Controlling for exposure changes the relationship between ethnicity, deprivation and injury: an observational study of child pedestrian injury rates in London

Background Research has suggested that inequalities in risk exposure may help explain identified ethnic inequalities in child pedestrian injury risk. However, addressing risk exposure in epidemiological research presents theoretical and methodological challenges. This article conceptualises the risk of child pedestrian injury as related to both exposure levels (the quantity of time children spend as pedestrians) and the probability of a hazard where that exposure takes place (the quality of the road environment). Objective To investigate the sensitivity of results on ethnic inequalities in child pedestrian injury risk in London to control for exposure and hazard levels. Methods Using police records of injury 2000-2009, we modelled the relationship between ethnicity, deprivation and child pedestrian injury rates in London using characteristics of the road environment to control for hazard levels and restricting the analysis to the time of the weekday morning commute (7.00-9.00 am), when most children must make a journey to school, to control for exposure levels. Results Controlling for risk exposure in this way fundamentally changed the nature of the relationship between ethnicity, deprivation and child pedestrian injury. During the time of the morning commute to school, ` Black' children were found to have higher pedestrian injury rates in the least-deprived areas. Conclusions To inform effective strategies for reducing injury inequality, it is vital that exposure to risk is both acknowledged and considered.