Geographic variations in mortality from motor vehicle crashes

N Engl J Med. 1987 May 28;316(22):1384-7. doi: 10.1056/NEJM198705283162206.


Using a new technique to study the mortality associated with motor vehicle crashes, we calculated population-based death rates of occupants of motor vehicles during the period 1979 through 1981 and mapped them according to county for the 48 contiguous states of the United States. Mortality was highest in counties of low population density (r = 0.57; P less than 0.0001) and was also inversely correlated with per capita income (r = 0.23; P less than 0.0001). Death rates varied more than 100-fold; for example, Esmeralda County, Nevada, with 0.2 residents per square mile (2.6 km2), had a death rate of 558 per 100,000 population, as compared with Manhattan, New York, with 64,000 residents per square mile and a death rate of 2.5 per 100,000. Differences in road characteristics, travel speeds, seat-belt use, types of vehicles, and availability of emergency care may have been major contributors to these relations.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Accidents, Traffic*
  • Automobiles
  • Humans
  • Mortality
  • Rural Population
  • Seat Belts
  • United States