Online Program

The Use of Quasi Induced Exposure to Identify Risk Factors for Motor Vehicle Crashes (307893)

Allison E Curry, Center for Injury Research and Prevention, The Children’ s Hospital of Philadelphia, Philadelphia, P 
Robert D Foss, University of North Carolina 
*Nina r Joyce, Brown University School of Public Health 
Kristina B Metzger, Center for Injury Research and Prevention, The Children’ s Hospital of Philadelphia, Philadelphia, P 
Gregory A Wellenius, Brown University School of Public Health, Department of Epidemiology 
Andrew R Zullo, Brown University School of Public Health, Center for Gerontology 

Keywords: case control, motor vehicle crash, quasi induced exposure

Although analyses of crash data have advanced our understanding of motor vehicle crashes, many studies are unable to account for the amount of time drivers spend driving and are thus “at risk” for a crash. In response, researchers developed the Quasi-Induced Exposure (QIE) technique. Using data from crash reports, QIE assumes the responsible driver in a crash “randomly selects” the non-responsible drivers from all those at the location and time of the crash. Under this assumption, QIE is methodologically equivalent to risk-set sampling in a case-control study and thus estimates reflect the same incidence rate ratio (IRR) as from a prospective cohort study that measured time-at-risk directly. However, few studies make this connection between QIE and risk-set sampling in a case-control study explicit. In this paper, we discuss this connection and the assumptions under which QIE produces unbiased estimates of the IRR. We demonstrate the utility of this method via an applied analyses related to Graduated Driver Licensing policy for teen drivers. We conclude with recommendations for the use of a QIE approach based on advances in methods for analyzing case-control studies.