Short communicationThe role of exposure on differences in driver death rates by gender and age: Results of a quasi-induced method on crash data in Spain
Introduction
Driver death rates are widely used in all countries as health indicators for road crashes, with several purposes; i.e., quantifying the magnitude of the problem, monitoring the usefulness of road safety measures, or identifying subgroups of high-risk drivers. Regarding the latter purpose, the experience in Europe, one of the safest regions in the world, has shown the success of the adoption of feasible quantitative targets based on the reduction of death rates in specific sub-groups of drivers (International Road Traffic and Accident Database, 2013). The large differences found in death rates from traffic injuries across age and gender groups of drivers (European Road Safety Observatory, 2013) has been used to identify subgroups of high-risk drivers (for example, young male drivers), in order to prioritize preventive strategies for these subgroups (Santamarina-Rubio et al., 2014). However, the real causes of these differences cannot be easily ascertained through the use of mortality rates based on the number of registered drivers by gender and age, because this denominator does not capture the amount of exposure yielded by each group of drivers (European Road Safety Observatory, 2005). Therefore, direct exposure indicators that truly fit these mortality rates to real conditions of the driver population are required. Different types of direct exposure measures are used for estimating mortality risk. Vehicle-kilometers (or person-kilometers, taking into account vehicle occupancy) and time spent in traffic are the closest to the theoretical definition of exposure (European Road Safety Observatory, 2005). Together with number of trips, they are usually measured by both national surveys and specific questionnaires (Blanchard et al., 2010, Chipman et al., 1992). Vehicle-kilometers is perhaps the most often used direct exposure measure because it captures regional and temporal variations in the use of roads in a particular area (European Road Safety Observatory, 2005). Unfortunately, in Spain there are no routine surveys addressing the amount of exposure for different subgroups of drivers and driving environments. Instead, the information on exposure is based on aggregate estimates of the average kilometers travelled, referred only to non-urban roads (Ministerio de Fomento, 2014), or on indirect estimates from the registered drivers or fleet (Novoa et al., 2009). This entails a strong limitation for investigating the traffic accidents-related factors, prioritizing road safety interventions and explaining the differences in the magnitude of death rates across subgroups of drivers.
In an effort to overcome the above limitation, in the present study we apply an indirect approach based on a quasi-induced exposure method (Lardelli-Claret et al., 2006, Stamatiadis and Deacon, 1997), which may be useful for those countries that do not get direct exposure estimates, allowing to adjust for the differences in driver death rates found between age and gender subgroups of drivers by their respective amount of exposure. Therefore, the aim of the present study was to compare and explain the age and gender differences in driver death rates with and without adjustment by exposure in Spain, from 2004 to 2012.
Section snippets
Methods
The Spanish Dirección General de Tráfico (Traffic General Directorate) provided the two data sources for this study. The first one was the Spanish Register of Drivers of motorized vehicles (Dirección General de Tráfico, 2015), from which we obtained the number of licensed drivers stratified by year (from 2004 to 2012), by gender and age groups (15–17 years, 18–29 years, 30–44 years, 45–64 years, 65–74 years and >74 years). The study period started in 2004, when road safety was included on the
Results
In total, 17,975 drivers died in traffic accidents between 2004 and 2012 in Spain (90.3% were men and 9.7% were women). The annual number of drivers killed decreased from 2820 in 2004 to 1157 in 2012 (58.6% reduction among men and 61.9% among women).
Regarding variations in death rates across age groups, we found strong differences depending on the use of CDRR or ADRR. When we assessed the mortality rates in the overall population of drivers, CDRR showed that the youngest age group (15–17 years)
Discussion
The magnitude and, in some cases, the direction of the differences in death rates according to age and gender groups change substantially when adjustment by exposure is taken into account. When unadjusted rates were used, the highest mortality was found for the youngest drivers; however, the adjusted comparisons showed the highest rates in the oldest drivers. The high amount of driving exposure in the younger drivers explains most of their higher death rates. In the same way, a large part of
Conflicts of interest
None declared.
Funding
This work was supported by Health Strategic Action, co-financed by FEDER (European Regional Development) funds, “Una manera de hacer Europa” [grant number: PI14/00050] and Spanish Network on Addictive Disorders (RTA) [RD06/0001/1018 and RD12/0028/0018]. Writing of the paper was supported by “Ministerio de Economía y Competitividad, Actuación de Formación Posdoctoral” [FPDI-2013-15827].
Acknowledgements
The authors are grateful to the Directorate General of Traffic (DGT) for allowing access to their database of traffic accidents with victims and sending reports on traffic injury deaths. Thanks also to Mónica Ruiz for her help in searching for information.
References (22)
- et al.
Correspondence between self-reported and objective measures of driving exposure and patterns in older drivers
Accid. Anal. Prev.
(2010) - et al.
Time vs. distance as measures of exposure in driving surveys
Accid. Anal. Prev.
(1992) Older drivers' accident risk: conceptual and methodological issues
Accid. Anal. Prev.
(1998)- et al.
Difficulties with quasi-induced exposure when speed varies systematically by vehicle type
Accid. Anal. Prev.
(2007) - et al.
A review of the validity of the underlying assumptions of quasi-induced exposure
Accid. Anal. Prev.
(2010) - et al.
A comprehensive review on the quasi-induced exposure technique
Accid. Anal. Prev.
(2014) - et al.
Comparison of two methods to assess the effect of age and sex on the risk of car crashes
Accid. Anal. Prev.
(2011) - et al.
Gender differences in road traffic injury rate using time travelled as a measure of exposure
Accid. Anal. Prev.
(2014) - et al.
Quasi-induced exposure: methodology and insight
Accid. Anal. Prev.
(1997) - Estadísticas e indicadores. Available at: http://www.dgt.es/es/seguridad-vial/estadisticas-e-indicadores/ (accessed...
Cited by (18)
Investigation on the driver-victim pairs in pedestrian and bicyclist crashes by latent class clustering and random forest algorithm
2023, Accident Analysis and PreventionThe problem of women's "road unsafety" in Africa
2023, Transportation Research ProcediaLong-term impact of the 2008 economic crisis in Spain on road traffic collisions mortality by socioeconomic position
2021, Health and PlaceCitation Excerpt :On the other hand, overall RTC mortality differs by some sociodemographic characteristics. It is known that rates are higher in men, in older drivers, in single people, and is inversely related to population density (Pulido et al., 2016; Spoerri et al., 2011). The findings of various studies have shown an interaction between gender and age among different road users in non-fatal RTC (Santamarina-Rubio et al., 2014).
Validating the underlying assumption of quasi-induced exposure technique disaggregated by crash injury severity
2021, Journal of Safety ResearchCitation Excerpt :Moreover, the QIE can be finely disaggregated into detailed levels (e.g., specific time periods, locations, and driver characteristics), which is usually impossible for the direct exposures (Jiang et al., 2014). Due to the distinct advantages, the QIE methods have enjoyed the popularity in the field of traffic safety research, such as identifying the crash propensities of particular road users (DeYoung et al., 1997; Lardelli-Claret et al., 2011; Braitman et al., 2014; Martínez-Ruiz et al., 2015; Pulido et al., 2016; Naghawi, 2019), studying the specific crash types (Yan et al., 2005; Huang & Chin, 2009), assessing the safety impacts of traffic control strategy (Zhang et al., 2019), and evaluating the effectiveness of a driver license program (Jiang & Lyles, 2010; Curry et al., 2017; Curry, 2017; Palumbo et al., 2018). In the QIE theory, there are two basic assumptions (Jiang et al., 2014) that need to be strictly satisfied before the application: (a) it should be 2-vehicle crashes with one at-fault (D1) and one not-at-fault drivers (D2); and (b) the D2s are assumed to be the representatives of the driving population when and where crashes occur.
Development and validation of the Drivers’ Emotional Intelligence Inventory (EMOVIAL)
2020, Transportation Research Part F: Traffic Psychology and BehaviourCitation Excerpt :Research could also examine the association between drivers’ EI and variables such as risk perception (Arneu-Sabatés et al., 2012) or feelings of invulnerability. In all cases, it would be important to explore gender differences, not least so as to analyze whether EI may partly explain the existing differences in death rates between men and women (Pulido et al., 2016). To conclude, we believe that the availability of the EMOVIAL inventory will allow researchers to examine the relationship between EI and other key variables in the area of road safety and traffic psychology.