have reached different conclusions concerning the relationship of truck configuration to crash risk. A previous case-control study found a significant increase in crash risk for double-trailer trucks in the state of Washington. The present case-control study was done of tractor-trailers crashing on Indiana interstates during November 1989-March 1991. Controls were obtained for 25% of the crash sites and were all tractor-trailers passing the crash sites during a traffic observation session one to four weeks following a crash on the same day of the week for 30 minutes at the same time of day. Logistic regression identified day of week, time of day, urban/rural area, and specific highway as significant predictors of controls' truck configuration. This model was applied to the cases to estimate the expected number of double-trailer cases. For all crashes combined, no increased crash risk was observed for doubles (Standardized Crash Ratio (SCR) = 83). Doubles were significantly underinvolved in multiple-vehicle crashes (SCR = 74), crashes on dry roads (SCR = 61), and crashes on wet (other than snow, ice, or slush) roads (SCR = 54). Doubles were significantly overinvolved in crashes on roads with snow, ice, or slush (SCR = 153). Because truck configuration was highly associated with driver age and work operation attributes among trucks in crashes, the absence of control data on these potential confounders precluded definitive assessment of the intrinsic risk of multiple versus single-trailer vehicles.