Exposure-Based, ‘Like-for-Like’ Assessment of Road Safety by Travel Mode Using Routine Health Data

Including measures of distance and time can improve risk comparisons of travel modes and may show that bicycling is not as hazardous as commonly thought.


  • For drivers, risk was highest in youth, fell by a factor of 20 for middle-aged adults, and rose dramatically again for those over age 70. There was less variation in risk for walking and bicycling by age, but a general fall in risk after age 20 and gradual increase by age. Comparisons with data from the Netherlands showed similar trends in increasing risk by age.
  • Risk (in fatalities per million hours’ use) were highest for male drivers ages 17-20 (1.3 f/mhu) male bicyclists over age 70 (2.2 f/mhu), and female pedestrians over age 70 (0.95 f/mhu).
  • In this study, males ages 17-20 had five times greater risk per hour while driving than bicycling (1.3 f/mhu for drivers compared to 0.24 f/mhu for bicyclists).
  • Females had lower risks than males across modes at younger ages, but this trend reversed above age 50.
  • In this study, pedestrians had higher fatalities per km traveled than cyclists (45 fatalities per billion km for pedestrians and 34 fatalities per billion km for cyclists).
  • At a fatality rate of 34 per billion km, an individual who cycles 1 hour/day for 40 years would cover 180,000 km with a 1 in 150 chance of fatal injury.


  • This study examined hospital admissions and deaths in England from 2007-2009 for pedestrians, cyclists, and car/van drivers by age and sex. This study used distance and time collected through the National Travel Survey (NTS) as measures of exposure to reduce error in comparisons for deaths and injuries by travel mode. Risk was measured as fatalities per million hour use (f/mhu) or per billion km (/Bn km).

Mindell, J.S., Leslie, D., Wardlaw, M. (2012). Exposure-Based, ‘Like-for-Like’ Assessment of Road Safety by Travel Mode Using Routine Health Data. PLoS ONE, 7(12), e50606.

filed under