Road Traffic Deaths in Brazil: Rising Trends in Pedestrian and Motorcycle Occupant Deaths

Abstract

Objective: According to the World Health Organization, the global burden of road traffic mortality exceeds 1.27 million people annually; over 90 percent occur in low- and middle-income countries. Brazil's road traffic mortality rate of ∼20 per 100,000 is significantly higher than nearby Chile or Argentina. To date, there has been very little information published on road traffic fatalities among vulnerable road users (VRUs) in Brazil.

Methods: Road traffic fatality data from 2000 to 2008 were extracted from Brazil's Mortality Information System (SIM). Road traffic deaths were extracted using the International Classification of Diseases (ICD-10) V-codes (V01–V89) and then subcategorized by VRU categories. Information was then disaggregated by gender, age, and region.

Results: In 2008, 39,211 deaths due to road traffic injuries were recorded in Brazil, resulting in a crude mortality rate of 20.7 per 100,000 inhabitants. Pedestrian mortality averaged 5.46 deaths per 100,000 between 2000 and 2008. The mortality rate for elderly pedestrians (80+ years) is 20.1 per 100,000, over 10 times that of 0- to 9-year-olds. In the past decade, motorcycle occupant mortality has dramatically increased by over 300 percent from 1.5 per 100,000 in 2000 to 4.7 per 100,000 in 2008. The 20- to 29-year age group remains most affected by motorcycle deaths, with a peak fatality rate of 10.76 per 100,000 in 2008. The north and northeast regions, with the lower per capita gross domestic product (GDP), have higher proportions of VRU deaths compared with other regions.

Conclusions: Vulnerable road users are contributing an increasing proportion of the road traffic fatalities in Brazil. Nationally, elderly pedestrians are at particularly high risk and motorcycle fatalities are increasing at a rapid rate. Less prosperous regions have higher proportions of VRU deaths. Understanding the epidemiology of road traffic mortality in vulnerable road user categories will better allow for targeted interventions to reduce these preventable deaths.

Keywords:

• Road traffic crashes
• Mortality
• Unintentional injuries
• Brazil
• Vulnerable road users
• Pedestrians
• Motorcyclists
• Bicyclists

INTRODUCTION

According to the World Health Organization (WHO), the global burden of road traffic mortality exceeds 1.27 million people annually and is ranked as the ninth leading cause of death, responsible for 2.2 percent of all deaths regardless of socioeconomic status (WHO 2008). Over 90 percent (nearly 1.16 million) of road traffic deaths and 92 percent (over 38 million) of road traffic crash–associated disability-adjusted life years (DALYs) occur in low- and middle-income countries (LMICs) despite these countries having only 48 percent of the world's registered vehicles (WHO 2008). The WHO-defined region of the Americas had an estimated 152,000 road traffic deaths and 4.58 million DALYs in 2004; traffic deaths comprised over 44 percent of all unintentional injury deaths that year. In a report summarizing published cost estimate studies, the annual direct (injury and vehicle) economic costs of road crashes in Latin America and the Caribbean were estimated to be US$19 billion (Jacobs et al. 2000).

Vulnerable road users (VRUs) are traditionally defined as slow-moving exposed/unprotected road users that are at higher risk of injury when involved in a crash; normally these include pedestrians, nonmotorized cyclists, and motorcyclists. When breaking down road traffic injuries by user group, 60 percent of traffic fatalities occur in motorized 4-wheeled cars in high-income countries (HICs) compared to 34 percent in LMICs. In contrast, pedestrians account for an average of 45 percent of road traffic fatalities in LMICs compared to 18 percent in HICs (Naci et al. 2009). The increasing burden of pedestrian, bicycle, and motorcycle injuries has been shown in some Latin American countries in recent studies. In Argentina in 2007, VRUs accounted for 29.5 percent of road traffic deaths (Ubeda et al. 2011). Hijar et al. (2003) showed a pedestrian mortality rate of over 7 per 100,000 in Mexico City, with a markedly high rate of 35 per 100,000 in certain districts in the mid-1990s.

Brazil is a middle-income country that is the largest in South America. It covers an area of 8.5 million km² with an estimated population of 190,732,694. The WHO estimates an annual road traffic mortality rate of 21.9 per 100,000, and when Reichenheim et al. adjusted for underreporting, they found a slightly higher rate of 23.1 per 100,000 (Reichenheim, 2011; WHO 2008). Brazil's road traffic mortality rates have been increasing since the 1960s, which correlates to increases in the registered motor vehicle fleet size; the registered fleet size has increased from approximately 32 million in 2001 to 54.5 million in 2008 (Vasconcellos 1999). In 2003, road traffic crashes in Brazil were responsible for over 26 percent of fatal injuries (rate of 17 per 100,000 population; Gawryszewski and Rodrigues 2006). The most commonly affected population in road traffic fatalities was found to be males in the “20 to 29 year old” age group (Franca et al. 2008). Reichenheim et al. (2011) showed using Brazil's Ministry of Health's mortality database that pedestrians account for nearly 35 percent of traffic deaths, with a mortality rate of just over 6 per 100,000 in 2007. They also showed an alarming increase in motorcycle deaths by over 800 percent between 1996 and 2007. A 1996 study in Sao Paulo, which is the largest city in Brazil, found that whereas one in ∼200 vehicle crashes resulted in a fatality, one in 10 pedestrian crashes resulted in a fatality (Companhia de Engenharia de Trafego 1996).

To date, there has been very little information published on road traffic fatalities among vulnerable road users in Brazil. At a time when multiple road safety intervention projects have been launched throughout the country and at the outset of the Decade of Action for Road Safety, it is important to understand the epidemiology of road traffic deaths in all road user categories in order for more specific targeting of intervention programs (Maffei de Andrade et al. 2008; Salvarani et al. 2009). This article utilizes the Brazilian Ministry of Health's mortality database to explore in detail the epidemiology of road traffic deaths in Brazil disaggregated by vulnerable road user categories.

METHODS

Road traffic fatality data from 2000 to 2008 were extracted from Brazil's Mortality Information System (Sistema de Informações sobre Mortalidade; SIM) via the DATASUS Database (Ministerio de Saude 2011). SIM was created by the Ministry of Health in 1975, and the government provides training to the teams responsible for collecting mortality information. Information is collected primarily from hospitals; additional information is sometimes collected from medical examiner's offices at the municipal level (Fajardo et al. 2009).

The SIM data set is publicly available and contains information on over 40 different mortality variables. Mortality information for SIM is collected from death certificates, which since 1996 have been coded with V-codes according to the codes in the 10th revision of the International Classification of Diseases (ICD-10). Deaths through 2008 are currently available in this database. Deaths due to road traffic crashes are recorded as codes V01–V89 regardless of the length of time that elapses between the crash and the time of death. All road traffic deaths classified with one of these 3 codes were extracted. Specific ranges of V-codes were used to designate road user categories. Vulnerable road users are designated in the database using one of the following: V01–V09 for pedestrians, V10–V19 for bicyclists, V20–V39 for motorcyclists, and V40–49 for vehicle occupants. Information once subcategorized by the major vulnerable road user groups was then disaggregated by gender, age, and region. Regional analyses were performed using location of occurrence (as opposed to location of residence) in order to target high-traffic risk areas.

Crude mortality rates and standard mortality ratios (SMRs; ratio of regional to national road traffic mortality rate) were calculated using population data obtained from the Brazilian Unified Health System (Sistema Único de Saúde; SUS) database (DATASUS). Nationwide Brazilian censuses are done once in 10 years, and the last was conducted in 2010. Population data were interpolated for inter-census years for rate calculations, and region-specific populations were utilized in regional analyses. The national population was used for standardization of regional mortalities, and relevant adjustments were performed for age and/or gender. National and regional gross domestic product (GDP) information was obtained from the 2004–2008 Municipal Gross Domestic Product report (IBGE 2011) of the Brazilian Institute of Geography and Statistics (Instituto Brasileiro de Geografia e Estatistica; IBGE).

RESULTS

Road traffic mortality rates from transport crashes have increased steadily since the year 2000. In 2008, 39,211 deaths due to road traffic injuries were recorded in Brazil, resulting in a crude mortality rate of 20.7 per 100,000 inhabitants. When disaggregating the data by age and gender, the highest rates occur in the 70+ age group (30.9 per 100,000 for both genders). Males have a higher rate in all age groups, peaking at over 50 per 100,000 in the 20- to 29-year age group.

Disaggregating all road traffic deaths into vulnerable road user categories (pedestrians, motorcycle occupants, and bicyclists) shows that in Brazil in 2008, vulnerable road user mortalities account for more than half (51%) of all road traffic deaths and that the majority of the deaths are nearly equally distributed between pedestrian and motorcycle occupant fatalities.

Pedestrian mortality has consistently been highest, averaging 5.46 deaths per 100,000 between 2000 and 2008 (Figure 1). In the past decade, mortality rates for motorcycle occupants have dramatically increased by over 300 percent, from 1.5 per 100,000 in 2000 to 4.7 per 100,000 in 2008. In 2000, the relative proportion of fatalities by pedestrians were more than 3 times that of motorcyclists; however, due to recent trends in rising motorcycle mortality, the relative proportion of fatalities comparing pedestrians to motorcyclists are now almost equal (ratio 1.06 in 2008).

Figure 1 Road traffic mortality rate for vulnerable road users in Brazil, 2000 to 2008.

Table I Road traffic mortality in Brazil by region, 2000 to 2008 average

Region	Population size (2010)	Crude mortality rate (per 100,000 population)	Standardized crude mortality ratio	VRU mortality rate (per 100,000)	Per capita GDP (2008, US$)
North	15,864,454	19.5	0.94	10.0	5552
Northeast	53,081,950	17.8	0.86	9.2	4069
Central-west	14,058,094	29.7	1.43	15.1	11,072
Southeast	80,364,410	19.4	0.94	10.1	11,512
South	27,386,891	26.0	1.26	12.8	9922
National	190,755,799	20.7	Reference	10.6	8690

Regional Analysis

Disaggregating the mortality data by region, the highest crude road traffic mortality rate has consistently been the central-west region, with an average since 2000 of 28.7 per 100,000 population (Table I). The Federal District of Brazil (Brasilia, capital city) is located in this region, which may explain the increased mortality rate. Interestingly, Brazil's largest cities of Sao Paulo and Rio de Janeiro are in the southeast region, which has a relatively lower mortality rate. The northeast region has the lowest road traffic mortality rate, averaging 17.8 per 100,000 population from 2000 to 2008. Notably, this region also has the lowest per capita GDP of any region in Brazil, which may correspond to a lower motorization rate and therefore limit exposure to road traffic crashes. This region has 4.37 road traffic deaths per $1,000 per capita GDP; in contrast, the southeast region has 1.69 crashes per $1000 per capita GDP.

Utilizing national-level data as the reference, the SMR was calculated for each region. The central-west and south regions have SMRs of over 1.4 and 1.2, respectively, indicating that both of these regions have significantly higher road traffic mortality rates than the national average.

Due to the steady increase in mortality rates in vulnerable road users, the ratio of number of road traffic deaths in vulnerable road users vs. nonvulnerable road users has been increasing since 2006. Figure 2 shows the ratio of vulnerable to nonvulnerable road user death rates over time by region. The south, central-west, and southeast regions mirror the national-level increase, such that in contrast to 2000 when there were two thirds as many VRU deaths as non-VRU deaths, by 2008 there were approximately equal numbers of VRU and non-VRU deaths. In contrast, the north and northeast regions had a consistently higher number of VRU than non-VRU deaths over the past decade. Interestingly, these are the regions with the lowest per capita GDPs, which may indicate lower levels of 4-wheeled motorization.

Figure 2 Ratio of vulnerable road user to nonvulnerable road user death rates by region and year (color figure available online).

Pedestrians

In 2008, a total of 9474 pedestrian mortalities occurred in Brazil, resulting in a mortality rate of 5.0 per 100,000 population. This rate was dramatically higher at over 10 per 100,000 in 1996 but then declined steadily until the year 2000; it has since been stable at around 5 per 100,000. The reason for this prior decline is unclear. The majority of 2008 pedestrian fatalities occurred in the southeast region (4288 deaths), and the fewest cases occurred in the northern region (736 deaths).

Stratifying pedestrian deaths by age group shows a dramatic increase in mortality rate with age in each region in Brazil (Figure 3). The mortality rate for the elderly age group (80+ years old) is 20.1 per 100,000, over 10 times that of the youngest age group (0–9 years old).

Figure 3 Pedestrian mortality rate in Brazil by region and age group, 2008 (color figure available online).

Stratifying pedestrian mortality by gender shows a rate of 7.63 per 100,000 for males and 2.44 per 100,000 for females in 2008; the male to female mortality ratio for 2008 was 3.02. Adjusting by age and gender, pedestrian mortality rates peak for both genders in the 80+ age group, at 33.0 and 10.96 per 100,000 population for males and females, respectively.

The national trend of increased pedestrian mortality rates with increases in age is mirrored in each of the 5 regions within Brazil, with a peak mortality rate of 31.8 per 100,000 in the 80+ age group in the central-west region. Unlike with overall road traffic mortality rates, the central-west and north regions have the highest rates of pedestrian mortality.

Motorcyclists

In 2008, motorcycle fatalities resulted in nearly 9000 deaths in Brazil, accounting for nearly a quarter (22.5%) of all transport related deaths, and motorcycles accounted for 20.3 percent of the total registered vehicle fleet. Motorcycle fatalities have been an escalating problem in recent years, with the mortality rate per 100,000 increasing sharply from 1.47 in 2000 to 4.71 in 2008. This has coincided with a dramatic rise in the number of registered motorcycles in Brazil; the number of registered motorcycles more than doubled between 2003 and 2008 (5,332,056 to 11,045,686; increase by a factor of 2.07) and the total fleet increased by a factor of 1.46 in that same period of time.

The age group most affected by motorcycle fatalities has consistently been the 20- to 29-year-old group in the past decade (Figure 4). The 20- to 29-year age group remains most affected by motorcycle deaths, with a peak fatality rate of 10.76 per 100,000 in 2008, compared to 3.63 per 100,000 in 2000.

Figure 4 Motorcycle occupant mortality rate per 100,000 population by age group over time.

The majority of motorcycle deaths occur in males (89.6% in 2008), and gender-specific mortality rates peaked at 19.29 per 100,000 for males and 2.18 per 100,000 for females in the 20- to 29-year-old group. Most motorcycle fatalities occurred in the southeast region of Brazil (3186 deaths), and the fewest in the north region (653 cases) in 2008. Interestingly, before 2006, the northeast region had the most motorcycle fatalities of the regions in Brazil. The highest motorcycle mortality rate occurred in the midwest region (7.70/100,000) in 2008. When stratifying by region and age group, motorcycle deaths in males impacted elderly populations more heavily in both the north and northeast regions (compared to youth populations). Comparing motorcycle fatalities in Brazil by gender in 2008 showed that 8.6 times as many male motorcycle drivers died as female. The southeast region had the highest male-to-female ratio (9.40) and the northern region had the lowest (6.93). Unfortunately, the SIM database does not allow differentiation of motorcycle drivers versus passengers.

Bicyclists

In 2008, bicycle fatalities totaled over 1600 deaths. Nationally, the highest number of these deaths occurred in the 40- to 49-year age group (310 fatalities) and the lowest number was in the 80+ age group (26 deaths). The average mortality rate for bicyclists in Brazil from 2000 to 2008 was significantly lower than that for both pedestrians and motorcycle riders; the national average is 0.75 per 100,000, and the regional average for the central-west is significantly higher at 1.45 per 100,000 (Table II). Over the years, the mortality rate for bicyclists has been increasing slightly, from 0.47 per 100,000 in 2000 and peaking in 2007 at 0.90 per 100,000.

Table II Average bicyclist mortality in Brazil (2000 to 2008)

Region	Mortality rate (per 100,000)	Proportionate mortality (%)
North	0.68	3.90
Northeast	0.51	3.17
Southeast	0.59	3.12
South	1.36	5.39
Central-west	1.45	5.03
National	0.75	3.81

Gender patterns for bicyclist mortalities are similar to that for motorcyclists, with the 2000 to 2008 average male-to-female ratio of 8.39. Age group–specific mortality rates most heavily impact elderly populations, with the 60- to 69-year-old age group having a national mortality rate of 1.67 per 100,000 in 2008. On average, bicyclists in Brazil account for 3.8 percent of road traffic deaths (2000–2008). The region-specific proportional mortality was highest for the central-west region (5.7) and lowest for the northeast region (3.1) in 2008, though the overall average from 2000 to 2008 shows the south region as having the greatest regional proportionate mortality (5.4%).

DISCUSSION

Road traffic deaths are a significant problem in Brazil, resulting in a mortality rate averaging to 19.44 per 100,000 from 2000 to 2008. This is dramatically higher than the rates of nearby countries such as Argentina (9.9 per 100,000) and Chile (10.7 per 100,000; Maffei de Andrade et al. 2008). Pedestrian and motorcycle riders are particularly vulnerable road users, with average mortality rates of 5.46 and 2.98 per 100,000 between 2000 and 2008.

Our analysis shows that elderly pedestrians are at particularly high risk. Brazil's aging index comparing elderly (≥60 years old) to youth (<15 years old) populations has been steadily growing since 2000. Elderly populations are more susceptible to increased mortality due to factors of comorbidity, including decreased physiologic reserve and elevated incidence of preexisting medical conditions (Gawryszewski and Rodrigues 2006). Travel behavior in elderly populations also suggests increases in nonmotorized transport (Ubeda et al. 2011). With a growing elderly population that is more susceptible to specific age-related risks, targeted interventions addressing these risks are needed.

Demographic, economic, and urbanization differences may explain some of the variations in road traffic mortality data between regions. The southeast region of Brazil is the most populous, because Brazil's 2 largest cities Rio de Janeiro and Sao Paulo are in this area. Interestingly, this is not the region with the highest overall road traffic mortality rate. It is unclear whether this is due to improved traffic safety or enforcement in those areas or whether there is a difference in quality of data collection. Highest VRU death rates occur in the central-west and south regions; however, the least prosperous regions (north and northeast) have a higher proportion of VRU (vs. non-VRU) deaths. This is likely a result of there being fewer 4-wheeled motorized traffic in the more rural north and northeast.

This analysis is limited by the quality of data collection and reporting to the SIM national database. The accuracy of death certificate and hospital-coded death data has long been recognized as limited (Glasser 1981; Lu et al. 2001), in particular with road traffic deaths (Harris 1990; Hijar et al. 2012; Lapidus et al. 1994). Previous studies from Brazil have shown that cause-of-death coding in the SIM database can be limited; when reviewing coded deaths from the southern city of Porto Alegre, Fajardo et al. (2009) found that nearly 13 percent needed to be recoded by researchers. In addition, information in SIM is not available in real time; as of Spring 2011, only mortality data through 2008 were available. In addition, we only utilized Ministry of Health mortality data for this study; numbers reported by the police are markedly different. We also did not include crash or injury data in this analysis because this information is collected by different systems using different data gathering techniques. Further study adjusting for differences in data collection would allow for calculation of crash, injury, and case fatality data.

In summary, this analysis highlights the significant burden of road traffic mortality in Brazil and, in particular, for vulnerable road users. Pedestrian safety has not improved since the year 2000. Motorcycle occupant deaths are increasing at an alarming rate coincident with a sharp increase in the size of the motorcycle fleet. Although bicycle mortality rates are the lowest of the VRU groups, they still require targeted intervention to prevent further increase. Road traffic safety interventions should target these obviously high-risk vulnerable road user populations.

ACKNOWLEDGMENTS

The authors thank Dr. Adnan A. Hyder for his support and guidance. This work was conducted as part of the Road Safety in 10 Countries project funded by the Bloomberg Philanthropies.

*Vida no Transito Evaluation Team: Federal University of Minas Gerais, Waleska Teixeira Caiaffa, Celeste de Souza Rodrigues, Amélia Augusta de Lima Friche, Federal University of Rio Grande do Sul, Veralice Maria Gonçalves Fernandes, Pontificial Catholic University of Parana, Samuel Jorge Moysés, and Sandra Lúcia V. Ulinski Aguilera.