Abstract

INTRODUCTION

Traffic accidents (TA) are an important public health problem worldwide that requires joint efforts for effective and sustainable prevention¹1. World Health Organization. Global status report on road safety: time for action. Geneva: World Health Organization; 2009.. Brazil ranks fifth among countries with the highest rates for traffic accident-related deaths, with estimation of 40,000 deaths per year²2. Ministério da Saúde. DATASUS. Disponível em: http://www2.datasus.gov.br/DATASUS/index.php?area=02. (Acessado em 08 de julho de 2014).
http://www2.datasus.gov.br/DATASUS/index... that, added to seriously injured victims, account for over 150 thousand people and total costs around R$ 28 billion per year³3. Bacchieri G, Barros AJD. Acidentes de trânsito no Brasil de 1998 a 2010: muitas mudanças e poucos resultados. Rev Saúde Pública 2011; 45(5): 949-63.. The most relevant impacts of CA morbimortality are seen among young males⁴4. Morais Neto OL, Montenegro MMS, Monteiro RA, Siqueira Junior JB, Silva MMA, Lima CM, et al. Mortalidade por acidentes de transporte terrestre no Brasil na última década: tendência e aglomerados de risco. Ciênc Saúde Coletiva 2012; 17(9): 2223-36., with great repercussion to the health system.

The development and implementation of traffic accident prevention policies depend directly on information aimed for surveillance and monitoring systems, and on the analysis of situations in order to determine the magnitude of occurrences, victims' profile, means of transportation involved, and the mapping of risk areas⁵5. Cabral APS, Souza WV, Lima MLC. Serviço de Atendimento Móvel de Urgência: um observatório dos acidentes de transportes terrestre em nível local. Rev Bras Epidemiol 2011; 14(1): 3-14.. Such information is not fully available in most countries, Brazil included.

Measures adopted in the country to reduce CA morbimortality have not presented the expected results³3. Bacchieri G, Barros AJD. Acidentes de trânsito no Brasil de 1998 a 2010: muitas mudanças e poucos resultados. Rev Saúde Pública 2011; 45(5): 949-63.. Among the reasons pointed in literature, data about victims, which can support planning and monitoring, are spread in different databases and, in isolation, are not able to show the actual magnitude of the problem. The World Health Organization (WHO) recommends that more than one data source be used in such cases as a shorter and less tortuous path to a broader understanding of health problems⁶6. Gawryszewski VP, Kahn T, Mello Jorge MHP. Informações sobre homicídios e sua integração com o setor de segurança pública. Rev Saúde Publica 2005; 39(4): 627-33.. Traffic accidents with victims are routinely registered by local police authorities in police reports (BO, in Portuguese). This legal tool brings information on type of accident, number of people involved, and presence of injured or fatal victims. Cases that demand hospitalization are registered in medical reports. Fatal victims, regardless of the length of time between the accident and one's death, are registered in death certificates issued by the Institute of Legal Medicine (IML), a mandatory procedure in cases of death by external causes.

Records managed by different institutions must be fully considered to better understand this contemporary phenomenon which strongly impacts Public Health.

This study, as a part of Avaliação do Projeto Vida no Trânsito⁷7. Morais Neto OL, Silva MMA, Lima CM, Malta DC, Silva Jr JB, et al. Vida no Trânsito Project: evaluation of the implementation in five Brazilian state capitals, 2011-2012. Epidemiol Serv Saúde 2013; 22(3): 373-82., seeks to evaluate the completeness of different information systems in order to improve characterization of road accidents and victims in Belo Horizonte, Minas Gerais, Brazil.

METHODS

Exploratory study performed with three databases: Information System of BHTRANS, Belo Horizonte's Transportations and Traffic Company (BH10), Hospital Information System (SIH) and Mortality Information System (SIM), from 2008 to 2010.

BH10 includes all traffic accidents that happen within the limits of the municipality based on police reports issued at the moment and place of accident, covering information about all people involved and the accident circumstances.

All people included in the Police reports were considered involved in cases (drivers, victims, witnesses, car owners). If applicable, information about age, gender, presence of injury, deaths, condition of passenger or driver, use of safety equipment (seat belt, helmet, child seat), license, and alcohol intake were made available when it came to drivers. Data about type and circumstances of the accident included: type of accident (collision, run over, fall); vehicles involved (car, bicycle, bus, motorcycle, truck); severity (fatal or non fatal); month, day of the week or weekend (Saturday and Sunday) of occurrence, time and place; highway conditions (traffic signs, paving - asphalt or others); weather conditions (good, rainy, cloudy).

SIH holds information about Hospital Admissions (AIH) for payment of services delivered by the National Public Health System (SUS). Records containing primary or secondary diagnosis of traffic accident according to codes V00 to V89, based on the 10^th International Disease Classification (IDC-10) were selected. Records whose discharge was marked as "permanence" were excluded from the sample, once they corresponded to an administrative act performed while the patient was still hospitalized, but submitted to new treatment procedures. However, these records were also used in analysis of costs, type of procedures used and need of intensive care assistance. The following variables related to patients admitted to hospitals of the SUS network (hired or associated) were: age group, gender, educational level, occupation, city of living and of hospital admission (Belo Horizonte, others). In order to obtain data about the highway user (pedestrian, cyclist, motorcycle riders, car drivers, bus drivers, truck drivers, unspecified vehicle), IDC-10 codes at secondary diagnosis were used.

Data about hospitalization included reasons for admission (primary diagnosis), date of admission and discharge, reason for discharge (cure, improvement, death, referral, other), type of procedures (clinical or surgical), and cost of hospitalization, assistance at intensive care unit, and legal nature of the institution (public, hired, associated).

Information about deaths was obtained from death certificates and reports issued by the IML and registered at SIM. Data was selected in databases with the codes V00 to V89 of IDC-10.

Variables analyzed were age group, gender, educational level, marital status (single, married, widow(er), divorced or legally separated), ethnicity (white, black, brown) and place of death (hospital, health institution, scene of accident). Type of highway user was selected as basic cause. Variables related to city of living and city of accident were grouped as Belo Horizonte and others.

Post-mortem relation to alcohol, marijuana, cocaine and crack was assessed based on investigative reports issued by IML, obtained in active search and included in SIM by the epidemiological surveillance Committee. Results of toxicology tests were classified according to IDC-10. Positive blood alcohol level was identified by Y90.0 to Y90.9; blood marijuana metabolites level, by F12.0 to F12.9, and cocaine/crack, by F14.0 to F14.9.

Analyses included absolute and relative frequencies, estimates of accident rates and fatal victims per 10,000 vehicles ration, considering Belo Horizonte's fleet as denominator of due years⁸8. Empresa de Transporte e Trânsito de Belo Horizonte (BHTRANS). Frota de veículos em Belo Horizonte - SisMob-BH. Disponível em: http://www.bhtrans.pbh.gov.br/portal/page/portal/portalpublicodl/Temas/Observatorio/SISMOBBH-2013/Tabelas14_FrotaSisMob-BH-Tab14b190413.pdf. (Acessado em 5 de junho de 2013).
http://www.bhtrans.pbh.gov.br/portal/pag... , and hospitalization rate per traffic accident among Belo Horizonte inhabitants⁹9. Instituto Brasileiro de Geografia e Estatística. Censo 2010 - Dado Belo Horizonte. Disponível em: http://www.censo2010.ibge.gov.br/painel/?nivel=mn. (Acessado em 15 de janeiro de 2013).
http://www.censo2010.ibge.gov.br/painel/... , by city of death occurrence.

Univariate analysis and risk estimates with 95% confidence interval (95%CI) for fatal accidents on weekends (Saturdays and Sundays); deaths on the scene of accident - driver or passenger/pedestrian - and injury or death on the scene of accident as regarding gender of drivers. Data were organized and statistical analyses were made in EpiInfo 3.5.1.

The study was approved by the Ethics Committee of UFMG and by the State Health Department of Belo Horizonte, protocols 158.014/2012 and 182.177/2012.

RESULTS

Each database provided information that was essential to the understanding of traffic accidents with victims (Table 1).

The traffic information system allowed a characterization of accidents and provided data about people involved in accidents: identification of injured or fatal victims, gender and age, as well as the drivers' characteristics. These form blanks were fulfilled in 93% of cases. Although this was the only source of variables relevant for the identification of factors associated with accidents (use of safety equipment, alcohol intake by the driver, highway conditions, local traffic signs, type of driver license), these blanks were found incomplete more often, which limited the analysis. There were no data available about socioeconomic status and place of residence of people involved, nor information about severity of injuries, being the record limited to registration of death occurrence on the scene of accident of within the first hours after the happening (Table 1).

Databases on hospitalizations and mortality covered more complete data about victims characteristics, as socioeconomic variables (educational level, ethnicity/skin color) were only fully available at SIM and presented high completeness rates (above 96%). Fulfilling of information about educational level and occupation at SIH was satisfactory (below 10%). Hospitalization costs, procedures, and assistance at intensive care unit were able to estimate the impact on health services in SIH. Use of less specific codes (IDC-10) in secondary diagnosis at SIH and primary cause at SIM limited the analysis of accident characteristics. The absence of this information and non correlation to hospitalization or death by traffic accidents underestimates the number of victims and makes accidents characterization more difficult (Table 1).

At SIM, records about occupation and work-related accidents were inadequately fulfilled. The place of accident was not informed in 80% of the records between 2008 and 2009, and in 55% of records in 2010. SIM also allowed identifying deaths on public highways and out of the city. Between 2008 and 2010, 48,918 traffic accidents with victims happened within the limits of Belo Horizonte, including roads crossing the city, thus accounting 109,322 people involved in accidents and 61,950 victims, according to BH10. All 10,374 hospital admissions (SIH) and 1,869 deaths (SIM) were related to accidents inside and out the city. Deaths occurred on the scene of accident or at health institutions.

ACCIDENTS WITH VICTIMS ACCORDING TO POLICE REPORTS - BH 10

There was an increase of 6% in number of victims, especially males (72%), Young people aging 18 to 29 (45%), or adults aging 30 to 39 (22%). Table 2 shows a reduction in the ratio fatal victims and obits among pedestrians (41%).

Cars (46%) and motorcycles (34%) were the vehicles more often involved in accidents with pedestrians, to which 18% of the elderly and 15% of teenagers younger than 15 years fall victim. Males were more often among injured drivers (OR = 1.66; 95%CI 1.59 - 1.73) or those dead on the scene of accident (OR = 3.25; 95%CI 1.87 - 5.67). Pedestrians or passengers were almost twice more likely to die on the scene of accident (OR = 1.81; 95%CI 1.58 - 2.08) when compared to drivers. Increase in traffic accidents (7%) was below the increase of vehicle fleet (20%), with a reduction in the accidents/fleet ratio from 142 to 126 per 10,000 vehicles (Table 2). Considering the fleet, the proportion of 6.8% of motorcycles involved in accidents was quite superior to that related to cars (1.6%).

Severity of motorcycle accidents was evidenced by the proportion of fatal victims, which is nearly three times higher than fatal victims by car accidents (Table 2). Car collisions (70%) and accidents with pedestrians (19%) were the most common types of accident. While accidents were more frequent on Fridays (16%) and between 12p.m. and 9p.m. (54%), fatal cases were 50% more likely to happen on weekends (95%CI 1.20 - 2.05), mostly (42%) between 9p.m. and 6a.m. Good weather conditions (91%) and paved roads (97%) were registered in the vast majority of Police reports.

TRAFFIC ACCIDENTS VICTIMS ACCORDING TO HOSPITAL ADMISSIONS - SIH/SUS

Three public hospitals held more than 90% of admissions due to TA. The progressive increase in hospitalization due to TA in the years studied and predominance of males (81%), young people aging 18 to 29 (39%) or adults aging 30 to 39 years (19%) among victims validate information obtained from BH10. Among residents of the city capital (47%), hospitalization rates increased from 6.2 to 8.3 per 10,000 inhabitants. Table 3 shows predominance of motorcycle riders among victims of traffic accidents. In this period, hospital admissions due to car accidents increased by three times, as well as those involving motorcycles (43%) and pedestrians (13%), with reduction in cases involving bicycles.

Head and neck trauma (31%), lower limb trauma (28%), and polytrauma (6%) were the leading causes for hospital admissions, as 23% required intensive care. Surgical procedures were performed in more than 70% of cases. In this period, an increase of 12% was seen in the proportion of hospital discharge due to healing or improvement, and reduction of 20% in deaths. In 2010, 3,191 (79%) of hospital admissions had favorable outcomes and 283 (7%) patients died. Referrals and other reasons accounted for 543 (14%) of all hospital admissions (Table 3). Costs increased by 53% between 2008 and 2010, going from R$ 9.3 to 14.2 million due to admissions in intensive care units, accounting for more than half of all costs in 2010.

TRAFFIC ACCIDENTS FATAL VICTIMS ACCORDING TO SIM DATA

1,869 deaths by TA occurred in the period, accounting for 26% of deaths by external causes. Most fatal victims (53%) were residents of Belo Horizonte, males (79%), young people aging 18 to 29 (30%) or adults aging 30 to 39 (18%), single (60%), brown or black-skinned (62%), with less than eight years of study (56%). The level of the aged among deaths due to TA (21%) was higher than that regarding hospital admissions (7%) or police reports (6%). About 30% of deaths of Belo Horizonte residents happened in other cities. There was a 40% increase in mortality rates among residents, from 13.9 to 19.4 deaths per 100,000 inhabitants, when deaths in other cities were included.

Among annual deaths by TA, more than two thirds (479 on average) of cases happened in Health Institutions, almost exclusively in public hospitals (96%). There was a 12% increase in deaths on public highways, on the scene of accident (129 in 2008 to 145 in 2010). A higher rate of death of pedestrians was also found, followed by the death of motorcycle riders and car drivers (Table 3). In 2009, an increase in deaths by "unspecific" TA was reported, as related to 2008 and 2010, reaching 26% of total number of deaths. However, the distribution of road users was similar in all years. In 2010, automobiles were mostly involved in accidents with pedestrians (28%), followed by heavy vehicles such as bus and trucks (22%), and motorcycles (21%).

The leading injuries reported in death certificates were polytrauma (53%), head and neck trauma (35%), and chest and abdominal trauma (6%).

In the course of this three-year study, toxicological test was positive for alcohol, marijuana and/or cocaine/crack in 348 (18.6%) cases of death. Among 432 fatal victims on the scene of accident, 55% were positive for toxicology, while among obits in hospitals (1,437) only 11% of tests were positive. In 275 (79%) of positive tests, alcohol was detected in isolation. Association of alcohol and illicit drugs (marijuana, cocaine/crack) was found in 38 death cases (11%), and illicit drugs only were detected in 35 cases (10%).

DISCUSSION

This study allowed a broader understanding of traffic accidents in Belo Horizonte, for it brought information from different sources: police officers at the scene of accident, health professionals at health institutions, and forensic doctors through death certificates. Our findings show the liability of different sources with different purposes and point of view to collect data, and justify the increasing use of such databases. Linking information from different, more complete and trustworthy sources¹⁰10. Mello Jorge MHP, Koizumi MS. Acidentes de transito no Brasil: breve análise de suas fontes de dados. Rev ABRAMET 2001; 38: 49-57.^,1111. Tomimatsu MFAI, Andrade SM, Soares DA, Mathias TAF, Sapata MPM, Soares DFPP, et al. Qualidade da informação sobre causa externas no Sistema de Informações Hospitalares. Rev Saúde Pública 2009; 43(3): 413-20. can be a reliable strategy for the assessment and characterization of Information Systems, and also for the improvement of actions to reduce TA.

All three databases lacked some information, which confirms the findings of previous studies on Information Systems¹²12. Lima CRA, Schramm JMA, Coeli CM, Silva MEM. Revisão das dimensões de qualidade dos dados e métodos aplicados na avaliação dos sistemas de informação em saúde. Cad Saúde Pública 2009; 25(10): 2095-109., reflecting inadequate fulfilling of data regarding important risk factors for severe and fatal accidents: alcohol intake by the driver and misuse of safety equipment¹1. World Health Organization. Global status report on road safety: time for action. Geneva: World Health Organization; 2009.. The adequate report of this type of information should be continuously encouraged¹²12. Lima CRA, Schramm JMA, Coeli CM, Silva MEM. Revisão das dimensões de qualidade dos dados e métodos aplicados na avaliação dos sistemas de informação em saúde. Cad Saúde Pública 2009; 25(10): 2095-109.. The limited availability of information about the place of accident (such as road conditions and traffic signs) suggests the use of tools such as Google Street View to help analyze these features¹³13. Clarke P, Ailshire J, Melendez R, Bader M, Morenoff J. Using Google Earth to conduct a neighborhood audit: Reliability of a virtual audit instrument. Health & Place 2010; 16(6): 1224-9.. Although SIH only holds information about hospital admissions at SUS, it covers well the most severe accidents, a gateway to the public health system. Deaths in hospitals may reflect the severity of accidents or poor quality of health care, as well as hospital discharges. Universal coverage of SIM shows total number of deaths, including those happening on the streets, hospitals or out of the city. Fluctuation in death rates due to TA classified as unspecific in the studied period, which increased in 2009, represents a challenge in continued qualification of secondary databases and points the importance of completing data with information from other sources, including search in newspapers¹⁴14. Villela LCM, Rezende EM, Drumond EF, Ishitani LH, Carvalho GML. Utilização da imprensa escrita na qualificação das causas externas de morte. Rev Saúde Pública 2012; 46(4): 730-6. and consultation to legal medical reports and certificates. The distribution of deaths per institutions as found at SIM, allows one to know services with greater demand of care in cases of death.

The reduction in TA in proportion to the city fleet was also seen in Campinas by Marín-León et al.¹⁵15. Marín-León L, Belon AP, Barros MBA, Almeida SDM, Restitutti MC. Tendência dos acidentes de trânsito em Campinas, São Paulo, Brasil: importância crescente dos motociclistas. Cad Saúde Pública 2012; 28(1): 39-51., who attributed this finding to the slowness in traffic due to the large number of cars on streets. Isolated analysis of this indicator should be avoided, once it does not point an effective reduction in the risk of injuries and death. Although the increase in number of accidents were not exactly related to the urban fleet increase, the magnitude of the problem was reflected in the annual increase of accidents with fatal and injured victims, which places TA as one of the most significant public health problems not only in Belo Horizonte, but in all cities of Brazil⁴4. Morais Neto OL, Montenegro MMS, Monteiro RA, Siqueira Junior JB, Silva MMA, Lima CM, et al. Mortalidade por acidentes de transporte terrestre no Brasil na última década: tendência e aglomerados de risco. Ciênc Saúde Coletiva 2012; 17(9): 2223-36.. Besides that, mortality rates due to TA are high, indicating the need for effective interventions to substantially reduce these indexes³3. Bacchieri G, Barros AJD. Acidentes de trânsito no Brasil de 1998 a 2010: muitas mudanças e poucos resultados. Rev Saúde Pública 2011; 45(5): 949-63.^,44. Morais Neto OL, Montenegro MMS, Monteiro RA, Siqueira Junior JB, Silva MMA, Lima CM, et al. Mortalidade por acidentes de transporte terrestre no Brasil na última década: tendência e aglomerados de risco. Ciênc Saúde Coletiva 2012; 17(9): 2223-36.. Despite improvements in legislation, the implementation of traffic control agencies, improvement in automobiles' safety, and use of electronic monitoring in traffic, no reduction in accidents, deaths and disabilities resulting from them were seen in our study³3. Bacchieri G, Barros AJD. Acidentes de trânsito no Brasil de 1998 a 2010: muitas mudanças e poucos resultados. Rev Saúde Pública 2011; 45(5): 949-63..

The profile of victims identified in these Information Systems was similar to that reported in literature³3. Bacchieri G, Barros AJD. Acidentes de trânsito no Brasil de 1998 a 2010: muitas mudanças e poucos resultados. Rev Saúde Pública 2011; 45(5): 949-63.

4. Morais Neto OL, Montenegro MMS, Monteiro RA, Siqueira Junior JB, Silva MMA, Lima CM, et al. Mortalidade por acidentes de transporte terrestre no Brasil na última década: tendência e aglomerados de risco. Ciênc Saúde Coletiva 2012; 17(9): 2223-36.^-55. Cabral APS, Souza WV, Lima MLC. Serviço de Atendimento Móvel de Urgência: um observatório dos acidentes de transportes terrestre em nível local. Rev Bras Epidemiol 2011; 14(1): 3-14.. The high rate of fatal victims compared to hospital and traffic reports suggests high severity of lesions in this population as a result of fragility and presence of comorbidities common to this group of people¹⁶16. Silveira R, Rodrigues RAP, Costa Junior ML. Idosos que foram vítimas de acidentes de trânsito no município de Ribeirão Preto, em 1998. Rev Lat Am Enferm 2002; 10(6): 765-71.. Moreover, most of the victims were pedestrians, that is, more vulnerable to direct impact by cars¹⁶16. Silveira R, Rodrigues RAP, Costa Junior ML. Idosos que foram vítimas de acidentes de trânsito no município de Ribeirão Preto, em 1998. Rev Lat Am Enferm 2002; 10(6): 765-71..

The finding of predominantly black/brown-skinned or less educated fatal victims points to inequity in TA deaths, rooted in social determinants¹⁷17. Montenegro MMS, Duarte EC, Prado RR, Nascimento AF. Mortalidade de motociclistas em acidentes de transporte no Distrito Federal, 1996-2007. Rev Saúde Pública 2011; 45(3): 529-38.^,1818. Caiaffa WT, Friche AAL. Urbanização, globalização e segurança viária: um diálogo possível em busca da equidade? Ciênc Saúde Coletiva 2012; 17(9): 2238-41.. This information, made available only by SIM, is essential for the definition of public policies and strategies aimed at the most vulnerable groups.

The greater involvement of young male motorcycle riders or car drivers, on weekends³3. Bacchieri G, Barros AJD. Acidentes de trânsito no Brasil de 1998 a 2010: muitas mudanças e poucos resultados. Rev Saúde Pública 2011; 45(5): 949-63., suggests a relation between TA and the use of psychoactive substances, evidenced by blood alcohol and drugs level detected in more than half of death cases on the scene of accident. A study on fatal victims in the Federal District showed high rates of alcohol use (43%)¹⁹19. Modelli MES, Pratesi R, Tauil PL. Alcoolemia em vítimas fatais de acidentes de trânsito no Distrito Federal, Brasil. Rev Saúde Pública 2008; 42(2): 350-2.. A survey on alcohol intake in Brazil pointed a 35% prevalence of drinking-driving relation, 6 times higher chances of males drive after using alcohol, and 7 times higher chances of previous involvement in alcohol-related accidents²⁰20. Pechansky E, De Boni R, Diemen LV, Bumaguin D, Pinsky I, Zaleski M, et al. Highly reported prevalence of drinking and driving in Brazil: data from first representative household study. Rev Bras Psiquiatr 2009;31(2): 125-30.. More rigidly in road monitoring with the use of breath alcohol analyzers could reduce these numbers¹⁹19. Modelli MES, Pratesi R, Tauil PL. Alcoolemia em vítimas fatais de acidentes de trânsito no Distrito Federal, Brasil. Rev Saúde Pública 2008; 42(2): 350-2.^,2121. Mascarenhas MDM, Malta DC, Silva MMA, Gazal-Carvalho C, Monteiro RA, Morais Neto OL. Consumo de álcool entre vítimas de acidentes e violências atendidas em serviços de emergência no Brasil, 2006 e 2007. Ciênc Saúde Coletiva 2009; 14(5): 1789-96.. Previous use of illicit drugs in 21% of fatal victims point to the need of surveillance aimed at these substances, as there are laws on the subject already²²22. Brasil. Lei 11.705 de 19 de junho de 2008. Altera a Lei no 9.503, de 23 de setembro de 1997, que institui o Código de Trânsito Brasileiro, e a Lei no 9.294, de 15 de julho de 1996. Diário Oficial da União 2008; Jun 20..

The range of motorcycles in cities and the vulnerability of riders, ranking first in hospital admissions and second in death cases, have made them target in prevention actions²³23. Veronese AM, Oliveira DLLC. Os riscos dos acidentes de trânsito na perspectiva dos moto-boys: subsídios para a promoção da saúde. Cad Saúde Pública 2006; 22(12): 2717-21.. Motorcycles are used as work tools, and some riders are subjected to long workshifts and low wages by clients and employers²⁴24. Silva DW, Andrade SM, Soares DA, Soares DFPP, Mathias TAF. Perfil do trabalho e acidentes de trânsito entre motociclistas de entregas em dois municípios de médio porte do Estado do Paraná, Brasil. Cad Saúde Pública 2008; 24(11): 2643-52..

The study showed that motorcycle riders have been responsible for an expressive portion of accidents with pedestrians³3. Bacchieri G, Barros AJD. Acidentes de trânsito no Brasil de 1998 a 2010: muitas mudanças e poucos resultados. Rev Saúde Pública 2011; 45(5): 949-63.^,2525. Barros AJ, Amaral RL, Oliveira MS, Lima SC, Goncalves EV. Acidentes de trânsito com vítimas: sub-registro, caracterização e letalidade. Cad Saúde Pública 2003; 19: 979-86., which happen especially with people under the age of 15 and with the elderly. The authors gave emphasis to the double risk of accidents with pedestrians by motorcycle riders compared to car drivers²⁵25. Barros AJ, Amaral RL, Oliveira MS, Lima SC, Goncalves EV. Acidentes de trânsito com vítimas: sub-registro, caracterização e letalidade. Cad Saúde Pública 2003; 19: 979-86.. Freitas et al. analyzed traffic accidents involving children and reported that 76% of them resulted from accidents with pedestrians or with bicycle riders²⁶26. Freitas JPP, Ribeiro LA, Jorge MT. Vítimas de acidentes de trânsito na faixa etária pediátrica atendidas em um hospital universitário: aspectos epidemiológicos e clínicos. Cad Saúde Pública 2007; 23(12): 3055-60.. A study about victims of trauma suggested that accidents with pedestrians are the second leading cause of trauma among the elderly, being fall the first¹⁶16. Silveira R, Rodrigues RAP, Costa Junior ML. Idosos que foram vítimas de acidentes de trânsito no município de Ribeirão Preto, em 1998. Rev Lat Am Enferm 2002; 10(6): 765-71.. Vulnerability of pedestrians was shown in the analysis of hospital admissions and deaths³3. Bacchieri G, Barros AJD. Acidentes de trânsito no Brasil de 1998 a 2010: muitas mudanças e poucos resultados. Rev Saúde Pública 2011; 45(5): 949-63..

The large number of fatal victims in the scene of accident, including passengers or drivers, reinforces the idea that, despite the increase in cars' safety and the presence of speed radars, most severe accidents resulted from overspeeding, even on urban streets²⁷27. Thielen IP, Hartmann RC, Soares DP. Percepção de risco e excesso de velocidade. Cad Saúde Pública 2008; 24(1): 131-9.. The increase in hospitalization rates reflects the severity of traffic accidents, as well as the types of injury found in death certificates and hospital admission reports. Polytrauma was the main cause of death, mostly immediate, and head trauma were the leading cause for hospital admission. In literature reviews, head/neck was the second mostly affected region of the body in accidents, preceded by upper/lower limbs injuries²⁸28. Calil AM, Sallum EA, Domingues CA, Nogueira LS. Mapping injuries in traffic accident victims: a literature review. Rev Lat Am Enfermagem 2009; 17(1): 120-5.. The high rate of hospital discharge by improvement of patient's condition and the decrease in death rates among patients admitted to hospitals may suggest a better pre-hospital and hospital care. However, the large number of deaths on the scene of accident reveals the need for further specific studies on this theme.

The public health emergency and urgency network in Belo Horizonte has three reference trauma centers, where almost 100% of patients were admitted due to TA, which points the need of actions aimed at improving the assistance to these victims at these institutions⁶6. Gawryszewski VP, Kahn T, Mello Jorge MHP. Informações sobre homicídios e sua integração com o setor de segurança pública. Rev Saúde Publica 2005; 39(4): 627-33.. The correct codification of diagnosis at AIH must also be stimulated, for these data generate accurate information about hospital admissions by TA, a significant public health problem in Brazil.

High costs of hospitalization, especially for the public health network, were also found in other studies²⁹29. Mello-Jorge MHP, Koizumi MS. Gastos governamentais do SUS com internações por causas externas. Rev Bras Epidemiol 2004; 7(2): 228-38.^,3030. Melione LP, Mello-Jorge MHP. Gastos do Sistema Único de Saúde com internações por causas externas em São José dos Campos, São Paulo, Brasil. Cad Saúde Pública 2008; 24(8): 1814-24.. These costs arise from complex procedures such surgeries and intensive care assistance. Hospital admissions by external causes were shown to have higher costs than those by natural causes, and admissions by TA have a higher average cost²⁹29. Mello-Jorge MHP, Koizumi MS. Gastos governamentais do SUS com internações por causas externas. Rev Bras Epidemiol 2004; 7(2): 228-38.. Reduction in hospital admissions, severity of injuries and costs could be reached if prevention measures were taken, correct use of safety equipment included³3. Bacchieri G, Barros AJD. Acidentes de trânsito no Brasil de 1998 a 2010: muitas mudanças e poucos resultados. Rev Saúde Pública 2011; 45(5): 949-63..

In 2001, the Ministry of Health issued the National Policy for the Reduction of Morbidity and Mortality from Violence and Accidents³¹31. Brasil. Ministério da Saúde. Portaria n. 723, de 16 de maio de 2001. Dispõe sobre a Política Nacional de Redução de Morbimortalidade por Acidentes e Violência. Disponível em: http://portal.saude.gov.br/portal/arquivos/pdf/politicapromocao.pdf
http://portal.saude.gov.br/portal/arquiv... , implemented in Belo Horizonte, based on the magnitude, risk and importance of TA in overall mortality rates⁴4. Morais Neto OL, Montenegro MMS, Monteiro RA, Siqueira Junior JB, Silva MMA, Lima CM, et al. Mortalidade por acidentes de transporte terrestre no Brasil na última década: tendência e aglomerados de risco. Ciênc Saúde Coletiva 2012; 17(9): 2223-36.^,3232. Souza ER, Minayo MCS, Franco LG. Avaliação do processo de implantação e implementação do Programa de Redução da Morbimortalidade por acidentes de Trânsito. Epidemiol Serv Saúde 2007; 16(1): 19-31., which was confirmed by our findings. In 2010, these proposals were broadened and strengthened with the creation of Project Life in Traffic⁷7. Morais Neto OL, Silva MMA, Lima CM, Malta DC, Silva Jr JB, et al. Vida no Trânsito Project: evaluation of the implementation in five Brazilian state capitals, 2011-2012. Epidemiol Serv Saúde 2013; 22(3): 373-82..

Although there was no reduction in mortality related to TA, it is important to note the mobilization of the Municipal Health Department and intersectorial articulations that allowed integration of data from Police reports and Health reports. This initiative allowed implementing inspection and educational actions regarding traffic, with special emphasis to the most vulnerable groups.

The use of secondary databases with inadequate fulfilling of information limited the analysis of important features such as alcohol and safety equipment use by car drivers. Isolated assessment of different databases cannot assure analysis of the same individuals. Not all injured victims reported in Police reports were admitted to hospitals, nor did hospitalizations and deaths within the limits of the city exclusively correspond to TA that took place in the city, that is, the ones that can be managed by local traffic agencies.

Linking data from different sources allows completing information available at each database, and also relating victims' reports to the type of accident and outcomes, expanding understanding of this relevant public health issue.

CONCLUSION

Our study was able to show the completeness of all three sources of information used, with their respective purposes, and allowed pointing out important aspects of the events related to traffic accidents, from the moment they happen to possible death occurrences.

Differences in the estimates of number of injured and fatal victims found in databases represent a difficulty in measuring the actual magnitude of traffic accidents. Despite incompleteness of data and the use of unspecific codes in records, which hinders the exact listing of all victims, our study showed the vulnerability of young males, especially motorcycle riders, and pedestrians, namely the elderly. Our findings could also show the severity of accidents, reflected by the large number of fatal victims on the scene of accident and the type of lesions reported, raising the hypothesis of overspeeding as associated factor, as well as alcohol and/or drug abuse, related to more than half of deaths on the scene of accident.

The increase in mortality coefficient related to TA among Belo Horizonte inhabitants, added by deaths occurred out of the city limits, mostly on roads administered by State or Federal entities, underlies the need for broader policies and the importance of intersectoral and interinstitutional actions.

The identification of vulnerable groups and risk factors related to traffic accidents are essential to the development of public policies based on evidence, supporting planning and more effective interventions aimed at the reduction of TA morbimortality.

Publication Dates

History