Traffi C Accidents in Brazil from 1998 to 2010: Many Changes and Few Effects

The paper describes the situation of road traffi c accidents in Brazil since 1998, when a new Brazilian traffi c law was approved, up to the year 2010. A review of both academic and non-academic literature was carried out, including journals (both indexed and non-indexed), technical reports, author searches, searches in paper reference lists and direct contact with researchers. The main problems related to road traffi c accidents in Brazil identifi ed were the increase in the absolute number of deaths and in the mortality rates, a rapid increase in the number of motorcycles, and drink & driving. Infl uent authors in the fi eld and centers of expertise were identifi ed. Some potential solutions are presented by the authors, who suggest that the public offi ces related to traffi c regulation and control are not taking suitable measures for control and reduction of road traffi c accidents.


INTRODUCTION
Around 1.2 million people die worldwide from road traffi c accidents (RTA) every year and more than 90% of these deaths occur in low-and middle-income countries.a In addition to people's suffering from deaths and physical disabilities, RTA also pose a heavy economic burden on health systems.The World Health Organization (WHO) estimates that annual losses due to RTA exceed $500 billion.b In Brazil, the annual number of killed and severely injured people exceeds 150,000 c and the Brazilian Institute of Applied Economic Research (IPEA) has estimated that total RTA costs amount to R$ 28 billion per year.d,e Brazil ranks high among countries with the world's most dangerous roads and it has sought to reduce the high number of accidents in recent years.After the implementation of the new Brazilian Vehicle Code (BVC) f in 1998 RTA mortality rate has remained stable (around 20 deaths per 100,000 inhabitants), which is higher than the rates reported in Japan, Sweden and Canada (fi ve to eight deaths per 100,000).a New laws, local traffi c control, improved vehicle safety and increased electronic traffi c monitoring has all failed to signifi cantly reduce deaths and disabilities.This article aimed to present a panorama of road traffi c since the implementation of the new BVC.

METHODS
In light of the small number of scientifi c publications on RTA and low interaction between areas such of knowledge such as epidemiology, engineering, psychology, medicine and others, a descriptive review was conducted with extensive literature search.This review study included scientifi c articles and publications in Portuguese, English and Spanish published between 1998 and 2010 with data collected in Brazil and that evaluated RTA as a primary or secondary outcome.There was conducted a search of indexed journals in electronic databases, non-indexed journals, technical reports, search by specifi c authors, reference lists of articles and communications with researchers.The electronic search was conducted in MEDLINE (National Library of Medicine), LILACS (Latin American and Caribbean Health Sciences), SciELO (Scientifi c Electronic Library Online) and Google Scholar databases.It included the following descriptors: "traffic accidents", "road accidents", "epidemiology" and "external causes".The descriptor "Brazil" was added to restrict the search to publications in Brazil.There was no restriction on demographic and socioeconomic factors and types of studies.The search for articles was carried out between February and October 2010.Publications were pre-selected based on their titles and abstracts.The full text of the publications selected was obtained and examined.Brazilian websites such as the National Traffic Department (Denatran), National Department of Transport and Infrastructure (DNIT), Brazilian Federal Highway Authority (DPRF) and Brazilian Ministry of Health Mortality Database (Datasus) were searched.Manual search for non-indexed journals was also carried out at the Universidade Federal de Pelotas (UFPEL) and Universidade Católica de Pelotas (UCPEL) libraries.
The data here presented follow the WHO 10th Revision of the International Classifi cation of Diseases and Related Health Problems (ICD-10).g Data on mortality and morbidity (hospital admission) between 1998 and 2005 was obtained from the "Traffi c Accidents in Brazil -An Atlas of their Distribution," a publication of the Brazilian Traffi c Medicine Association (ABRAMET).h Data on mortality and morbidity between 2006 and 2008 was obtained from Datasus database.i The 2008 mortality data are preliminary and morbidity data include the year of 2009.The DENATRAN j provided information on the motor vehicle fl eet.

DATA AND TRENDS FROM 1998 TO 2008
Law No. 9503 established in 1998 -the fi rst year of this review study-the new BVC.f The new BVC was regarded as a promising action to reduce the growing number of RTA and address this issue.It included stringent laws, higher fi nes and chapters on traffi c education in traffi c.However, RTA mortality and hospital admission rates remained the same or even increased.Figure 1 shows mortality rates from RTA with denominators commonly used in the Brazilian scientifi c literature (deaths per 100,000 inhabitants and per 10,000 vehicles).
The mortality rate from RTA declined during the fi rst two years of follow-up (from 19.1 deaths per 100,000 in 1998 to 17.1 in 2000).There was a steady increase between 2001 and 2004, reaching 19.6 deaths per 100,000, which remained stable and close to 20 deaths per 100,000 in subsequent years.The RTA mortality rate per vehicles remained stable (around 10 deaths per 100,000 vehicles) until 2002.From 2003 to 2008 there was a slow decrease, 6.7 deaths per 100,000.This decrease may be explained by an 85% increase of the Figure 2 shows RTA according to categories of victims."Other" included occupants of three-wheeled motor vehicle, heavy vehicles (trucks), and buses, other road accidents and unspecifi ed road accidents.The category "Car occupants" included occupants of vans.
In "Other" category there were approximately 14,500 deaths in the beginning of the study period, accounting for 45% of all deaths.After a steady decline, there were seen 9,603 deaths in 2008, just over 26%.This decrease may be resulting from improved data quality as it also included unspecifi ed road accidents.
There was a 12% reduction in relative pedestrian deaths (36% in 1998 to 24% in 2008).However, the number of deaths remained quite constant (from 9,000 to 10,000 deaths/year).Similar to other countries, Brazilian males died more from RTA than females (approximately 4.5 male: female).There was a reduction of RTA mortality rate among males from 1998 to 2000, reaching the lowest rate of the study period (28 deaths per 100,000 males) (Figure 3).After a steady rise, it reached 33 deaths per 100,000 in 2007.Among females, after a decline

MOTORCYCLES
Motorcycles have taken up urban spaces as an efficient means of transportation and work to avoid traffi c congestion in large cities.Defi cient public transport, the growing delivery service industry, the possibility of generating income for unskilled young people and ease of purchasing a motorcycle have all lead to a 5-fold increase in motorcycle fl eet compared to car fl eet.There are more than 14 million circulating motorcycles, accounting for 25% of the national fl eet.j The motorcycle has become the most popular single means of transportation in Brazil. 36k Motorcycle drivers are a priority group in prevention programs 9,11,59 because they are seven times more at risk of dying, four times more likely of being injured and two times more likely to hit a pedestrian compared to car drivers. 94 The motorcycle as a means of work is regarded as a potential factor for increased motorcycle casualties, 2,9,11,59 and motorcycle couriers are at high risk of RTA due to their job demands. 19,36,64,73In São Paulo k and Porto Alegre 74 37% of deaths and 28% of emergency hospital admissions of motorcyclists were among motorcycle couriers.Pressure from customers and employers for fast deliveries, 19,73 pay on the basis of productivity and daily work shifts exceeding 10 hours 19,64 have been identifi ed as determinants of inappropriate driving.Increased productivity is a justifi cation for their disrespect for traffi c laws. 19,32,65,73Typical behaviors of young males including boldness, immaturity, a sense of invulnerability, a tendency to overestimate their abilities and driving inexperience allied to the need to become part of a group make them push the limits and transgress them, 1 which makes it more complex the issue of RTA.The executive veto of BVC Article 56 that prohibited driving between the lanes of traffi c on the grounds of offering "greater fl exibility of movement" further jeopardizes the safety of road users.
The expanding market of delivery services and motorcycle taxis indicates that these workers' job is demanded in our society.A qualitative study conducted in Porto Alegre showed the social dichotomy involving motorcycle couriers: while they are discriminated as drivers, their services are needed for fast deliveries. 73his pressure -recognizably a causing factor for traffi c accidents-is key to secure their jobs and is encouraged by employers and clients. 65niz et al 19 have suggested the work reality of these workers outgrows the thrill of riding a motorcycle.They often have to cope with poor work conditions, lack of formal contracts, long working hours, stress, psychological pressure for greater productivity, low income and high risk of accidents.Motorcycle couriers live with the dialectic of risk-need as described by Veronese & Oliveira: 73 "To work at risk of an RTA or not to work, that's the question.All that motorcycle couriers can do is to manage the risk." The phenomenon of motorcycle couriers is a major issue for traffi c authorities in Brazil. 14  Employers and clients have to take part of this responsibility, reducing their demands for deliveries that involve unsafe behaviors and extenuating working hours. 32,73Companies must be required to hire professional motorcycle couriers who have been trained on defensive driving and use of safety equipment, in addition to greater motorcycle inspection and use of speed control devices. 32,73However, a new debate on BVC Article 56 14 and targeted actions to promote motorcycle courier behavior change 73 are unlikely to be effective.Vasconcellos said the motorcycle-related deaths have cancelled out the gains achieved with BVC in reducing RTA by other vehicles.k The motorcycle has defi nitely a social role, either at work, for commuting or during leisure time, and occupies 82% less public space than cars. 63Its new status is part of an irreversible change of the motor vehicle fl eet composition. 36The challenge is to ensure safety of road users.

DRINKING AND DRIVING
The association of alcohol and RTA has been well documented in the literature.It is a major cause of morbidity and mortality affecting mostly young males.l The First Household Survey on Patterns of Alcohol Consumption conducted in 143 Brazilian cities in 2009 showed a 35% prevalence of drinking and driving (43% among men and 9% among women). 54m Among non-fatal injuries treated at trauma care centers and emergency rooms in São Paulo and Uberlândia (southeastern state of Minas Gerais), the prevalence of alcohol use was 24% and 29%, respectively, 24,27 Alcohol use was suspected in nearly 17% of RTA cases seen at emergency services in the cities covered by the Sentinel Surveillance System on Violence and Injuries. 37der the 1998 BVC driving under the infl uence of alcohol (BAC ≥0.6 grams of alcohol per liter of blood) is a criminal offense.Law No. 11,705 that went into effect on June 20, 2008 changed blood alcohol limit for drivers to zero (with a tolerance of up to 0.2 g/L) punishable with more strict penalties up to imprisonment in fl agrante delicto if BAC >0.6 g/L.Few studies have assessed the effectiveness of the Prohibition Act or Zero Tolerance law.Mello Jorge & Koizumi 47 found 28% reduction in hospital admissions between the two halves of 2008, as well as signifi cantly reduced hospital stay (42%), hospital costs (39.2%) and mortality (13.6%).In the city of São Paulo, there was a 48% to 36% reduction in BAC in fatal injuries when compared data from the second half of 2007 and 2008. 28Despite not having been based on a time series that could support or refute the fi ndings, the results suggest that the new law was effective.
Moura et al 50 monitored through phone calls the frequency adults drove after using alcohol in 27 Brazilian cities.They found a decrease in the months immediately following the enactment of the law but further increase was seen in the following months.These studies, together with morbidity data from the Brazilian National Health System Database (SIH-SUS), suggest that the Zero Tolerance law of 2008 was effective in reducing alcohol-related RTA for a limited period.
The city of Porto Alegre passed a law against alcohol consumption at gas stations in 2006.De Boni et al 17 showed no difference in BAC among groups of young patrons pre-and post-law enforcement.Studies using breathalyzers to estimate BAC at sobriety checkpoints on weekends showed very high prevalences in other two cities.In the city of Belo Horizonte BAC was detected in 38% of drivers and was above the legal limit in 20%. 13In the city of Diadema, state of São Paulo, the proportions were 24% and 19%, respectively, which are up to six times higher than those reported in international studies. 21A study conducted post-law enforcement with individuals potentially at risk for alcohol-related RTA found 51% with BAC and they intended to drive.n Studies 13,17,21,n pointed defi cient oversight as a factor for failed law enforcement, in addition to high prevalence of BAC.Pinsky et al 56 showed that 85% of young people applying for a driver's license did not know anyone who has been legally punished for drinking and driving, 74% believed that no offender would receive the statutory penalty and 64% said it was very unlikely they would be stopped by police or subject to penalties.A study conducted post-enforcement of the Zero Tolerance law found that, while 86% of respondents said they had drunk and driven in the last year, 9% were stopped and asked to perform a breathalyzer test.n Twenty percent of professional drivers working in Brazilian federal highways said they had been asked to take the breathalyzer test while 9% and 8% of car and motorcycle drivers were asked to do so, respectively.BAC was detected in 1% of bus drivers to up to 5% of truck drivers.o Souza et al p reported that 47% of RTA cost is related to excess alcohol use in Porto Alegre.Mello Jorge & Koizumi 47 showed 28% reduction in hospital admissions and 36% cost reduction with a public saving of more than R$ 23 million after the enforcement of Zero Tolerance law.They stressed this law has saved lives since it came into force.
o The Zero Tolerance law would not be required if there were effective oversight and enforcement of the previous law.Countries with low RTA mortality rates (<8 deaths per 100,000) enforce BAC limits close to that provided in the 1998 BVC such as 0.3 g/L in Japan,0.5 g/L in France, 0.5 g/L in Germany, 0.5 g/L in the Netherlands, 0.8 g/L in England, and 0.8 g/L in Canada.a They all have strict enforcement of alcohol laws.
The Global Road Safety Partnership (GRSP) q recommends actions to reduce alcohol-related accidents such as political commitment to reduce drinking and driving, clear laws for alcohol limits and penalties for offenders, mass campaigns and dissemination of law enforcement, education to change behavior and strict penalties for offenders.
WHO recommends the implementation of sobriety checkpoints with breathalyzer tests, which is a costeffective action than can reduce accidents by 20%.a They should be operated primarily on weekend nights when most alcohol-related RTA occur. 24,28,49cording to the Brazilian Consensus on Public Policies on Alcohol 29 (2004), evidence-based measures are an approach with great potential for success in preventing alcohol-related RTA and should include low BAC tolerance, evident and frequent oversight, suspension of offenders' driver's license and increasing the certainty of punishment through spot checks (oversight of checkpoints).
It is critical to improve control and oversight of alcohol sale and advertising.Beer must be categorized as an "alcoholic beverage" and subject to restrictions on advertising and broadcasting times since the greatest investment are made in alcohol products. 34,57e First Household Survey on Patterns of Alcohol Consumption in the Brazilian population showed that people support public policies that address the drinking and driving problem including increasing taxes on alcohol, restrictions on alcohol advertising and alcohol industry sponsorship of cultural and sports events and stricter penalties.r Nascimento & Garcia 51 have examined many public policies, emphasizing the need to defi nitely set the alcohol issue high in the Brazilian political agenda.Yet the power of the parties of interest including alcohol industry, retailers and consumers should not be underestimated.These policies must be implemented as the people's wish should prevail.

PEDESTRIANS AND CYCLISTS
Pedestrians account for almost one-third of deaths from RTA in Brazil (approximately 10,000 deaths per year) and an increasing number of deaths are seen among cyclists (over 1,500 in 2005).h Pedestrians are the third largest group of victims, following motorcyclists and car occupants in southern Brazil, but they are the leading group in casualties (38% of deaths and 11% mortality rate). 9There is high underreporting of RTA in pedestrians and cyclists, 53% and 33%, respectively, with a corrected mortality rate of 5%, the highest for both groups.Among injured pedestrians 70% were children or elderly. 9 the city of Maringá, southern state of Paraná, 68 pedestrians and cyclists had a 119% and 65% higher risk of RTA hospital admission, respectively, compared to car occupants.In 1998, 47% of the deaths occurred in pedestrians and cyclists 62 and almost 90% of deaths in elderly (65 years or more) were caused by run overs.Andrade & Mello Jorge 2 found that even if pedestrians are 12% of total RTA in Londrina, southern state of Paraná, they account for almost one-third of the deaths and have the highest rates of hospital admission (18%) and mortality (5%).Soares & Soares study 67 showed that children and elderly are the main groups of injured pedestrians, with high rates of RTA incidence and mortality among the elderly (256 and 44 per 100,000, respectively).Pedestrians accounted for 42% while cyclists accounted for 15% of RTA deaths in the city of Fortaleza, northeastern state of Ceará, in 2001-2002, These two groups accounted for 38% of all injuries. 16e elderly are the main injured group among pedestrians.Gawryszewski et al 26  Ribeirão Preto, southeastern state of São Paulo, 41% of RTA in individuals over 60 were in pedestrians, of which 60% had sequelae due to their injuries, and 12% were cyclists. 66About 52% of the elderly injured were pedestrians and 13% were cyclists. 71They were all able to perform daily activities before but RTA have prevented them from their activities causing serious fi nancial diffi culties.The authors noted the importance of the elderly in society: many work in the community, care for relatives, support their families, work and consume goods and services.In the city of Londrina, of 121 elderly who were injured and treated in hospitals, 9% died and the main underlying cause was run over.According to the families, the elderly were totally independent before they were injured. 12milar to the elderly, child pedestrians are frequently injured.There were more than 1,000 deaths per year due to run overs in children under 15 years between 2000 and 2005 (2.2 deaths per 100,000 inhabitants). 15The male-to-female ratio was 1.8:1 and the most affected age group was fi ve to nine.Because of poor data quality or run away from the accident scene 83% of reports did not specify the causing vehicle in 2000, decreasing to 70% in 2005.Among the vehicles specifi ed, there was a higher proportion of passenger vehicles, and the highest number of deaths occurred in pedestrians under 15 when compared to other RTA in this age group (64% in 2000 and 67% in 2005).In the city of Uberlândia, 46% of injured victims from zero to 14 years of age 25 were cyclists and 31% pedestrians.Among the most severely injured (admission to an intensive care unit), 81% were pedestrians and 7% cyclists.Two-thirds of deaths occurred in pedestrians.The large proportion of RTA involving bikes can be possibly explained by the fact that they are commonly used for recreation in the fl at, sunny streets of Uberlândia.In Londrina, most injured people were cyclists (66%) and pedestrians (20%) and the rates of hospital admission were 38% and 35%, respectively.Of eight deaths reported, fi ve were pedestrians and two cyclists. 35th an estimated fl eet of 75 million units, bikes are the main means of transportation among workers in Brazil.s In the city of Pelotas, southern state of Rio Grande do Sul, 17% of the working population use bikes for daily commute.It is the main means of transportation among men (27%). 6Over a 12-month period, 6% of cyclists have suffered some RTA during work commute, a proportion that is higher than that reported in offi cial local data.In Maringá, cyclists accounted for 18% of deaths from RTA in 1992.The use of bikes as a means of transportation is common in Brazilian inner cities with mild temperatures and fl at topography, and is an economic means especially among lower income population. 62st studies have recommended measures to reduce RTA in pedestrians and cyclists, from the reassessment of public policies that encourage one-person-per-car commute 2 to the design of more friendly public spaces to pedestrians especially children and elderly. 2,15,68The main recommendations include creating safe spaces for recreation, 35 educational programs for drivers 15,66 children, parents 15 and the elderly. 66Faria & Braga 22 (1999) proposed a program of actions to reduce the risk of RTA involving children and adolescents, especially pedestrians.These recommendations included assessment of routes of circulation, traffi c engineering and educational activities at pedestrian crossings, improved monitoring and a new proposal for traffi c education.
The creation of exclusive spaces for bicycle travel (bicycle paths and lanes) is the main recommendation for increasing the safety of cyclists, 6,8,35,62 besides increasing public awareness about traffi c rules. 6Studies conducted in Pelotas have shown that behavior change of cyclists and educational interventions will not have a signifi cant impact on reducing RTA in the city. 7,8though it makes no reference to RTA, it is worth mentioning the publication of the Brazilian Program for Bike Mobility -Bike Brazil by the Brazilian Ministry of Cities.s

FREIGHT AND PASSENGER TRANSPORTATION
Sixty-two percent of freight and 96% of passengers are transported in the Brazilian highway system, t with a fl eet estimated at more than two million trucks and about 700,000 buses and minibuses.j Between 2004 and 2007 road traffi c accidents involving trucks and freight vehicles on federal highways increased 14%, from 40,107 to 45,833 with 3,124 deaths and more than 20,000 injured.u In the United States 25 truck drivers per 10,000 die in road accidents every year compared to 281 per 10,000 in Brazil. 14Truck drivers fall into two categories: employees of truckload services companies with better work conditions and safety, and self-employed. 14In addition to their stressful job, selfemployed truck drivers have to deal with low pay, little or no training, great pressure to deliver their freight on time, and excess workload. 14However, the association with RTA cannot be verifi ed due to a lack of scientifi c publications on these workers working on the Brazilian Studies with truck drivers at gas stations and roadside restaurants showed that two-thirds would drive nine hours or more per day, 50% of them over 12 hours a day. 52Another study showed 43% of truck drivers would drive more than 16 hours a day. 70This work routine is associated with high prevalences of excessive daytime sleepiness (32% 70 to 46% 55 ) and RTA.Many truck drivers use amphetamines to reduce their fatigue on long journeys.Nascimento et al 52 reported that two-thirds of truck drivers use amphetamines: 27% reported daily use and 60% reported using it two to three times a week.Most (76%) reported using it to get to their destination in less time and 27% reported road accidents associated to its use.Amphetamines are purchased at gas stations (58%), drugstores (38%), and in their truck companies (8%).Nearly half of truck companies in the southern state of Santa Catarina reported drug use by drivers in 2004; 65% of them reported amphetamine use.Most reported RTA causing deaths or permanent disabilities. 77study showed that 91% of truck drivers interviewed reported alcohol use during working hours, 24% of them reported daily use.Forty-three percent of drivers reported gas stations as their preferred place of consumption, and 17% who reported alcohol use were involved in road traffi c accidents. 52Seventy-one percent of truck drivers working in the metropolitan areas of 26 Brazilian capitals reported at least one episode of binge drinking in the last year.About 10% had used alcohol on the day of the interview and 9% reported drinking and driving.o There were found no studies on RTA associated with passenger transportation.Data from the Federal Highway Authority, apud Correa, 14 showed that 4% of more than 177,000 RTA occurring on federal highways involved buses and minibuses in 2004.These accidents kill at least 2,000 people every year. 14A study found that 42% of bus drivers were involved in interstate road accidents and in 8% they were associated with excessive sleepiness. 75Excessive daytime sleepiness was found in 28% of drivers, 48% reported sleepiness while driving and 12% took daily medication to help them stay awake.It should be noted that these proportions may be underestimated since these interviews were conducted in the transportation company.Another study found that 16% of the drivers interviewed reported having fallen asleep at the wheel of the bus while on duty. 48searchers have recommended prevention programs, increased oversight of roads on the sale of alcohol and amphetamines, 52,70 and implementation of education programs, specifi c treatments 55 and management of sleep disorders. 75

DATA QUALITY
Mello Jorge & Koizumi 43 have asserted that knowledge, research and understanding of these issues would allow the implementation of actions aimed at changing the current panorama of road traffi c in Brazil.RTA data come from three main data sources: police reports, SIH-SUS and Mortality Database (SIM).Data from these sources are not fully complete and accurate and are neither complementary nor interconnected. 43e police report is a record of RTA fi led at civil police and civil-military federal highway authority stations.This is not a standard record systematized at the national level. 43Studies with police reports have demonstrated signifi cant underreporting of RTA and deaths.Barros et al 9 verifi ed that 39% of RTA in Pelotas were underreported, varying according to the category of the victim, type of injury and time of occurrence.Andrade & Mello Jorge 3 reported only 33% of police reports compared to records of emergency room and hospital admissions and death certifi cates in Londrina.Police report coverage was 72% of accidents involving car occupants and 8% and 25% of those involving cyclists and pedestrians, respectively.These authors have argued that police reports can provide major information on RTA that is crucial for the development of prevention actions.
Although SIH-SUS includes only data from hospital admissions paid by the Brazilian National Health System (SUS), these data are regularly updated and readily available. 58All information is collected from hospital admission authorizations (AIH).The AIH form has been modifi ed in 1998 to include a required specifi c fi eld to code "secondary diagnosis" or injury morbidity data from external causes.A 2008 study assessed the quality of AIH data by comparing the coding of primary and secondary diagnoses in SIH-SUS and medical records. 41RTA accounted for 32% of hospital admissions showing an optimal level of agreement (kappa = 0.90).But hospital admissions due to external causes coded as natural causes in the database were not included in the study.Tomimatsu et al 72 assessed coverage and quality of hospital admission information in Londrina and Maringá based on AIH medical reports.In Londrina there was high underreporting of RTA with an increase in the number of hospital admissions to 988 from 492 after reclassifi cation.They also found underreporting of hospital admissions due to external causes (41% in Londrina and 25% in Maringá).Researchers have argued that low appreciation and use of epidemiological information in the medical curricula could explain data underreporting.It is being implemented an intervention study that proposes awareness-raising actions and training of doctors and hospital coders on the importance of proper reporting of information in SUS-affi liated hospitals. 4eath certifi cates are SIM core data form.For deaths from external causes, a death certifi cate is completed by medical examiners at Institutes of Forensic Medicine (IMLs) who is ethically and legally responsible for providing complete death certifi cates according to the Brazilian Ministry of Health.v Unlike police reports, deaths are recorded in the SIM regardless of the time between the RTA and death, resulting in inconsistent data. 43ICD-10 allows to include information on RTA victim category (pedestrian, cyclist, motorcyclist, and others) and characteristics.A study with death certificate data collected in 1996 following the implementation of ICD-10 in Brazil showed that RTA was not reported as a cause of death in about 35% of deaths and 20% of death certifi cates did not include information on the category of victim. 3Limited knowledge and/ or low motivation of health providers could explain inadequate quality of data.Also in 1996, Drumond Jr. et al 20 found that the type of accident could be identifi ed in 53% of deaths from police reports sent with the bodies, and 33% were RTA suggesting that IMLs do not use information available at their own institute.A study carried out in 15 Brazilian cities based on data from IMLs, police stations and victim's homes showed an increase by 90% of RTA deaths compared to that reported in death certifi cates only. 44The authors argued that IMLs have data that allow to identify the cause of death but they are not included in death certifi cates.Jorge Mello et al have suggested there is a need to focus on the search for information, which is usually available but have to be retrieved. 46Matos et al 38 identifi ed 83% of the 70 deaths classifi ed as "unspecifi ed injury" by combining IML and SIM data from the Municipal Health Department of Belo Horizonte, and 41% were RTA-related.The cause of death was identifi ed in 71% of events of undetermined intent, and RTA were the most common cause, resulting in a 6% increase in deaths from run overs and 33% from car accidents.This study suggests the need to improve coding and selection of underlying cause, death certifi cate completion and quality of police and medical information.In the city of São Jose dos Campos, southeastern state of São Paulo, missing information on death certifi cates prevented the estimate of mortality rates by type of victim.Unspecified RTA was recorded in 57% to 70% of death certifi cates. 39A study in the state Paraná using SIM data showed a reduction in the proportion of deaths due to events of undetermined intent, 22% in 1980 to 2% in 2005. 33The reduction of 15 deaths per 100,000 during the triennium 1979-1981 to two deaths in 2003-2005 suggests improved quality of records as a result of local data processing (1993) and systematic training of technical coders.According to Jorge Mello et al, 46 SIM has greatly improved coverage and quality of data and is expected to have a coverage close to 100%.There is a need to encourage, raise awareness and guide managers, conduct research and learn from successful experiences to improve quality of data.

ROAD TRAFFIC ACCIDENT COSTS
Deslandes et al 18 estimated emergency room costs due for violence-related injuries in the city of Rio de Janeiro in 1996 in two local public hospitals, Miguel Couto Hospital (HMMC) and Salgado Filho Hospital (HMSF).RTA accounted for 67% and 46% of the costs in HMMC and HMSF, respectively, from R$ 78.00 to 236.00 for run overs and R$ 74.00 to 119.00 for other RTA, with an annual expenditure close to R$ 1 million.
The fi nancial impact on injured patients was assessed based on AIH records obtained from a university hospital in the city of Curitiba, southern state of Paraná. 76Between September 2000 and February 2001 RTA accounted for 54% of hospital admissions amounting to over US$ 110,000.The cost of each hospital admission was on average US$ 600, ranging from US$ 21 to 2,722.
Mello Jorge & Koizumi 45 estimated costs of hospital admissions due to external causes based on SIH data in the state of São Paulo in 2000.Because of costly procedures, external causes incurred in higher average cost and daily cost (about 4% and 50%, respectively) compared to patients admitted due to natural causes.RTA accounted for 17% of hospital admissions due to external causes and 26% of deaths post-admission.The average hospital stay was 5.7 days and hospital mortality rate was 4.7%, both above the average rates from other external causes.The average cost was R$ 721.15 for discharged patients, R$ 1,500.79 for deceased patients and the daily cost was R$ 132.18  d The total cost was estimated at R$ 3.6 billion in 49 large urban areas and R$ 5.3 billion for all urban areas in Brazil.Loss of productivity accounted for 43% of the total cost of RTA, followed by property damage (30%), medical and hospital costs (16%), and other costs (11%) such as legal proceedings, removal of vehicles, police assistance, among others.Cars, which are 74% of the motor vehicle fl eet, accounted for 56% of all costs.Motorcycles, urban buses and trucks, which are a lower percentage of the fl eet, accounted for a high proportion of costs (11% to 19%, 1% to 13%, and 4% to 11%, respectively).A RTA had an average cost of R$ 8,782.49; a RTA without injured victims cost on average R$ 3,262.00;with injured victims more than R$ 17,000; and with deaths R$ 144,478.00.Although accidents with victims (injured or dead) represented 14% of all accidents, they accounted for 69% of the total cost (almost R$ 2.5 billion).
In 2006 the IPEA and the ANTP conducted a new study on the costs of RTA on Brazilian highways.e More than 100,000 RTA on federal highways had an estimated total cost of R$ 6.5 billion in 2005.Victim costs accounted for 68%.Loss of productivity and health care were the main components of cost while vehicle costs represented 31%.Accidents involving fatalities had an average cost of R$ 418,000 while accidents with injured victims and without victims had an average cost of R$ 86,000 and R$ 17,000, respectively.RTA costs in state highways, as determined by adjustment models, were estimated at R$ 14.1 billion and R$ 1. 4   69 This study evaluated mainly the effectiveness and impact of this intervention using a quantitative approach to outline the epidemiological scenario of RTA and a qualitative approach (interviews with managers, program implementation staff, users, among others, and fi eld observations).PRMMAT implementation in the fi ve cities studied occurred at different time points from October 2003 to August 2004.The authors pointed out weaknesses including lack of familiarity with intersectoral work; poor visibility of RTA as a problem; frequent changes in coordination; and political interference in the program projects.The strengths included the experience of local departments in dealing with RTA; the process of intersectoral coordination; raising public awareness on RTA; and the production of a set of indicators and methods for program evaluation.
There is no assessment of the effectiveness of PRMMAT based on RTA mortality and morbidity reduction data.PRMMAT reduction goals to be achieved and their terms have not been reported.

CONCLUSIONS
Although scientifi c production on RTA is still modest in Brazil, there was a 450% growth in publications during the review period, from 22 until 1997 to 100 from 1998 to 2010 (using the descriptor "traffi c accidents" in the SciELO database).It is of note focal points of knowledge production and leading authors in the areas studied, particularly the production of an atlas of RTA w Ministério da Saúde, Secretaria de Políticas de Saúde, Projeto Promoção da Saúde.Programa de Redução da Morbimortalidade por Acidentes de Trânsito: mobilizando a sociedade e promovendo a saúde.Rev Saude Publica 2002;36(1):114-7.DOI:10.1590/S0034-89102002000100018distribution in Brazil.h However, there are gaps in RTA research in Brazil including ongoing studies on accidents and traffi c fl ow on highways, use of more reliable denominators for risk comparison (e.g., per mileage traveled) and assessment of intervention effectiveness.
Brazil is off track in the fi ght against RTA which is evidenced by a growing number of deaths and sustained high mortality and hospital admission rates.The BVC had at its implementation the required legal framework to produce an effective reduction of RTA.Over the years factors including modern electronic data systems, improved technologies (breathalyzers, radars, and electronic speed control devices), improved vehicle safety, and development of traffi c engineering and pre-hospital emergency services have all contributed to increase its effectiveness.In addition, several cities have invested in traffi c safety, highways were privatized and properly maintained and the media has given due visibility to the problem helping to raise public awareness.
The process of reducing RTA has been hindered by several actions, or lack of actions, including the requirement to warn the public on the location of traffi c control devices, veto of a law prohibiting motorcycles to ride between lanes of traffi c, reduced penalties and payment of fi nes in installments, non-implementation of a traffi c education and mild restrictions on advertising related to drinking and driving, among others.Marin & Queiroz (2000)  34 stressed that state authorities shall have the responsibility for the implementation of effective public policies and law enforcement.The original BVC and most recently Law No. 11,705/08 have constituted legal advancements.Yet, public policies and law enforcement lacked strength with minor actions or actions limited to short periods after their implementation.Studies on the implementation of the new BVC have emphasized that strict law enforcement and punitive measures may be effective in reducing traffi c accidents 10,23 and increasing traffi c safety practices. 31ndrade et al 5 demonstrated that the greatest impact on RTA mortality reduction observed from 1994 to 2005 was a result of the implementation of the new BVC.But there were no major advances in reducing mortality after 1999, possibly due to loosening law enforcement and penalties provided for under the new code.
Studies that have shown RTA reduction after implementation of BVC and Law No. 11,705/08 suggest that state authorities have the power to reduce a great deal of accidents and save lives.It is key to have ongoing vigorous actions.Educational actions are also important and should be widely implemented, but they are not effective as a single measure. 53Improvements in public transport, investment in alternative healthier means of transport such as bikes and promotion of rational use of private cars are essential to improve quality of life in cities and reduce traffi c accidents.
The signifi cant increase in the car and motorcycle fl eet fueled by low interest loans and economic growth and coupled with the rapid population aging are enormous challenges to the current national traffi c policy.In 2004 RTAs were the ninth major cause of death worldwide with 1.2 million victims.The WHO estimates that RTA will be the fi fth leading cause of death in 2030, killing more than 2 million people.a The way Brazil will address this issue can either help reduce this burden or confi rm this prediction.
By the end of 1998, a total of 30,890 people died from RTA.It increased by 19% in 2008 (36,666 deaths) while the Brazilian population increased 17%.From 1998 to 2000 there was a decrease in the fatality rate (28,995 deaths) but it increased steadily in the following years, amounting to 37,407 deaths in 2007.

ROAD TRAFFIC ACCIDENT MORBIDITY AND MORTALITY REDUCTION PROGRAM The
billion in local highways.RTA on Brazilian highways cost about R$ 22 billion every year.The annual RTA costs in Brazil exceed R$ 27.2 billion.d,e 2001 Brazilian Ministry of Health Decree 737 established the National Policy for the Reduction of Morbidity and Mortality from Violence and Injuries.The Program for Road Traffi c Accident Morbidity and Mortality Reduction: Mobilizing Society and Promoting Health (PRMMAT) w was launched to put into practice the new public policy.Its starting point was the Brazilian Vehicle Code.The program aimed to implement health promotion activities in partnership with government, non-government organizations and society in large urban areas with extensive road systems and high rate of accidents.It was coordinated by the Brazilian Ministry of Health and implemented at state and local levels and its priority was intersectoral actions with involvement of civil society.The areas of actions were defi ned based on the magnitude, risk and relevance of RTA in overall mortality.There were selected 26 urban areas, covering 84 municipalities in 14 states, totaling over 46 million people.Between 2003 and 2006 the Latin American Center for Studies on Violence and Health, the National School of Public Health, Fundação Instituto Oswaldo Cruz and the Brazilian Ministry of Health evaluated the implementation of PRMMAT in fi ve cities: Recife, Belo Horizonte, Goiânia (central-west Brazil), São Paulo and Curitiba.