TY - JOUR
AU1 - Nazif-Muñoz, José Ignacio
AU2 - Nandi, Arijit
AU3 - Ruiz-Casares, Mónica
AB - Abstract Background In 2010, Brazil introduced child restraint legislation (CRL). We assessed the effectiveness of CRL in reducing child (aged 0–8 years) injuries and fatalities by race. We performed an evaluation study with an interrupted time–series design. Methods We measured the effect of CRL on two outcomes—number of child deaths and number of child injured in traffic collisions per child population, stratified by race, from 2008 to 2014. We controlled for time, unemployment rate and oil consumption (barrels/day in thousands). Results The CRL was associated with a 3% reduction in the rate of child injuries among whites (incidence rate ratio (IRR): 0.97; 95% CI: 0.96–0.99), but no reduction in child injuries among non-whites (IRR: 0.99; 95% CI: 0.99–1.00). In the first month after the implementation of Brazil’s CRL we observed a 39% reduction in all child fatalities (IRR: 0.61; 95% CI: 0.44–0.84), including a 52% reduction among whites (IRR: 0.48; 95% CI: 0.33–0.68), but no reduction in non-white fatalities (IRR: 0.87; 95% CI: 0.55–1.37). Conclusions Our results support the hypothesis that socially advantaged populations were more likely to consistently adopt and employ restraint devices following the reform. Countries should also consider complementary policies that facilitate an equitable distribution of safety devices that reach vulnerable populations. accidents, children, race Background In the last 2 decades, deaths as a result of road traffic injuries (RTI) have increased by 46%, leading to a yearly worldwide loss of ~80 million healthy years of life.1 Furthermore, the burden of RTI has increased substantially in countries that have experienced rapid economic growth (e.g. the BRICS group—Brazil, Russia, India, China and South Africa). These countries have both prioritized road infrastructure investment and promoted industrialization with increased vehicle production and purchases, particularly cars and motorcycles2; however, they have left behind public transportation projects3–6 as well as road safety policies.7 Brazil is an emerging economic powerhouse and the largest middle-income country in Latin America. Yet, Brazil is among countries with the highest socioeconomic and racial inequalities in the world.8,9 There are signs that population health in Brazil is improving on account of reductions in infectious disease morbidity and mortality,10 and improvements in child health.11–13 However, many challenges remain, including a high burden of mortality from injury.14 In fact, RTI are overlooked compared with problems such as circulatory diseases15 and are emerging as important causes of mortality in Brazil.14,16 In Brazil, two trends in traffic mortality and injury rates have been observed for the period 1998–2012: between 1998 and 2003 there was a reduction for all types of road users, followed by an upward trend between 2003 and 2012, when a reduction in pedestrian deaths but increased fatalities among vehicle occupants, especially among motorcyclists, was observed.17 Regarding non-fatal RTIs, a 55% increase in emergency department visits was observed between 2003 and 2013.18 According to Morais-Neto et al.19 and Liberatti et al.20 the reduction of the first period was mostly due to the introduction of the new Brazilian Traffic Code in 1998. The likely contributors to the subsequent upward trend in mortality rates were increased household income and the rapid increase in motorization rates of cars and motorcycles.19,21 To target the upward trend different legislative initiatives were introduced including the enactment of child restraint legislation (CRL) in 2010,16 in which child restraint use was made mandatory.22 More particularly, this reform determined that drivers of motor vehicles are responsible for installing age-appropriate child restraint systems (seats for infants under the age of 7 years and 6 months of age), and ensuring proper use of such devices. Drivers who fail to comply with this legislation can be fined $191.47 Brazilian Reals (~60 US dollars) and have their vehicle impounded, which is nontrivial considering the monthly minimum wage in Brazil for 2016 was $ 880.00, or ~228 US dollars.23 A series of measures to promote child safety in Brazil were introduced after the enactment of the Brazilian CRL, among these the launching of the campaign ‘A criança no trânsito’ (‘Children in traffic’) to raise awareness of children and parents regarding road safety practices.24 One evaluation of Brazil’s CRL was carried out one year after its enactment. In this study, no evidence was found to suggest that this policy was effective in reducing children’s fatality rates.25 This study, however, had three limitations. First, it did not consider child injuries. It is well established that in order to assess a road safety measure more comprehensively both injuries and fatalities should be considered.26 Further, in the realm of CRLs there is evidence that, due to the low number of children’s fatalities, these evaluations may be underpowered to distinguish the impact of CRLs on mortality from sampling variability.27,28 Secondly, it only considered fatalities as crude numbers and not rates, without normalizing them by including the number of children or motor vehicles in the denominator. Lastly, the post-enactment period consisted of only one year. This limited the potential measured impact of the policy since it only allowed for an abrupt change; however, the effects of CRLs may be delayed28 since the purchase of child restraints and learning to install these devices properly may not be immediate. In this study, we evaluated the impact of Brazil’s CRL on rates of child injuries and fatalities per child population and the number of motor vehicles, allowing for a post-enactment period of more than four years, which allows us to more accurately capture its effect. More specifically, our objective was to determine whether Brazil’s CRL was effective in reducing child (aged 0–8) occupant injuries and fatalities. Because Brazil has a history of racial inequality that is manifested in health outcomes,29 particularly in children, we also compared effects across racial groups. Evaluating whether the impact of CRL varied for white and non-white children provides an opportunity to identify differences and/or similarities among these groups and might inform targeted interventions for reducing racial inequality in the realm of road safety. Methods Study design We evaluated the impact of CRL in Brazil using an interrupted time–series design, which is a time–series analysis where the series is divided, or interrupted, by the intervention into pre-intervention and post-intervention periods, which are compared.30,31 Two study populations were considered: (i) child occupants who were injured in vehicle collisions in Brazil between 2008 and 2014 and (ii) child occupants who died in vehicle collisions for the same period. From these two populations we also considered two subsets: white and non-white children. Data The main explanatory variable was the intervention: the implementation of Brazil’s CRL in 2010. We created a dummy variable to compare the post-intervention period (September 2010 to December 2014) with the pre-intervention period (January 2008 to August 2010). Outcomes and other covariates were retrieved from: the ‘DATASUS Database,’ the Brazilian Information System on Hospital Admissions;32 the 2010 Brazilian census obtained from the Brazilian Institute of Geography and Statistics;33 the ‘Vehicle fleet Database’ from the National Department of Traffic;34 and the ‘Economic Indicators database’ from the Central Bank of Brazil.35 The Information System on Hospital Admissions is the most comprehensive source of data on traffic-related injuries and deaths in Brazil, since it covers deaths at any time after the event and not only those at the time of the vehicle-crash.16 Nationwide Brazilian censuses, last conducted in 2010, are done once in 10 years. The ‘Vehicle fleet database’ has monthly information since 1998 considering the type of vehicles and geographical location by states and its Federal district. The economic indicators database from the Central Bank of Brazil contains monthly information regarding unemployment rates and oil consumption (barrels/day in thousands). The dependent variables were the number of children ages 8 years and younger who died or were injured as passengers in traffic collisions. We used International Classification of Diseases (10th revision) V30–V69 codes for occupant traffic-related deaths and injuries.36 We calculated the rates for each variable by month between January 2008 and December 2014, specifically child fatalities and injuries per 1 000 000 motor vehicles and 1 000 000 child population. We accounted for several potential confounders based on the road safety literature: unemployment rates,37 oil consumption38 and rates of child fatalities and injuries among 9–14 years old per 1 000 000 vehicles and per 1 000 000 child population.39 We included these variables in the analyses to account for economic and social changes occurring during the study period that might have influenced our primary outcomes. More specifically, these confounders are intended to proxy other secular changes associated with vehicle technology and infrastructure, which may influence child injuries and fatalities. Table A1 presents the sources of information and main descriptive variables used in this study. Statistical analysis The Poisson model equation estimating monthly injury and morality rates was expressed as follows: Log(E(Y))=β0+β1lagY+β2Time+β3CRL+β4(Time*CRL)+βk(X)+e (1) where Y denotes the outcome (monthly count of child deaths or injuries), β0 is the model intercept, β1 represents the outcome Y lagged by one month to avoid incorrect specification of intervention effects, β2 is the coefficient for the monthly time trend variable, β3 is the coefficient of the indicator variable for CRL implementation, β4 is the coefficient of the interaction between the indicator variable for CRL and the monthly time trend, βk includes a vector of potentially confounding covariates (i.e. oil consumption, unemployment and child fatalities and injuries age 9–14), and e is the model error term. In the pre-enforcement period, with CRL = 0, the model takes the form: Log(E(Y))=β0+β1lagY+β2Time+βk(X)+e (2) In the post-enforcement period, with CRL = 1, the model takes the form: Log(E(Y))=(β0+β1lagY+β3)+(β2+β4)Time+βk(X)+e (3) where β3 is the change in the log rate ratio measuring the immediate effect of the CRL and (β2 + β4) is the post-implementation rate of change in injury and mortality, with β4 representing the change in slope after the enactment of CRL. We reported estimates on the incidence rate ratio (IRR) scale, with the immediate effect calculated as [exp (β3)] and the monthly trend as [exp (β4)]. All analyses were conducted using STATA version 14. Models used the ‘exposure’ function for the population of children age 0–8 per year when assessing rates per population and the ‘exposure’ function for the vehicle fleet per year when assessing rates per vehicle. We also used the Prais–Winsten estimation procedure in order to account for serial correlation type AR(1).40 We used this method to test the robustness of our Poisson models. Results Overall trends Figure 1 shows crude trends for injury rates for all children, white children and non-white children, per vehicle fleet and per population. In the pre-legislation period (January 2008 to August 2010), injury rates for all children per 1 000 000 vehicle fleet were highly variable and their averages slightly higher than in the post-legislation period (September 2010 to December 2014). For all children, the rate of child injuries per 1 000 000 vehicle fleet was 0.59 [95% CI: 0.55–0.62] in the pre-legislation period and 0.61 [95% CI: 0.55–0.68] in the post-legislation period. For whites, the injury rates per vehicle fleet were 0.44 in the pre-legislation period [95% CI: 0.39–0.50] and 0.36 [95% CI: 0.32–0.39] in the post-legislation period. The rates for non-white children per vehicle fleet were 0.17 [95% CI: 0.14–0.20] in the pre-legislation period and 0.22 [95% CI: 0.21–0.24] in the post-legislation period. The rate is slightly higher for the white children population since they travel more than non-white children in motor-vehicles. Fig. 1 View largeDownload slide Child injury rates per 1 000 000 motor vehicles and 1 000 000 child population, Brazil, 2008–14. Observed Monthly Child Injury Rates Child in Brazil, 2008–14. The vertical line represents the month in which Brazil started its CRL implementation. Fig. 1 View largeDownload slide Child injury rates per 1 000 000 motor vehicles and 1 000 000 child population, Brazil, 2008–14. Observed Monthly Child Injury Rates Child in Brazil, 2008–14. The vertical line represents the month in which Brazil started its CRL implementation. For the injury rates per 1 000 000 child population, the injury rates among all children were 2.81 [95% CI: 2.44–3.17] and 3.69 [95% CI: 3.48–3.89] in the pre- versus post-legislation periods. For white children the rates were 4.06 [95% CI: 3.50–4.62] and 4.48 [95% CI: 4.14–4.82], respectively, compared to rates of 1.57 [95% CI: 1.30–1.83] and 2.90 [95% CI: 2.67–3.12] for non-white children. Figure 2 shows trends for child fatalities. For all children, the fatality rates per 1 000 000 vehicles in the pre-legislation and post-legislation periods were 0.44 [95% CI: 0.38–0.50] and 0.34 [95% CI: 0.31–0.36], respectively. The corresponding rates were 0.28 [95% CI: 0.23–0.33] and 0.19 [0.17–0.21] for white children and 0.16 [95% CI: 0.13–0.18] and 0.14 [95% CI: 0.12–0.15] for non-white children. In reference to fatalities per 1 000 000 child population, we observed the following outcomes: for all children, rates of 1.99 [95% CI: 1.73–2.25] and 2.15 [95% CI: 1.95–2.35]; for white children rates of 2.53 [95% CI: 2.12–2.95] and 2.52 [95% CI: 2.25–2.79]; and lastly, for non-white children, rates of 1.45 [95% CI: 1.22–1.68] and 1.79 [95% CI: 1.58–1.99]. Figure 2 View largeDownload slide Child fatality rates per 1 000 000 motor vehicles and 1 000 000 child population, Brazil, 2008–14. Observed Monthly Child Fatality Rates Child in Brazil, 2008–14. The vertical line represents the month in which Brazil started its CRL implementation. Figure 2 View largeDownload slide Child fatality rates per 1 000 000 motor vehicles and 1 000 000 child population, Brazil, 2008–14. Observed Monthly Child Fatality Rates Child in Brazil, 2008–14. The vertical line represents the month in which Brazil started its CRL implementation. In sum, for injuries, we observed lower rates per vehicle fleet in the post-CRL period for all children and white children, higher rates per vehicle fleet among non-white children, and higher rates per child population in the three groups in the post-legislation period. For fatalities, we observed decreases for all children and white children when the counts are normalized by vehicle fleet, and an increase for all children and non-white children when the rate is per child population. Although both figures show differences comparing the pre- and post-enactment periods, these changes cannot be attributed to the CRL unless other factors associated with the series (i.e. trends or seasonality) are accounted for. Effect of the policy reform on child occupant injuries Two main results are presented on Table 1. First, at the first month of the implementation of the CRL in Brazil, represented in columns ‘Immediate effects’ for the two statistical approaches (Poisson and Prais–Winsten AR(1)), we observed no immediate effect of the intervention on monthly injury rates for any of the three populations, regardless of the statistical model or the denominator chosen. Secondly, both models indicated a gradual reduction in monthly injuries over time, irrespective of the denominator, for all children and white children. In Poisson models, Brazil’s CRL was associated with a 2%, reduction in child injuries for the total and white populations, irrespective of the denominator used. However, the Prais–Winsten AR(1) results suggested a 6% reduction in injuries for the total population and a 11% reduction among whites when the denominator was the child population. Table 1 CRL effects on monthly injury rates per 1 000 000 child population and per 1 000 000 motor-vehicles, Brazil, 2008–2014, stratified by race Poisson Prais–Winsten AR(1) Immediate effects Gradual effects per month Immediate effects Gradual effects per month IRR (95% CI) IRR (95% CI) IRR (95% CI) IRR (95% CI) Total child injuries ages 0–8 per 1 000 000 motor-vehiclesa 0.82 (0.63–1.06) 0.98 (0.97–0.99) 0.89 (0.73–1.09) 0.98 (0.97–0.99) Total child injuries ages 0–8 per 1 000 000 children 0–8b 0.82 (0.63–1.07) 0.98 (0.97–0.99) 0.57 (0.19–1.74) 0.94 (0.90–0.99) White child injuries ages 0–8 per 1 000 000 motor-vehiclesc 0.90 (0.66–1.23) 0.97 (0.96–0.99) 0.99 (0.85–1.15) 0.98 (0.98–0.99) White child injuries ages 0–8 per 1 000 000 children 0–8d 0.91 (0.66–1.24) 0.97 (0.96–0.99) 0.71 (0.11–4.57) 0.89 (0.82–0.97) Non-white child injuries ages 0–8 per 1 000 000 motor-vehiclese 0.74 (0.53–1.03) 0.99 (0.97–1.00) 0.94 (0.87–1.00) 0.99 (0.99–1.00) Non-white child injuries ages 0–8 per 1 000 000 children 0–8f 0.74 (0.53–1.03) 0.99 (0.97–1.00) 0.49 (0.21–1.18) 1.00 (0.96–1.03) Poisson Prais–Winsten AR(1) Immediate effects Gradual effects per month Immediate effects Gradual effects per month IRR (95% CI) IRR (95% CI) IRR (95% CI) IRR (95% CI) Total child injuries ages 0–8 per 1 000 000 motor-vehiclesa 0.82 (0.63–1.06) 0.98 (0.97–0.99) 0.89 (0.73–1.09) 0.98 (0.97–0.99) Total child injuries ages 0–8 per 1 000 000 children 0–8b 0.82 (0.63–1.07) 0.98 (0.97–0.99) 0.57 (0.19–1.74) 0.94 (0.90–0.99) White child injuries ages 0–8 per 1 000 000 motor-vehiclesc 0.90 (0.66–1.23) 0.97 (0.96–0.99) 0.99 (0.85–1.15) 0.98 (0.98–0.99) White child injuries ages 0–8 per 1 000 000 children 0–8d 0.91 (0.66–1.24) 0.97 (0.96–0.99) 0.71 (0.11–4.57) 0.89 (0.82–0.97) Non-white child injuries ages 0–8 per 1 000 000 motor-vehiclese 0.74 (0.53–1.03) 0.99 (0.97–1.00) 0.94 (0.87–1.00) 0.99 (0.99–1.00) Non-white child injuries ages 0–8 per 1 000 000 children 0–8f 0.74 (0.53–1.03) 0.99 (0.97–1.00) 0.49 (0.21–1.18) 1.00 (0.96–1.03) aAdjusted for time trend, lag of child injuries, child injuries ages 9–14 per 1 000 000 vehicles, oil consumption and unemployment. bAdjusted for time trend, lag of child injuries, child injuries ages 9–14 per 1 000 000 child population, oil consumption and unemployment. cAdjusted for time trend, lag of white child injuries, child injuries ages 9–14 per 1 000 000 vehicles, oil consumption and unemployment. dAdjusted for time trend, lag of white child injuries, child injuries ages 9–14 per 1 000 000 child population, oil consumption and unemployment. eAdjusted for time trend, lag of non-white child injuries, child injuries ages 9–14 per 1 000 000 vehicles, oil consumption and unemployment. fAdjusted for time trend, lag of non-white child injuries, child injuries ages 9–14 per 1 000 000 child population, oil consumption and unemployment. IRR = Incidence rate ratios. Bold values indicate associations with P-values <0.05. Table 1 CRL effects on monthly injury rates per 1 000 000 child population and per 1 000 000 motor-vehicles, Brazil, 2008–2014, stratified by race Poisson Prais–Winsten AR(1) Immediate effects Gradual effects per month Immediate effects Gradual effects per month IRR (95% CI) IRR (95% CI) IRR (95% CI) IRR (95% CI) Total child injuries ages 0–8 per 1 000 000 motor-vehiclesa 0.82 (0.63–1.06) 0.98 (0.97–0.99) 0.89 (0.73–1.09) 0.98 (0.97–0.99) Total child injuries ages 0–8 per 1 000 000 children 0–8b 0.82 (0.63–1.07) 0.98 (0.97–0.99) 0.57 (0.19–1.74) 0.94 (0.90–0.99) White child injuries ages 0–8 per 1 000 000 motor-vehiclesc 0.90 (0.66–1.23) 0.97 (0.96–0.99) 0.99 (0.85–1.15) 0.98 (0.98–0.99) White child injuries ages 0–8 per 1 000 000 children 0–8d 0.91 (0.66–1.24) 0.97 (0.96–0.99) 0.71 (0.11–4.57) 0.89 (0.82–0.97) Non-white child injuries ages 0–8 per 1 000 000 motor-vehiclese 0.74 (0.53–1.03) 0.99 (0.97–1.00) 0.94 (0.87–1.00) 0.99 (0.99–1.00) Non-white child injuries ages 0–8 per 1 000 000 children 0–8f 0.74 (0.53–1.03) 0.99 (0.97–1.00) 0.49 (0.21–1.18) 1.00 (0.96–1.03) Poisson Prais–Winsten AR(1) Immediate effects Gradual effects per month Immediate effects Gradual effects per month IRR (95% CI) IRR (95% CI) IRR (95% CI) IRR (95% CI) Total child injuries ages 0–8 per 1 000 000 motor-vehiclesa 0.82 (0.63–1.06) 0.98 (0.97–0.99) 0.89 (0.73–1.09) 0.98 (0.97–0.99) Total child injuries ages 0–8 per 1 000 000 children 0–8b 0.82 (0.63–1.07) 0.98 (0.97–0.99) 0.57 (0.19–1.74) 0.94 (0.90–0.99) White child injuries ages 0–8 per 1 000 000 motor-vehiclesc 0.90 (0.66–1.23) 0.97 (0.96–0.99) 0.99 (0.85–1.15) 0.98 (0.98–0.99) White child injuries ages 0–8 per 1 000 000 children 0–8d 0.91 (0.66–1.24) 0.97 (0.96–0.99) 0.71 (0.11–4.57) 0.89 (0.82–0.97) Non-white child injuries ages 0–8 per 1 000 000 motor-vehiclese 0.74 (0.53–1.03) 0.99 (0.97–1.00) 0.94 (0.87–1.00) 0.99 (0.99–1.00) Non-white child injuries ages 0–8 per 1 000 000 children 0–8f 0.74 (0.53–1.03) 0.99 (0.97–1.00) 0.49 (0.21–1.18) 1.00 (0.96–1.03) aAdjusted for time trend, lag of child injuries, child injuries ages 9–14 per 1 000 000 vehicles, oil consumption and unemployment. bAdjusted for time trend, lag of child injuries, child injuries ages 9–14 per 1 000 000 child population, oil consumption and unemployment. cAdjusted for time trend, lag of white child injuries, child injuries ages 9–14 per 1 000 000 vehicles, oil consumption and unemployment. dAdjusted for time trend, lag of white child injuries, child injuries ages 9–14 per 1 000 000 child population, oil consumption and unemployment. eAdjusted for time trend, lag of non-white child injuries, child injuries ages 9–14 per 1 000 000 vehicles, oil consumption and unemployment. fAdjusted for time trend, lag of non-white child injuries, child injuries ages 9–14 per 1 000 000 child population, oil consumption and unemployment. IRR = Incidence rate ratios. Bold values indicate associations with P-values <0.05. Effect of the policy reform on child occupant fatalities In Table 2 we observe three main results. First, there was no association between the introduction of CRL and fatality rates in the non-white population. This is consistent across the two statistical approaches, for the different denominators, and when examining immediate and gradual effects. Second, there was an immediate reduction in fatalities observed in the total and white populations after the introduction of the CRL. The Poisson models estimated an immediate 39% reduction in monthly fatality rates after the introduction of the CRL, whereas the corresponding rates ranged from 15 to 66% in the Prais–Winsten AR(1) models, depending on the denominator. For whites, the Poisson models estimated a 52–53% decrease in monthly child fatality rates after the introduction of the CRL, whereas in the Prais Winsten AR(1) models we observed a 14% decrease with the vehicle fleet as the denominator and a 82% reduction when the child population was the denominator. Lastly, the CRL was not associated with a change in slope in the monthly rates of child occupant fatalities, suggesting an immediate effect of this policy. Table 2 CRL effects on monthly fatality rates per 1 000 000 child population and per 1 000 000 motor vehicles, Brazil, 2008–14, stratified by race Poisson Prais–Winsten AR(1) Immediate effects Gradual effects per month Immediate effects Gradual effects per month IRR (95% CI) IRR (95% CI) IRR (95% CI) IRR (95% CI) Total child fatalities ages 0–8 per 1 000 000 motor vehiclesa 0.61 (0.44–0.84) 1.00 (0.99–1.01) 0.85 (0.74–0.98) 1.00 (0.99–1.00) Total child fatalities ages 0–8 per 1 000 000 children 0–8b 0.61 (0.45–0.84) 1.00 (0.99–1.01) 0.34 (0.20–0.76) 1.01 (0.99–1.03) White child fatalities ages 0–8 per 1 000 000 motor vehiclesc 0.47 (0.33–0.68) 1.00 (0.99–1.02) 0.86 (0.76–0.96) 1.00 (0.99–1.00) White child fatalities ages 0–8 per 1 000 000 children 0–8d 0.48 (0.33–0.68) 1.00 (0.99–1.02) 0.18 (0.06–0.54) 1.02 (0.99–1.06) Non-white child fatalities ages 0–8 per 1 000 000 motor vehiclese 0.86 (0.54–1.36) 0.99 (0.98–1.01) 0.98 (0.92–1.06) 0.99 (0.99–1.00) Non-white child fatalities ages 0–8 per 1 000 000 children 0–8f 0.87 (0.55–1.37) 0.99 (0.97–1.01) 0.83 (0.38–1.78) 0.99 (0.97–1.02) Poisson Prais–Winsten AR(1) Immediate effects Gradual effects per month Immediate effects Gradual effects per month IRR (95% CI) IRR (95% CI) IRR (95% CI) IRR (95% CI) Total child fatalities ages 0–8 per 1 000 000 motor vehiclesa 0.61 (0.44–0.84) 1.00 (0.99–1.01) 0.85 (0.74–0.98) 1.00 (0.99–1.00) Total child fatalities ages 0–8 per 1 000 000 children 0–8b 0.61 (0.45–0.84) 1.00 (0.99–1.01) 0.34 (0.20–0.76) 1.01 (0.99–1.03) White child fatalities ages 0–8 per 1 000 000 motor vehiclesc 0.47 (0.33–0.68) 1.00 (0.99–1.02) 0.86 (0.76–0.96) 1.00 (0.99–1.00) White child fatalities ages 0–8 per 1 000 000 children 0–8d 0.48 (0.33–0.68) 1.00 (0.99–1.02) 0.18 (0.06–0.54) 1.02 (0.99–1.06) Non-white child fatalities ages 0–8 per 1 000 000 motor vehiclese 0.86 (0.54–1.36) 0.99 (0.98–1.01) 0.98 (0.92–1.06) 0.99 (0.99–1.00) Non-white child fatalities ages 0–8 per 1 000 000 children 0–8f 0.87 (0.55–1.37) 0.99 (0.97–1.01) 0.83 (0.38–1.78) 0.99 (0.97–1.02) aAdjusted for time trend, lag of child fatalities, child fatalities ages 9–14 per 1 000 000 motor vehicles, oil consumption and unemployment. bAdjusted for time trend, lag of child fatalities, child fatalities ages 9–14 per 1 000 000 child population, oil consumption and unemployment. cAdjusted for time trend, lag of white child fatalities, child fatalities ages 9–14 per 1 000 000 motor vehicles, oil consumption and unemployment. dAdjusted for time trend, lag of white child fatalities, child fatalities ages 9–14 per 1 000 000 child population, oil consumption and unemployment. eAdjusted for time trend, lag of non-white child fatalities, child fatalities ages 9–14 per 1 000 000 motor vehicles, oil consumption and unemployment. fAdjusted for time trend, lag of non-white child fatalities, child fatalities ages 9–14 per 1 000 000 child population, oil consumption and unemployment. IRR = Incidence rate ratios. Bold values indicate associations with P-values <0.05. Table 2 CRL effects on monthly fatality rates per 1 000 000 child population and per 1 000 000 motor vehicles, Brazil, 2008–14, stratified by race Poisson Prais–Winsten AR(1) Immediate effects Gradual effects per month Immediate effects Gradual effects per month IRR (95% CI) IRR (95% CI) IRR (95% CI) IRR (95% CI) Total child fatalities ages 0–8 per 1 000 000 motor vehiclesa 0.61 (0.44–0.84) 1.00 (0.99–1.01) 0.85 (0.74–0.98) 1.00 (0.99–1.00) Total child fatalities ages 0–8 per 1 000 000 children 0–8b 0.61 (0.45–0.84) 1.00 (0.99–1.01) 0.34 (0.20–0.76) 1.01 (0.99–1.03) White child fatalities ages 0–8 per 1 000 000 motor vehiclesc 0.47 (0.33–0.68) 1.00 (0.99–1.02) 0.86 (0.76–0.96) 1.00 (0.99–1.00) White child fatalities ages 0–8 per 1 000 000 children 0–8d 0.48 (0.33–0.68) 1.00 (0.99–1.02) 0.18 (0.06–0.54) 1.02 (0.99–1.06) Non-white child fatalities ages 0–8 per 1 000 000 motor vehiclese 0.86 (0.54–1.36) 0.99 (0.98–1.01) 0.98 (0.92–1.06) 0.99 (0.99–1.00) Non-white child fatalities ages 0–8 per 1 000 000 children 0–8f 0.87 (0.55–1.37) 0.99 (0.97–1.01) 0.83 (0.38–1.78) 0.99 (0.97–1.02) Poisson Prais–Winsten AR(1) Immediate effects Gradual effects per month Immediate effects Gradual effects per month IRR (95% CI) IRR (95% CI) IRR (95% CI) IRR (95% CI) Total child fatalities ages 0–8 per 1 000 000 motor vehiclesa 0.61 (0.44–0.84) 1.00 (0.99–1.01) 0.85 (0.74–0.98) 1.00 (0.99–1.00) Total child fatalities ages 0–8 per 1 000 000 children 0–8b 0.61 (0.45–0.84) 1.00 (0.99–1.01) 0.34 (0.20–0.76) 1.01 (0.99–1.03) White child fatalities ages 0–8 per 1 000 000 motor vehiclesc 0.47 (0.33–0.68) 1.00 (0.99–1.02) 0.86 (0.76–0.96) 1.00 (0.99–1.00) White child fatalities ages 0–8 per 1 000 000 children 0–8d 0.48 (0.33–0.68) 1.00 (0.99–1.02) 0.18 (0.06–0.54) 1.02 (0.99–1.06) Non-white child fatalities ages 0–8 per 1 000 000 motor vehiclese 0.86 (0.54–1.36) 0.99 (0.98–1.01) 0.98 (0.92–1.06) 0.99 (0.99–1.00) Non-white child fatalities ages 0–8 per 1 000 000 children 0–8f 0.87 (0.55–1.37) 0.99 (0.97–1.01) 0.83 (0.38–1.78) 0.99 (0.97–1.02) aAdjusted for time trend, lag of child fatalities, child fatalities ages 9–14 per 1 000 000 motor vehicles, oil consumption and unemployment. bAdjusted for time trend, lag of child fatalities, child fatalities ages 9–14 per 1 000 000 child population, oil consumption and unemployment. cAdjusted for time trend, lag of white child fatalities, child fatalities ages 9–14 per 1 000 000 motor vehicles, oil consumption and unemployment. dAdjusted for time trend, lag of white child fatalities, child fatalities ages 9–14 per 1 000 000 child population, oil consumption and unemployment. eAdjusted for time trend, lag of non-white child fatalities, child fatalities ages 9–14 per 1 000 000 motor vehicles, oil consumption and unemployment. fAdjusted for time trend, lag of non-white child fatalities, child fatalities ages 9–14 per 1 000 000 child population, oil consumption and unemployment. IRR = Incidence rate ratios. Bold values indicate associations with P-values <0.05. Discussion Main findings of this study Our results suggest that the introduction of Brazil’s CRL was followed by an immediate reduction in child fatalities and a gradual reduction in child injuries in the long term. However, these results were driven by significant reductions among the white population. Our estimates suggest that the CRL reduced the fatality rate by 39% for all children and 52% for white children in the first month following the reform. In terms of injuries, the CRL was associated with a 2% reduction in the overall population and a 3% reduction among white children. These figures are more optimistic than those presented by Garcia et al.25 which showed no reduction in fatalities within 1 year of CRL’s implementation. The introduction of Brazil’s CRL did not appear to affect fatality and injury rates among non-white children. What is already known on this topic Our results have implications for authorities seeking to optimize the deployment of CRL and use it as a road safety measure to protect children. First, as reported in studies carried out in other countries, the introduction of CRL was followed by a gradual decline in traffic injuries and abrupt reduction in fatalities.27 This is similar to the Brazilian case, where a 2% gradual decrease in the monthly injury rate and 39% immediate reduction in the monthly fatality rate was observed following its introduction. Second, these reductions were concentrated among the white population. It is important is to notice that while the white population has higher fatality and injury rates than the non-white population this is due mostly to different travel patterns. White children are more likely to travel in motor vehicles than non-white and therefore are more exposed as occupants to car crashes. Nevertheless, our result suggests that the partial success of the policy may be associated with access to resources which the white population in Brazil has had historically. Research in Brazil has documented that individuals who transport children in child restraints are more likely to have higher levels of education than individuals who do not transport children using these devices.41 In this regard, it is important to recall that a strong correlation between level of education and race predominate in Brazil.42,43 Lastly, it is noteworthy that the overall proportion of the population that use child restraint devices in Brazil is 57%, which is low compared to, for instance Canada, where this rate is higher than 95%.44 As such, policies to increase both use and proper installation of these devices should consider race when designing public campaigns, as well as subsidises for populations which lack monetary resources to acquire them. What this study adds This study questions the body of literature which finds that legislation can be effective in reducing injuries, but less so when assessing fatality reductions.28 Wagenaar et al.28 and Nazif-Muñoz et al.27 suggest that one may observe differences in these two outcomes since traffic fatalities are relatively rare events. However, longer study periods are likely necessary to be adequately powered to detect impacts on mortality. McLeod and Vingilis45 have indeed noted that time series evaluations of traffic safety measures increase their statistical power the longer the pre- and post-intervention periods are. We increased the power of our analysis by analyzing monthly rates, rather than annual rates as in prior studies.27,28 In our models we observed an immediate effect of introducing CRL for the reduction of fatalities and a gradual decrease of injuries, suggesting that these outcomes follow from different causal pathways. Thus, when assessing the impact of CRL, and more broadly passive safety devices, it is important to consider that these systems operate to reduce the consequences of crashes, but their effectiveness depends on factors such as speed, vehicle characteristics, road infrastructure and provision of emergency services. Further, along with the severity of the outcome one should consider their frequency since milder crashes occur more often than severe ones.46 Therefore, when contrasting the effects of a given intervention on fatality and injuries one should indeed consider an overlap between these phenomena but also their qualitatively different elements. Another element that deserves attention is the different results obtained when using Poisson and Prais–Winsten AR(1) particularly when assessing child fatalities per population. The magnitude of the results is more pronounced for the Prais–Winsten AR(1) models. This technique, however, requires the normality assumption of distribution of errors, which may have been violated in this case when examining non-negative integer-valued data such as traffic crashes. By contrast, our Poisson models preserved the integer structure of the data.47 Limitations Our conclusions regarding the effectiveness of the CRL are supported by the inclusion of control variables and time trends to account for potential confounding. This includes control for the rates of fatalities and injuries among 9–14-year-old children, which aim to capture unobserved trends. The use of two different denominators (children population and vehicle fleet) further emphasize the robustness of our findings. Despite these strengths, there were limitations to our analyses. We did not have access to information regarding the percentage of use and proper use of children restraints before the enactment of the law, preventing us from adding this component to our analyses. Second, we lacked information on police enforcement, drinking and driving behaviors and vehicle characteristics. Changes in these population-level characteristics, to the extent that they corresponded with the introduction of the CRL and influenced our outcomes, could have biased our estimates. Another Brazilian study suggested that potential explanations for a downward trend in fatal traffic injuries was related to the introduction of various policies since 1997, including improvements of the driving licence system, restrictions on drinking and driving, alcohol sales limitations, and use of airbags.16 These reforms may have confounded our estimates to the extent that they coincided with the introduction of the CRL in 2010. Table A1 Descriptive statistics (2008–14) Variable Source Mean SD Min Max Dependent variables Child injuries 0–8 per 1 000 000 motor-vehicles Information System on Hospital Admissions and National Department of Traffic 0.57 0.16 0.11 1.12 Child injuries 0–8 per 1 000 000 children 0–8 Information System on Hospital Admissions and Brazilian Institute of Geography and Statistics 3.38 0.96 0.97 5.59 White child injuries 0–8 per 1 000 000 motor-vehicles Information System on Hospital Admissions and National Department of Traffic 0.36 0.14 0.05 0.80 White child injuries 0–8 per 1 000 000 white children 0–8 Information System on Hospital Admissions and Brazilian Institute of Geography and Statistics 4.23 1.37 0.88 7.68 Non-white child injuries 0–8 per 1 000 000 motor-vehicles Information System on Hospital Admissions and National Department of Traffic 0.20 0.06 0.03 0.43 Non-white Child injuries 0–8 per 1 000 000 non-white children 0–8 Information System on Hospital Admissions and Brazilian Institute of Geography and Statistics 2.53 1.07 0.31 5.91 Child fatalities 0–8 per 1 000 000 motor-vehicles Ministry of Health’s Mortality Information System and National Department of Traffic 0.43 0.18 0.18 1.11 Child fatalities 0–8 per 1 000 000 children 0–8 Ministry of Health’s Mortality Information System and Brazilian Institute of Geography and Statistics 1.98 0.69 0.82 3.7 White child fatalities 0–8 per 1 000 000 motor-vehicles Ministry of Health’s Mortality Information System and National Department of Traffic 0.27 0.14 0.05 0.76 White child fatalities 0–8 per 1 000 000 white children 0–8 Brazilian Institute of Geography and Statistics 2.43 0.99 0.64 5.68 Non-white child fatalities 0–8 per 1 000 000 motor-vehicles Ministry of Health’s Mortality Information System and National Department of Traffic 0.16 0.07 0.03 0.37 Non-white child fatalities 0–8 per 1 000 000 non-white children 0–8 Ministry of Health’s Mortality Information System and Brazilian Institute of Geography and Statistics 1.54 0.70 0.29 3.71 Other covariates Child injuries 9–14 per 1 000 000 motor-vehicles Information System on Hospital Admissions and National Department of Traffic 0.55 0.16 0.19 0.98 Child injuries 9–14 per 1 000 000 children 10–14 Information System on Hospital Admissions and Brazilian Institute of Geography and Statistics 5.86 1.47 2.55 8.81 Child fatalities 9–14 per 1 000 000 motor-vehicles Ministry of Health’s Mortality Information System and National Department of Traffic 0.31 0.14 0.04 0.81 Child fatalities 9–14 per 1 000 000 children 10–14 Brazilian Institute of Geography and Statistics 2.25 0.80 0.44 4.81 Oil consumption (barrels/day) (thousand) Central Bank of Brazil 99.29 30.44 51.00 196.00 Unemployment (%) Central Bank of Brazil 8.32 2.47 4.60 13.10 Variable Source Mean SD Min Max Dependent variables Child injuries 0–8 per 1 000 000 motor-vehicles Information System on Hospital Admissions and National Department of Traffic 0.57 0.16 0.11 1.12 Child injuries 0–8 per 1 000 000 children 0–8 Information System on Hospital Admissions and Brazilian Institute of Geography and Statistics 3.38 0.96 0.97 5.59 White child injuries 0–8 per 1 000 000 motor-vehicles Information System on Hospital Admissions and National Department of Traffic 0.36 0.14 0.05 0.80 White child injuries 0–8 per 1 000 000 white children 0–8 Information System on Hospital Admissions and Brazilian Institute of Geography and Statistics 4.23 1.37 0.88 7.68 Non-white child injuries 0–8 per 1 000 000 motor-vehicles Information System on Hospital Admissions and National Department of Traffic 0.20 0.06 0.03 0.43 Non-white Child injuries 0–8 per 1 000 000 non-white children 0–8 Information System on Hospital Admissions and Brazilian Institute of Geography and Statistics 2.53 1.07 0.31 5.91 Child fatalities 0–8 per 1 000 000 motor-vehicles Ministry of Health’s Mortality Information System and National Department of Traffic 0.43 0.18 0.18 1.11 Child fatalities 0–8 per 1 000 000 children 0–8 Ministry of Health’s Mortality Information System and Brazilian Institute of Geography and Statistics 1.98 0.69 0.82 3.7 White child fatalities 0–8 per 1 000 000 motor-vehicles Ministry of Health’s Mortality Information System and National Department of Traffic 0.27 0.14 0.05 0.76 White child fatalities 0–8 per 1 000 000 white children 0–8 Brazilian Institute of Geography and Statistics 2.43 0.99 0.64 5.68 Non-white child fatalities 0–8 per 1 000 000 motor-vehicles Ministry of Health’s Mortality Information System and National Department of Traffic 0.16 0.07 0.03 0.37 Non-white child fatalities 0–8 per 1 000 000 non-white children 0–8 Ministry of Health’s Mortality Information System and Brazilian Institute of Geography and Statistics 1.54 0.70 0.29 3.71 Other covariates Child injuries 9–14 per 1 000 000 motor-vehicles Information System on Hospital Admissions and National Department of Traffic 0.55 0.16 0.19 0.98 Child injuries 9–14 per 1 000 000 children 10–14 Information System on Hospital Admissions and Brazilian Institute of Geography and Statistics 5.86 1.47 2.55 8.81 Child fatalities 9–14 per 1 000 000 motor-vehicles Ministry of Health’s Mortality Information System and National Department of Traffic 0.31 0.14 0.04 0.81 Child fatalities 9–14 per 1 000 000 children 10–14 Brazilian Institute of Geography and Statistics 2.25 0.80 0.44 4.81 Oil consumption (barrels/day) (thousand) Central Bank of Brazil 99.29 30.44 51.00 196.00 Unemployment (%) Central Bank of Brazil 8.32 2.47 4.60 13.10 Table A1 Descriptive statistics (2008–14) Variable Source Mean SD Min Max Dependent variables Child injuries 0–8 per 1 000 000 motor-vehicles Information System on Hospital Admissions and National Department of Traffic 0.57 0.16 0.11 1.12 Child injuries 0–8 per 1 000 000 children 0–8 Information System on Hospital Admissions and Brazilian Institute of Geography and Statistics 3.38 0.96 0.97 5.59 White child injuries 0–8 per 1 000 000 motor-vehicles Information System on Hospital Admissions and National Department of Traffic 0.36 0.14 0.05 0.80 White child injuries 0–8 per 1 000 000 white children 0–8 Information System on Hospital Admissions and Brazilian Institute of Geography and Statistics 4.23 1.37 0.88 7.68 Non-white child injuries 0–8 per 1 000 000 motor-vehicles Information System on Hospital Admissions and National Department of Traffic 0.20 0.06 0.03 0.43 Non-white Child injuries 0–8 per 1 000 000 non-white children 0–8 Information System on Hospital Admissions and Brazilian Institute of Geography and Statistics 2.53 1.07 0.31 5.91 Child fatalities 0–8 per 1 000 000 motor-vehicles Ministry of Health’s Mortality Information System and National Department of Traffic 0.43 0.18 0.18 1.11 Child fatalities 0–8 per 1 000 000 children 0–8 Ministry of Health’s Mortality Information System and Brazilian Institute of Geography and Statistics 1.98 0.69 0.82 3.7 White child fatalities 0–8 per 1 000 000 motor-vehicles Ministry of Health’s Mortality Information System and National Department of Traffic 0.27 0.14 0.05 0.76 White child fatalities 0–8 per 1 000 000 white children 0–8 Brazilian Institute of Geography and Statistics 2.43 0.99 0.64 5.68 Non-white child fatalities 0–8 per 1 000 000 motor-vehicles Ministry of Health’s Mortality Information System and National Department of Traffic 0.16 0.07 0.03 0.37 Non-white child fatalities 0–8 per 1 000 000 non-white children 0–8 Ministry of Health’s Mortality Information System and Brazilian Institute of Geography and Statistics 1.54 0.70 0.29 3.71 Other covariates Child injuries 9–14 per 1 000 000 motor-vehicles Information System on Hospital Admissions and National Department of Traffic 0.55 0.16 0.19 0.98 Child injuries 9–14 per 1 000 000 children 10–14 Information System on Hospital Admissions and Brazilian Institute of Geography and Statistics 5.86 1.47 2.55 8.81 Child fatalities 9–14 per 1 000 000 motor-vehicles Ministry of Health’s Mortality Information System and National Department of Traffic 0.31 0.14 0.04 0.81 Child fatalities 9–14 per 1 000 000 children 10–14 Brazilian Institute of Geography and Statistics 2.25 0.80 0.44 4.81 Oil consumption (barrels/day) (thousand) Central Bank of Brazil 99.29 30.44 51.00 196.00 Unemployment (%) Central Bank of Brazil 8.32 2.47 4.60 13.10 Variable Source Mean SD Min Max Dependent variables Child injuries 0–8 per 1 000 000 motor-vehicles Information System on Hospital Admissions and National Department of Traffic 0.57 0.16 0.11 1.12 Child injuries 0–8 per 1 000 000 children 0–8 Information System on Hospital Admissions and Brazilian Institute of Geography and Statistics 3.38 0.96 0.97 5.59 White child injuries 0–8 per 1 000 000 motor-vehicles Information System on Hospital Admissions and National Department of Traffic 0.36 0.14 0.05 0.80 White child injuries 0–8 per 1 000 000 white children 0–8 Information System on Hospital Admissions and Brazilian Institute of Geography and Statistics 4.23 1.37 0.88 7.68 Non-white child injuries 0–8 per 1 000 000 motor-vehicles Information System on Hospital Admissions and National Department of Traffic 0.20 0.06 0.03 0.43 Non-white Child injuries 0–8 per 1 000 000 non-white children 0–8 Information System on Hospital Admissions and Brazilian Institute of Geography and Statistics 2.53 1.07 0.31 5.91 Child fatalities 0–8 per 1 000 000 motor-vehicles Ministry of Health’s Mortality Information System and National Department of Traffic 0.43 0.18 0.18 1.11 Child fatalities 0–8 per 1 000 000 children 0–8 Ministry of Health’s Mortality Information System and Brazilian Institute of Geography and Statistics 1.98 0.69 0.82 3.7 White child fatalities 0–8 per 1 000 000 motor-vehicles Ministry of Health’s Mortality Information System and National Department of Traffic 0.27 0.14 0.05 0.76 White child fatalities 0–8 per 1 000 000 white children 0–8 Brazilian Institute of Geography and Statistics 2.43 0.99 0.64 5.68 Non-white child fatalities 0–8 per 1 000 000 motor-vehicles Ministry of Health’s Mortality Information System and National Department of Traffic 0.16 0.07 0.03 0.37 Non-white child fatalities 0–8 per 1 000 000 non-white children 0–8 Ministry of Health’s Mortality Information System and Brazilian Institute of Geography and Statistics 1.54 0.70 0.29 3.71 Other covariates Child injuries 9–14 per 1 000 000 motor-vehicles Information System on Hospital Admissions and National Department of Traffic 0.55 0.16 0.19 0.98 Child injuries 9–14 per 1 000 000 children 10–14 Information System on Hospital Admissions and Brazilian Institute of Geography and Statistics 5.86 1.47 2.55 8.81 Child fatalities 9–14 per 1 000 000 motor-vehicles Ministry of Health’s Mortality Information System and National Department of Traffic 0.31 0.14 0.04 0.81 Child fatalities 9–14 per 1 000 000 children 10–14 Brazilian Institute of Geography and Statistics 2.25 0.80 0.44 4.81 Oil consumption (barrels/day) (thousand) Central Bank of Brazil 99.29 30.44 51.00 196.00 Unemployment (%) Central Bank of Brazil 8.32 2.47 4.60 13.10 Conclusions Our results suggest that CRLs are partially effective in reducing child injuries following motor vehicle collisions. The enactment of the CRL in Brazil in September 2010 resulted in a gradual reduction in child occupant injuries in the long term, as well as an immediate reduction in child fatalities. However, these trends were concentrated among the white population. These findings are in line with the World Health Organization’s and the American Association of Pediatrics’ recommendations to introduce child safety measures to tackle transport-related injuries in children. Nonetheless complementary efforts need to be targeted to reach and protect vulnerable populations such as non-white children in Brazil. References 1 Bhalla K , Shotten M , Cohen A et al. ( 2014 ) Transport for Health: The Global Burden of Disease From Motorized Road Transport. http://documents.worldbank.org/curated/en/984261468327002120/pdf/863040IHME0T4H0ORLD0BANK0compressed.pdf (date last accessed, 24 November 2017). 2 Hyder AA , Vecino-Ortiz AI . BRICS: opportunities to improve road safety . Bull World Health Organ 2014 ; 92 ( 6 ): 423 – 8 . Google Scholar Crossref Search ADS PubMed WorldCat 3 Oh JJE . Russia. In: Pojani D , Stead D (eds) . The Urban Transport Crisis in Emerging Economies . Zurich : Springer International Publishing , 2017 : 197 – 219 . Google Preview WorldCat COPAC 4 Pojani D , Stead D . The urban transport crisis in emerging economies: an introduction. In: Pojani D , Stead D (eds) . In The Urban Transport Crisis in Emerging Economies . Zurich : Springer International Publishing , 2017 : 1 – 10 . Google Preview WorldCat COPAC 5 Armijo LE , Rhodes SD . Explaining infrastructure underperformance in Brazil: cash, political institutions, corruption, and policy Gestalts . Policy Stud 2017 ; 38 ( 3 ): 231 – 47 . Google Scholar Crossref Search ADS WorldCat 6 Reddy BS , Balachandra P . Urban mobility: a comparative analysis of megacities of India . Transport Policy 2012 ; 21 : 152 – 64 . Google Scholar Crossref Search ADS WorldCat 7 Hyder AA , Paichadze N , Toroyan T et al. Monitoring the decade of action for global road safety 2011–2020: an update . Global Public Health 2016 ; 12 : 1492 – 1505 . Google Scholar Crossref Search ADS PubMed WorldCat 8 Chiavegatto Filho ADP , Beltrán-Sánchez H , Kawachi I . Racial disparities in life expectancy in Brazil: challenges from a multiracial society . Am J Public Health 2014 ; 104 ( 11 ): 2156 – 62 . Google Scholar Crossref Search ADS PubMed WorldCat 9 de Azevedo Barros MB , Lima MG , Medina LDPB et al. Social inequalities in health behaviors among Brazilian adults: National Health Survey, 2013 . Int J Equity Health 2016 ; 15 ( 1 ): 148 . Google Scholar Crossref Search ADS PubMed WorldCat 10 Barreto ML , Teixeira MG , Bastos FI et al. Successes and failures in the control of infectious diseases in Brazil: social and environmental context, policies, interventions, and research needs . Lancet 2011 ; 377 ( 9780 ): 1877 – 89 . Google Scholar Crossref Search ADS PubMed WorldCat 11 Paim J , Travassos C , Almeida C et al. The Brazilian health system: history, advances, and challenges . Lancet 2011 ; 377 ( 9779 ): 1778 – 97 . Google Scholar Crossref Search ADS PubMed WorldCat 12 Rasella D , Aquino R , Santos CA et al. Effect of a conditional cash transfer programme on childhood mortality: a nationwide analysis of Brazilian municipalities . Lancet 2013 ; 382 ( 9886 ): 57 – 64 . Google Scholar Crossref Search ADS PubMed WorldCat 13 Victora CG , Barreto ML , do Carmo Leal M et al. Health conditions and health-policy innovations in Brazil: the way forward . Lancet 2011 ; 377 ( 9782 ): 2042 – 53 . Google Scholar Crossref Search ADS PubMed WorldCat 14 Auger N , Le Serbon E , Rasella D et al. Impact of homicide and traffic crashes on life expectancy in the largest Latin American country . J Public Health (Bangkok) 2016 ; 38 ( 3 ): 467 – 73 . Google Scholar Crossref Search ADS WorldCat 15 de Souza MDFM , Gawryszewski VP , Ordunez P et al. Cardiovascular disease mortality in the Americas: current trends and disparities . Heart 2012 ; 98 ( 16 ): 1207 – 12 . Google Scholar Crossref Search ADS PubMed WorldCat 16 Reichenheim ME , De Souza ER , Moraes CL et al. Violence and injuries in Brazil: the effect, progress made, and challenges ahead . The Lancet 2011 ; 377 ( 9781 ): 1962 – 75 . Google Scholar Crossref Search ADS WorldCat 17 Morais-Neto OL , Montenegro MDMS , Monteiro RA et al. Mortalidade por acidentes de transporte terrestre no Brasil na última década: tendência e aglomerados de risco . Ciência Saúde Coletiva 2012 ; 17 ( 9 ): 2223 – 36 . Google Scholar Crossref Search ADS PubMed WorldCat 18 Morais-Neto OL , Montenegro MDMS , Monteiro RA et al. Road traffic injuries—profile and trends—Brasil, 2004–2013. In: Health Brazil 2014 a Situational Analysis of Road Injuries and Other External Causes . Brasilia : Ministry of Health of Brazil Publisher , 2015 : 117 – 44 . Google Preview WorldCat COPAC 19 Morais-Neto OLM , Andrade AL , Guimarães RA et al. Regional disparities in road traffic injuries and their determinants in Brazil, 2013 . Int J Equity Health 2016 ; 15 ( 1 ): 142 . Google Scholar Crossref Search ADS PubMed WorldCat 20 Liberatti CLB , Andrade SD , Soares DA . The new Brazilian traffic code and some characteristics of victims in southern Brazil . Injury Prevent 2001 ; 7 ( 3 ): 190 – 3 . Google Scholar Crossref Search ADS WorldCat 21 ANTP. Associação Nacional de Transportes Públicos . ( 2015 ) Sistema de Informações da Mobilidade Urbana Relatório Comparativo 2003–201. Retrieved from Assoc Nac Transp Públicos. http://files-derver.antp.org.br/_5dotSystem/userFiles/SIMOB/Relatorio%20Comparativo%202013.pdf (date last accessed, 24 November 2017). 22 República Federativa de Brasil [Federative Republic of Brazil] . ( 2008 ) Resolução no. 277, de 28 de maio de 2008. [Resolution 277, edited on 28 May 2008] Dispõe sobre o transporte de menores de 10 anos e a utilização do dispositivo de retenção para o transporte de crianças em veículos. [Provision for the transport of children under 10 years and the use of the restraint device for the transport of children in motor vehicles]. Diário Oficial da União [Federal Official Gazette of Brazil]. 23 Federative Republic of Brazil . ( 2016 ) Lei no. 9.503, de 23 de setembro de 1997 [Law 9503, edited on 23 September 1997] Diário Oficial da União [Federal Official Gazette of Brazil]. 24 Conselho Estadual de Trânsito . ( 2009 ) A Criança no Trânsito. Retrived from Conselho Estadual de Trânsito website: http://www.atividadeseducativas.com.br/atividades/11900_Cartilha_Crianca.pdf 25 Garcia LP , Freitas LRSD , Duarte EC . Avaliação preliminar do impacto da Lei da Cadeirinha sobre os óbitos por acidentes de automóveis em menores de dez anos de idade, no Brasil: estudo de séries temporais no período de 2005 a 2011 . Epidemiol Serv Saúde 2012 ; 21 ( 3 ): 367 – 74 . Google Scholar Crossref Search ADS WorldCat 26 Elvik R , Vaa T , Hoye A et al. (eds) . The Handbook of Road Safety Measures . Emerald Group Publishing , 2009 . Google Preview WorldCat COPAC 27 Nazif-Muñoz JI , Falconer J , Gong A . Are child passenger fatalities and child passenger severe injuries equally affected by child restraint legislation? The case of Chile . Int J Inj Contr Saf Promot 2017 ; 24 : 501 – 509 . Google Scholar Crossref Search ADS PubMed WorldCat 28 Wagenaar AC , Webster DW , Maybee RG . Effects of child restraint laws on traffic fatalities in eleven states . J Trauma Acute Care Surg 1987 ; 27 ( 7 ): 726 – 32 . Google Scholar Crossref Search ADS WorldCat 29 Andrews GR . Racial inequality in Brazil and the United States, 1990-2010 . J Soc Hist 2014 ; 47 ( 4 ): 829 – 54 . Google Scholar Crossref Search ADS WorldCat 30 Linden A . Conducting interrupted time-series analysis for single-and multiple-group comparisons . Stata J 2015 ; 15 ( 2 ): 480 – 500 . Google Scholar Crossref Search ADS WorldCat 31 McCleary R , Hay R . Applied Time Series Analysis for the Social Sciences . Thousand Oaks, CA : Sage Publications, Inc , 1980 . 32 DATASUS . ( 2017 ) Informações de Saúde. Assistência à Saúde [Health information. Health and Assistance]. http://www2.datasus.gov.br/DATASUS/index.php?area=0202 (date last accessed, 24 November 2017). 33 Instituto Brasileiro de Geografia e Estatística [Brazilian Institute of Geography and Statistics] . ( 2010 ) Population Census. [Database on the Internet] Instituto Brasileiro de Geografia e Estatística [Brazilian Institute of Geography and Statistics]. http://www.ibge.gov.br/english/estatistica/populacao/projecao_da_populacao/2013/default.shtm (date last accessed, 24 November 2017). 34 Departamento Nacional de Trânsito [National Department of Traffic] . ( 2017 ) Frota de veiculos [Vehicle fleet]. http://www.denatran.gov.br/estatistica/237-frota-veiculos. (date last accessed, 24 November 2017). 35 Banco Central do Brasil [Central Bank of Brazil]. ( 2017 ) Economic Indicators. https://www.bcb.gov.br/pec/Indeco/Ingl/indecoi.asp (date last accessed, 24 November 2017). 36 World Health Organization . ICD-10: International Statistical Classification of Diseases and Related Health Problems: Tenth Revision . Geneva : World Health Organization , 2004 . Google Preview WorldCat COPAC 37 Scuffham PA . Economic factors and traffic crashes in New Zealand . Appl Econ 2003 ; 35 ( 2 ): 179 – 88 . Google Scholar Crossref Search ADS WorldCat 38 Sen A , Mizzen B . Estimating the impact of seat belt use on traffic fatalities: empirical evidence from Canada . Canad Public Policy 2007 ; 33 ( 3 ): 315 – 35 . Google Scholar Crossref Search ADS WorldCat 39 Brubacher JR , Desapriya E , Erdelyi S et al. The impact of child safety restraint legislation on child injuries in police-reported motor vehicle collisions in British Columbia: an interrupted time series analysis . Paediatr Child Health 2016 ; 21 ( 4 ): e27 – 31 . Google Scholar Crossref Search ADS PubMed WorldCat 40 Wooldridge JM . Introductory Econometrics: A Modern Approach . Nelson Education , 2015 . Google Preview WorldCat COPAC 41 Crianças Segura . ( 2012 ) Uso de Cadeirinha no Automovel [Child restraint use in motor-vehicles]. http://criancasegura.org.br/wp-content/uploads/2016/08/15-1.pdf (date last accessed, 24 November 2017). 42 Marteleto LJ . Educational inequality by race in Brazil, 1982–2007: structural changes and shifts in racial classification . Demography 2012 ; 49 ( 1 ): 337 – 58 . Google Scholar Crossref Search ADS PubMed WorldCat 43 Marteleto LJ , Dondero M . Racial inequality in education in Brazil: a twins fixed-effects approach . Demography 2016 ; 53 ( 4 ): 1185 – 1205 . Google Scholar Crossref Search ADS PubMed WorldCat 44 Bruce B , Cramm C , Mundle K et al. Roadside observation of child passenger restraint use . Adv Pediatr Res 2015 ; 2 : 24 . WorldCat 45 McLeod AI , Vingilis ER . Power computations in time series analyses for traffic safety interventions . Accid Anal Prevent 2008 ; 40 ( 3 ): 1244 – 8 . Google Scholar Crossref Search ADS WorldCat 46 World Health Organization . Global Status Report on Road Safety . Geneva : World Health Organization , 2015 . Google Preview WorldCat COPAC 47 Quddus MA . Time series count data models: an empirical application to traffic accidents . Accid Anal Prevent 2008 ; 40 ( 5 ): 1732 – 41 . Google Preview WorldCat COPAC Appendix © The Author(s) 2018. Published by Oxford University Press on behalf of Faculty of Public Health. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)
TI - Protecting only white children: the impact of child restraint legislation in Brazil
JF - Journal of Public Health
DO - 10.1093/pubmed/fdy105
DA - 2019-06-01
UR - https://www.deepdyve.com/lp/oxford-university-press/protecting-only-white-children-the-impact-of-child-restraint-gfxees3d0P
SP - 287
VL - 41
IS - 2
DP - DeepDyve
ER -