Factors related to the risk of pedestrian fatality after a crash in Spain, 1993–2013

Highlights

• Associations were estimated between pedestrian fatality and other factors.

• Pedestrian risk of death after a crash increased exponentially with age.

• Drivers’ visual defects were associated with a 21% higher risk of pedestrian fatality.

• More than one person in the vehicle increased pedestrian risk of death.

• Reinforce the assessment of drivers’ visual acuity would be an advisable strategy.

Abstract

Introduction

The aim of this study was to quantify the magnitude of association between pedestrian fatalities during the first 24 h after a crash and pedestrian-, driver-, vehicle- and environment-related characteristics in Spain from 1993 to 2013.

Methods

Data were analyzed for all 203,622 traffic crashes involving a pedestrian and a motor vehicle recorded in the Spanish Registry of Road Crashes with Victims. After multiple imputation for missing values, crude (CMRR) and adjusted mortality rate ratios (AMRR) were obtained for each variable with Poisson regression models.

Results

Pedestrian risk of death after a crash increased nearly exponentially with pedestrian age. Male sex, committing an infraction and having a physical defect were also associated with a higher risk of death (AMRR 1.27, 95%CI 1.17–1.37 for physical defect). Regarding driver-related factors associated with pedestrian fatalities, visual defects (AMRR 1.21, 95%CI 1.08–1.37) and the commission of a speed infraction (AMRR 2.59, 95%CI 2.43–2.76) increased the risk. Heavy vehicles (trucks, vans, buses) and the presence of passengers were also associated with a higher risk of pedestrian death. The risk of pedestrian death was lower for crashes that occurred between 12:00 and 14:00, in good light conditions, at intersections, and when the pedestrian was on a sidewalk. Risk was higher in crashes in rural areas with fewer than 5000 inhabitants.

Conclusions

We identified several factors strongly associated with the risk of pedestrian fatality; some of these factors are analyzed here for the first time. This knowledge is potentially useful in the design and prioritization of measures intended to increase pedestrian safety.

Introduction

Mortality in pedestrians due to road crashes is an important public health issue worldwide, even in the most developed countries. Among all traffic deaths, pedestrian fatalities account for 27% in Europe (World Health Organization, 2018) and 21% in Spain (Dirección General de Tráfico, 2017). This problem is expected to worsen due to the increase in the number of older people (who are the most likely to die as a consequence of a crash) and to the tendency toward increased walking in urban settings (Niebuhr et al., 2016, Rolison et al., 2012, Shinar, 2012). Consequently, it seems advisable to identify factors that influence the risk of pedestrian death after a vehicle impact, which can be subdivided into four groups: pedestrian-, driver-, vehicle- and environment- (geographic and temporal) related factors.

Many previous studies have assessed the impact of some factors in the aforementioned groups on the severity of crashes involving pedestrians. For example, older pedestrian age is recognized as a marker of higher fatality in most studies (Haleem et al., 2015, Islam and Hossain, 2015, Ma et al., 2017, Mohamed et al., 2013, Niebuhr et al., 2016, Pour-Rouholamin and Zhou, 2016, Vanlaar et al., 2016). Regarding vehicle-related factors, speed is undoubtedly the most relevant predictor of crash severity (Kröyer et al., 2014, Li et al., 2017, Matsui et al., 2016, Oikawa et al., 2016, Olszewski et al., 2015, Verzosa and Miles, 2016, Xu et al., 2016, Zhang et al., 2014). There is also agreement regarding the greater severity related with crashes involving heavier vehicles such as trucks or buses (Aziz et al., 2013, DiMaggio et al., 2006, Jang et al., 2013, Ma et al., 2017, Matsui et al., 2016, Mohamed et al., 2013, Oikawa et al., 2016, Tay et al., 2011, Zajac and Ivan, 2003). Finally, some environment-related characteristics have also consistently been related with severity in most previous studies, such as crashes occurring at night and/or with poor light conditions (Aziz et al., 2013, Coate and Markowitz, 2004, Islam and Hossain, 2015, Jang et al., 2013, Li et al., 2017, Mohamed et al., 2013, Moudon et al., 2011, Olszewski et al., 2015, Pour-Rouholamin and Zhou, 2016, p.; Tay et al., 2011; Verzosa and Miles, 2016; Zhang et al., 2014) and in rural areas compared to urban settings (Islam and Hossain, 2015, Islam and Jones, 2014, Pour-Rouholamin and Zhou, 2016, Tay et al., 2011).

However, few studies have attempted to jointly consider the following aspects of risk in connection with pedestrian fatalities:

• Treating death as the specific outcome. Fortunately, pedestrian death is a rare event after a road crash. Therefore the number of crashes required to obtain precise estimates of the association between this outcome and different risk factors is usually unattainable in datasets available at present.

• Including all types of crashes involving pedestrians, rather than specific subsets of them (based, for example, on characteristics of the pedestrian, the environment or the type of vehicle involved). This limits the external validity of the results.

• Using datasets from a nationwide register of road crashes to cover long sampling periods. Apart of solving the problems of achieving a large sample size and improving external validity, this would also make it possible to estimate the effect of temporal trends on the risk of death. Most published research is restricted to a single city, region or county, and usually comprises a period of only a few years.

• Considering, in the same design, the complete set of factors involved in the risk of pedestrian death – an important issue given that many factors are strongly interrelated. This is especially relevant for factors related with the characteristics of the driver of the vehicle involved in the crash, which are not well addressed in most studies.

All the above limitations are notable even within Spain, where there are no studies of the association between different types of traffic-related factors and the risk of pedestrian fatality after a crash.

In an attempt to overcome these limitations, we designed the present study to determine the adjusted association between pedestrian-, driver-, vehicle-, and environment-related factors and the risk of pedestrian death in the 24 hours after a crash in Spain for the period from 1993 to 2013.