Motor Vehicle Accidents and Vehicular Pollution in Los Angeles

Abstract

A simple, “robust” and reasonably efficient nonparametric statistical method has been used to compare Los Angeles motor vehicle accident (MVA) frequencies with the levels of two air pollutants, oxidant (OX) and carbon monoxide (CO), during two high OX and two high CO periods. Statistically significant association (P < .0001) is found for OX and MVA, but no association appears to exist between CO and MVA. Thus, more often than not, increasing OX levels are accompanied by an increase in the frequency of MVA, while MVA frequencies remain essentially unchanged when CO levels increase. Lag studies of one to three hours show decreasing OX-HVA association and continuing nonassociation for CO and MVA. The possible pitfalls in this study are discussed. Most potential confounding factors have been eliminated or minimized by the statistical design; but smoking, spatial factors, and other pollutants should be considered in future investigations.