Road Safety as a Public Health Problem: Case of Ecuador in the Period 2000–2019

: Background: Road safety is a signiﬁcant public health problem because it causes negative consequences on victims and families. The objective was to analyze the most signiﬁcant changes in trafﬁc crashes in Ecuador during the period from 2000 to 2019. With data obtained from the National Institute of Statistics and Census, we performed the analysis to identify: the number of trafﬁc crashes, the number of victims, and other study variables. Methods: Descriptive and analytical statistics and the contrast of proportions were used to analyze data from 2000 to 2019. Results: According to the ideal joinpoint analysis model, there was a signiﬁcant decrease in the number of recorded trafﬁc accidents from 2015 to 2019 of − 8.54 per year, while the tendency to die increased in females (2.05 per year) and males (3.29 per year). The most common crash was a collision, and the automobile appeared as the most involved vehicle from 2015 to 2019. The hypothesis test contrast is used to determine if statistically signiﬁcant differences exist between age groups by gender of the driver injured in the period 2017–2018. Conclusions: This study determines the most signiﬁcant changes in the variables related to trafﬁc crashes, where mortality due to this cause in the last four years has had a growth rate of 1.8% compared to collisions that presented a rate of − 31.12%. The contrast of the hypothesis test shows signiﬁcant differences in the injury level between males and female drivers, depending on the age group.


Introduction
Traffic accidents represent a problem that affects public health and road safety, giving rise to a mortality rate of 18.2 per 100,000 inhabitants worldwide, whereby the serious injuries generated by them are the eighth leading cause of death, causing 1.35 million deaths in 2016. They are currently the first cause of death in children and adolescents between 5 and 29 years old, generating 54% of deaths, which are vulnerable people such as pedestrians, cyclists, and motorcyclists. Ninety percent of deaths due to traffic accidents occur in low-and middle-income countries [1]. Therefore, the United Nations presented the Global Plan for the Decade of Action for Road Safety 2011-2020, where the goal is to halve the number of road traffic fatalities [2].
The Americas region shows an increasing trend in mortality due to traffic accidents [3]. In addition, Poswayo [4] states that traffic accidents are preventable, which means a reduction in the number of deaths and injuries. In 2016, deaths from this cause in this region cause 11% of deaths worldwide, with an average regional mortality rate of 15.6 per 100,000 inhabitants, and serious injuries represented the second leading cause of death in adolescents between 15 and 29 years of age.

State of the Art
Some studies analyze traffic accident patterns [12], risk behaviors [13,14], causes of traffic accidents [9,15], types of traffic collisions [8], and time of occurrence [16]. They also determine that the severity of accidents depends on the vehicles involved [17,18] and demographic characteristics [18][19][20][21][22]. In general, these investigations used historical data on traffic accidents and are descriptive. Several of them focus on recommendations to reduce the number of victims of traffic accidents and their consequences, such as the need to intervene to improve driver education for good driving behavior, improve vehicle safety, and provide better quality information for sound decision making.
Research on sex differences generally concludes that the crash rate or risk of being actively involved in a road traffic crash (especially fatal crashes) remains higher for men [24,25,[27][28][29][30][31], although the crash rate for both sexes is influenced by age in the first place, with higher rates for young [24,25,30,32,33] and older [30,32] drivers.
The researchers agreed, males have a higher accident traffic rate because they are more involved "voluntarily" in risky behavior, mainly among younger drivers, which are more prone to "search of excitement" and are less experienced drivers [24,25,[31][32][33]. The authors [28,29,31,33] have concluded that males and young drivers are more likely to break the law. The study [34] concluded that females are easily distracted and incur more perception faults than males. In addition, other authors have stated that male's higher accident traffic rates are related to higher exposure [31][32][33]. However, exposure data are hard to collect and unreliable, and recourse to alternative methods such as indirect or quasiinduced exposure is necessary [25,32,[35][36][37][38]. The investigation [39] driver collision patterns addressed by sex and age regarding offenses, collision type, and injury severity using clustering methodology and tests of proportions are used for a complimentary analysis when driver offense is present, concluding there are significant differences between males and females, showing that this tool is very useful to support conclusions.
The 2020 Jharkhand (Republic of India) study shows that 73.4% of males and 26.6% of females died from traffic crashes, with the majority being in the 21-30 age range [20]. Furthermore, in Ghana (Africa) in 2017, males are six times more likely to die from traffic crashes than females, with the highest incidence of fatalities in the age ranges of children and the elderly [21].
Previous studies carried out in Latin America have determined certain dangerous behaviors that affect road safety, among which are: exceeding speed limits, driving under the influence of alcohol, not wearing seat belts, or engaging in distracted driving [40,41]. In addition, there is research such as that carried out in Chile, which shows that the time of day when most traffic accidents occur is in the afternoon [16].
In a study conducted in Brazil [12], they determine that the rear-end collision is the most common type of traffic accident; furthermore, in another study in this country, in the state of Piauí, they analyze historical data from 2000 to 2017, where they define that the vehicle with the highest number of traffic accidents and responsible for almost half of the deaths was the motorcycle [18].
Traffic accidents represent a problem studied in different countries, which measure the impact and propose solutions according to the situation of each country. In the case of Ecuador, the Congacha study [42] determined that 96.3% of traffic accidents are explained by ten causes, with inattentive driving being the predominant one, with an incidence of 56.8%, coinciding with other studies [9,15]. Another study conducted in Ecuador [43] states that the national statistics on traffic accidents collected by the responsible institutions have few variables which are not sufficiently disaggregated, and if they are available, not all of them are of good quality, which does not allow for more in-depth accident rate studies.
In Bhatti work [44], they state that underreporting exists in low-income countries, especially for light injuries [45], partly due to the lack of a well-structured data collection system [46,47]. This underreporting can be identified and avoided through police and hospital data reviews and software to standardize information [48], which is why it is essential to implement improvements in data collection in order to know the progress of road safety [49] and to allow comparison with other countries [50].
The National Traffic Agency (ANT) is the entity that keeps statistics on accidents and road accidents in the country. In recent years, it has categorized the probable causes of traffic accidents into 12 variables until 2013 and into 22 variables as of 2014 [51]. The variables collected in the official form of traffic accidents are in the following factors: (a) human factor, (b) vehicle factor, (c) environment factor, (d) to be defined, and (e) others, and it has not been possible to obtain from the official source, the concepts and cases encompassed in groups (d) and (e). From the exhaustive review carried out, two aspects stand out, namely the lack of and deficiency of information, on some crucial variables used to develop models that explain the causality of traffic accidents. The information collected at the accident site only allows general descriptive studies and obtaining the overall mortality rate (traffic accidents per total population) for comparative purposes between countries and regions of the world.
In line with the above, the International Transport Forum recommends that Ecuador build and sustain capacity and investment in road safety research and data systems. In particular, it recommends the following: radically improving the vehicle registration system and providing a more effective means of vehicle regulation, collecting traffic and travel data to help assess risk exposure and analyze road safety problems [52]. From a Vision Zero perspective, it requires the continued analysis of historical crash data systems that must be accessible, structured, and secure in the first place. Countries leading the way in road safety have done so through robust, continuous, and sophisticated information systems. They can better assess the risks of road traffic deaths and injuries, prioritize road safety interventions based on accurate data, and track their outcomes.
Therefore, the main objective of this work is to analyze the most significant changes in the number of traffic accidents registered in Ecuador from 2000 to 2019 and if there are differences from the statistical point of view, in the level of injury of drivers by gender and age groups, according to data availability in two years (2017-2018).

Data Source
The data for this study came from the National Institute of Statistics and Census, specifically from: From the transport yearbooks published from the period 2000-2019 of traffic crashes, this base prepared by INEC with data from the National Traffic Agency, from which extracted information for the following variables: (a) the number of traffic accidents, (b) the number of victims, slightly injured and seriously injured (on-site severity), (c) causes of traffic accidents, (d) types of traffic accidents, (e) vehicle involved, and (f) time of traffic crashes; all these variables are collected on site (24 h).
The General Death Statistical Register provided data on fatalities from 2000 to 2019. From this database, the variables obtained were (a) number of deaths by traffic crashes, (b) sex, and (c) age; the latter recorded since 2008. According to INEC's methodology, data are collected monthly and published annually [53].
The data for inference and contrast of proportions between men and women drivers are prepared by INEC with data from the National Traffic Agency [53], where data on the severity of drivers according to sex and age groups are displayed in two years (2017-2018).

Study Variables
The information on the variables used in the study is shown in Table 1. It is limited to the variables available in the accessible data on traffic accidents in Ecuador.

Study Method
Descriptive statistics were performed, showing summarized information of the data, presenting averages. The relationship between demographic characteristics (sex) and onsite victims (deceased, seriously injured, and slightly injured) was analyzed from 2000 to 2013 using the chi-square test, because as of 2014, the entities in charge stopped collecting these variables and recorded them as a single variable (injured) without disaggregation.
The traffic crashes recorded evaluated by regression analysis using Joinpoint software (Joinpoint regression software version 4.8.0.1-April 2020, available from the US National Cancer Institute's Surveillance Research Program) [54]. The analysis performed in the software served to identify, through joinpoints, the most significant changes in the data entered, which helped identify the variables that are most relevant to the study, in addition to using the Monte Carlo permutation method present in the program to select the appropriate joinpoints by comparing two different joinpoints [55][56][57]. The Grid Search method was selected [58].
The program determined the Annual Percentage Change (APC), with a 95% Confidence Interval (CI) for each analysis performed, which adjusted the regression by the natural logarithm of the rate, where the independent variable was the years of the period analyzed. In addition, the Provided Average Annual Percentage Change (AAPC) was obtained, which is a measure of the entire range of data. In the data of gender and age of the victims, an advanced analysis called Pairwise Comparison was performed, which had the purpose of seeing if the two regression mean functions were similar using the Parallel test, where a value of p < 0.05 was statistically significant for the tests [59].

The Contrast Proportions and Rates' Methodology
This section evaluates the differences between male and female drivers in terms of severity of crashes that occurred in Ecuadorian roads from data encompassing the period 2017-2018. Starting from the general perception, according to which there are behavioral differences assigned in both groups toward traffic problems, such as those related to the levels of risk assumed and to the respect for traffic rules, which directly influence the number and types of road crashes and the injury level of drivers, it is necessary to make a risk comparison taking into account gender as well age. To this end, the exposure plays a fundamental role, but it is not available because the complexity and the cost to obtain directly. To face this problem, the methodology of contrast proportions is used for the analysis of statistically significant differences in the severity of male and female drivers, according to age intervals, as the available and disaggregated data allow it (2017-2018).

Statistically Significant Differences between Males and Females Drivers Injured in Crashes
For statistically significant differences determination between male and female drivers involved in road crashes, regarding the harmfulness by age group, we have applied proportions and rates' contrast methodology.
The test of proportions tries to analyze whether the sample data allow rejecting the hypothesis of equality of proportions of a determined characteristic of two populations or, on the contrary, the differences between the proportions of the sample, obtained from the data, are too much to accept it or, in statistical terminology, not to reject it.
The statistic, i.e., the function of the data applied in this case, is the difference between the proportions divided by an estimate of the standard deviation of this difference: (1) The above equation, under the null hypothesis of equality of population proportions, is distributed approximately as a standard Gaussian; thus, when (d) takes values in the tails of the distribution, typically outside the interval (−2,2), the hypothesis must be rejected, since the behavior of the samples is peripheral or inconsistent with what is expected if equality of proportions were to hold.

Data Descriptive and Regression Analysis by Joinpoint Results
The total number of traffic accidents and injury victims (fatalities and slightly and seriously injured), a ratio per 100,000 inhabitants, and their variations in the period 2000-2019 are presented in Table 2.  [60]. In addition, before the dollarization of the economy (the year 2000), the country had a stagnant gross domestic product growth rate of 1.8% per year. In the following years, GDP averaged 4.4% [61], and this growth may have contributed to the increase in traffic accidents [62,63].
Within the period analyzed, the highest number of traffic accidents occurred in the year 2014 with 8.63% of the total number of accidents, while the lowest was in the year 2003 with a 2.31%. This figure could be due to the change in the Regulation to the Law of Ground Transportation, Transit, and Road Safety, which entered in January of that same year, where it reinforced the curricula for non-professional driver training schools and also modified the maximum blood alcohol limit allowed for driving (0.8 g/L) [8].
The analysis of the database of the National Traffic Agency from 2000 to 2019 reveals the absence of some variables such as data of road users (pedestrian, passenger, and driver), type of driving license, type of road, road conditions or weather conditions, vehicle damage, type and age of vehicle, and, in some years, it lacks the number of serious and minor injuries, age and sex of the victims, which ostensibly limits the possibility of the scientific study of traffic accidents with a greater level of depth. The crash data collection system needs to be improved and that it is essential to carry out studies to broaden the knowledge of road safety in Ecuador. Tools and resources for this purpose are appearing every day and require exhaustive, continuous, rigorous, and complete information systems. The variations in criteria over time make it practically impossible to carry out rigorous studies on the evolution of accident conditions and to evaluate the effects of measures for improvement [64]. The definition of policies and measures requires more and better knowledge extracted from the data.
About the previous paragraphs, traffic accident data for Ecuador from 2000 to 2019 were analyzed. Regression analysis using Joinpoint software helps to identify the most significant changes in the period. See Figure 1. minor injuries, age and sex of the victims, which ostensibly limits the possibility of the scientific study of traffic accidents with a greater level of depth. The crash data collection system needs to be improved and that it is essential to carry out studies to broaden the knowledge of road safety in Ecuador. Tools and resources for this purpose are appearing every day and require exhaustive, continuous, rigorous, and complete information systems. The variations in criteria over time make it practically impossible to carry out rigorous studies on the evolution of accident conditions and to evaluate the effects of measures for improvement [64]. The definition of policies and measures requires more and better knowledge extracted from the data. About the previous paragraphs, traffic accident data for Ecuador from 2000 to 2019 were analyzed. Regression analysis using Joinpoint software helps to identify the most significant changes in the period. See Figure 1.   66], non-compliance with traffic law, vehicles with little or no safety equipment. As of 2016, they decreased by −8.54 per year (95% CI: −20.2 to 4.9), with 24.2% of the total number of traffic accidents; this decrease can probably be attributed has the reform of the land transport law regulation [67], in addition to the investment in road improvement, construction, and reconstruction [68], since these types of investments positively affect traffic accidents according to Calvo-Poyo study [69]. Likewise, the increase in breathalyzer tests for drivers and the implementation of computerized radars [70] have been part of the state surveillance program, which was demonstrated as a very effective measure to improve road safety in other countries. Table 3 shows the types of vehicles involved in traffic accidents from 2015 to 2019 in Ecuador. Before 2015, this variable was not in the records. It is the type of vehicle against which people are impacted or occupied at the time of the accident, which is of utmost importance in crashworthiness and severity studies. From Table 3 In the work of Pillajo-Quijia [71], they apply machine-learning methods to explain the severity of driver injuries in hit-and-run accidents involving light commercial vehicles, such as vans and pickup trucks, and vehicle types are one of the influential factors.
The study conducted in the state of Piauí (Brazil) indicates that the motorcycle produced the highest number of deaths [18], while in Ecuador between 2015 and 2019 did not present any similarity; however, this may change in the future and will motivate studies. It is worth mentioning that in Ecuador, the growth rate of motorcycle sales between 2018 and 2019 was 12.31%, while the sales of automobiles decreased by 7.9% [72]. Table 4 shows the causes that have generated traffic accidents in the period 2000-2019 in Ecuador. Table 4 shows that driver recklessness and carelessness is the predominant cause (representing 41.76%). Furthermore, the Joinpoint analysis indicates that from 2000 to 2013, this cause increased annually by 12.97 (95% CI: from 10.3 to 15.6), and since 2014, it decreased by −2.61 (95% CI: from −9.7 to 5.0). This result is similar to that obtained in previous studies from Ecuador [15,42]: the predominant cause is driver recklessness and carelessness, followed by speeding. A study conducted in Colombia on traffic fatalities also found that the causes with the highest fatality rate were reckless driving and speeding [41], and more examples can be found around the world.
The other causes did not show a decreasing trend over the years: disrespect of traffic signs (representing 9.49% of the causes of traffic accidents), invasion of oncoming traffic lanes with 9.07%, and speeding with 14.19%. The state of these results should motivate the monitoring and surveillance of compliance with the rules by the authorities in the field. Excessive speed is one of the principal causes of fatalities and minor injuries in traffic accidents, of which the WHO [1] points out that no corrective measures have been taken. There is sufficient evidence and agreement that an average 5% reduction in speed can lead to a 20% reduction in the number of fatal traffic accidents [5,73]. . With these types of updates, Ecuador demonstrates that it has taken actions in speed enforcement, which can be associated with the decrease in traffic accidents as of 2016. In addition, this may be due to drivers' behaviors and the training they receive, as they do not respect speed limits and traffic signs [75]. Table 5 shows the timetable of occurrence of traffic accidents in Ecuador from 2015 to 2019. The highest number of traffic accidents occurred between 12:00 am and 5:59 pm with 29.75%, coinciding with the study conducted in Chile [16], followed by 6:00 pm to 11:59 pm with 27.58%. The number of traffic accidents during the early morning hours was lower (16.55%) in Ecuador, which was a discrepancy with the study conducted in Cartagena (Colombia), where the highest number of traffic accidents occurred during the early morning hours [76].
The Joinpoint analysis showed a decreasing trend in traffic collisions during the following hours: from 0:00 to 5:59 h −2. 55    The comparison of traffic crashes between the morning (0:00 to 11:59 h) and the afternoon (12:00 to 23:59 h) showed that the decrease in traffic crashes between the two schedules was parallel (p = 0.10) with a total reduction of −8.63 per year as shown in Figure 2. Table 5 shows the number of fatalities by gender recorded in 24 h from 2000 to 2013.
Collision is one of the most frequent configurations, and the severity changes according to the type of collision. Head-on collisions are the most injurious compared to rear-end collisions, and being the predominant accident type in Ecuador, there was no categorization until 2013. In the last years, a side collision represented 58.49%, followed by rear collision at 21.53%, frontal collision at 13.81%, and other collisions at 6.18%. Research conducted in Colombia shows that collisions between vehicles are the fourth leading cause of death [77]. In the study conducted in Brazil [12], rear-end collisions are the leading cause of traffic accidents. Table 8 shows the variation of annual deaths concerning those registered in 24 h in the period 2000-2019 caused by traffic accidents and the ratio of deaths per 100,000 inhabitants in Ecuador.
The growth rate of the number of fatalities (annual) was 1.8% from 2015 to 2019. Table 8 shows that the number of males who die in traffic accidents is higher than that of females; approximately six males and two females die per day in traffic accidents in Ecuador. Other research [20,21] shows that sex is a factor that influences traffic accidents in several countries [23,26,27].
The average rate of deaths per 100,000 inhabitants is 18.2 in Ecuador between 2000 and 2019; the result obtained is close to those recorded in 2016 by countries in Latin America and the Caribbean, highlighting the worrying situation for several countries such as El Salvador, Paraguay, Guyana, and Belize with a rate value around 20 with a rate value around 20 deaths per 100,000 inhabitants [78]. If we compare this with the European region, which had a rate of 9.3 deaths per 100,000 inhabitants [79] in the same year, the dimension of the health problem to be faced in Ecuador becomes evident.
When comparing the annual death data between INEC and those recorded in situ (24 h) by ANT, there was an average variation of 35.97% between 2000 and 2019. This finding is very similar to that of other countries [43,45,48]. According to Furas's study, underreporting in Latin America varies from 12% in Colombia to 29% in Ecuador [80]. The situation is also affected by the lack of disaggregated variables in Ecuador, highlighting the need for a well-structured data collection system, as recommended by international institutions for Ecuador [1,52].  In the Joinpoint analysis, the most significant changes were identified in the number of deaths according to sex (per year), as shown in Figure 3, with females presenting an increase of 2.05 per year (95% CI: 1.2 to 2.9) and males increasing by 3.29 per year (95% CI: 2.3 to 4.3). In the comparability test, the number of deaths (males and females) due to In the Joinpoint analysis, the most significant changes were identified in the number of deaths according to sex (per year), as shown in Figure 3, with females presenting an increase of 2.05 per year (95% CI: 1.2 to 2.9) and males increasing by 3.29 per year (95% CI: 2.3 to 4.3). In the comparability test, the number of deaths (males and females) due to traffic accidents was not parallel (the selected model rejected parallelism, p = 0.0004). There are international studies that conclude that females are involved in fewer traffic accidents [23,26,31,34]. Underlying these numbers are socioeconomic factors determining a higher number of males in the driver census. The late incorporation of females into the labor market, family characteristics, and other cultural barriers determines this configuration of driver censuses in large regions worldwide. For a rigorous comparison, it is necessary to have a measure of risk exposure, which is the mainstay of accident studies and, at the same time, one of the costliest variables to obtain (surveys, naturalistic studies, among others) from large sectors of the population to ensure its representativeness. Around the world, researchers are working on alternative means of elicitation, exploring new data sources, and in the Quasi-Induced Exposure (QIE) methodology, from data contained in accident databases. In the paper [81], they have developed a methodology for improving driver crash liability useful for the QIE approach. In addition, in research papers, it can be found that the age is a factor to be taken into account. In this paper, the methodology of contrast of proportions is developed to analyze the statistically significant differences in the severity of male and female drivers, according to age intervals, as the available and disaggregated data allow it (2017-2018). The analysis of the results can be found in Section 4.2. Table 9 shows the age range of fatalities involved in traffic accidents registered in the period 2008-2019 in Ecuador.  2008  6  168  316  685  478  372  226  193  244  3  2009  21  230  369  797  556  389  287  236  287  4  2010  11  213  356  809  546  433  356  251  318  11  2011  20  224  349  884  590  435  292  215  335  7  2012  15  205  355  811  573  384  308  217  304  4  2013  14  172  342  856  574  368  314  208  292  6  2014  13  158  312  827  554  378  313  222  324  0  2015  16  162  339  845  554  392  285  193  335  2  2016  16  155  264  740  531  401  279  236  315  1  2017  11  130  293  830  542  366  298  213  334  0  2018  12  125  311  862  588  380  321  228  313  2  2019  16  127  264  926  579  409  311  224  320  3 From Table 9, it is noteworthy that of those killed in road traffic accidents 26.44% belong to the age range of 20-29 years (p = 0.03), followed by the group of 30-39 years (17.85%) (p = 0.01), 40-49 years (12.60%) (p = 0.04), over 69 years (9.96%) (p = 0.03), 50-59 years (9.61%) (p = 0.01), and 1-9 years (5.54%) (p = 0.01). The study conducted in Jharkhand (Republic of India) and those conducted by WHO, PAHO, and OISEVI coincide with the age range of the highest number of road traffic fatalities, which is 20-29 years [1, 3,20,75], but unlike the work conducted in Ghana [21], the results are different, with the most vulnerable road users being children under ten years of age and the elderly. Figure 4 shows the comparison analysis obtained from 2000 to 2019 in Ecuador, according to the age of those killed (annually) in traffic accidents.

Hypothesis Tests Results
We want to analyze the sample data to either not reject the hypothesis of equal proportions of a given characteristic of the feature of two populations or to do so if the differences between the sample proportions are too larger, considering the statistical (d) value (defined in Section 3.4) to accept or reject the hypothesis test.
In Table 10, the number of injured drivers in accidents traffic in Ecuador in the period 2017-2018 segregated by injury level according to the sex and age group is shown. The figures agreed with it is known: if absolute numbers are analyzed, there are more injured males than females, in all the levels of injury and each age group. It is important to keep in mind that the exposure is different between them, and as mentioned above, the exposure is not simple or directly available. To solve the problem, the risk comparison is made using the contrast proportions methodology to determine if there are significant differences in the proportion by injury level by sex and age intervals. The proportions are computed concerning the total of males or females, respectively, and we analyze if the differences between the sample proportions, as obtained from the data, are too large to accept in the statistical terminology, not to reject it. The results of the contrast are in Table 11.

Hypothesis Tests Results
We want to analyze the sample data to either not reject the hypothesis of equal proportions of a given characteristic of the feature of two populations or to do so if the differences between the sample proportions are too larger, considering the statistical (d) value (defined in Section 3.4) to accept or reject the hypothesis test.
In Table 10, the number of injured drivers in accidents traffic in Ecuador in the period 2017-2018 segregated by injury level according to the sex and age group is shown. The figures agreed with it is known: if absolute numbers are analyzed, there are more injured males than females, in all the levels of injury and each age group. It is important to keep in mind that the exposure is different between them, and as mentioned above, the exposure is not simple or directly available. To solve the problem, the risk comparison is made using the contrast proportions methodology to determine if there are significant differences in the proportion by injury level by sex and age intervals. The proportions are computed concerning the total of males or females, respectively, and we analyze if the differences between the sample proportions, as obtained from the data, are too large to accept in the statistical terminology, not to reject it. The results of the contrast are in Table 11. The existence of significant differences is measured by the 95% threshold for the normal statistic which is 2. In Table 11, the mean value and Low and Up CI are shown beside the normal statistic of the hypothesis test contrast, besides the ratio between the proportions.
In road collisions in Ecuador of the period analyzed, there exist significant sex differences for injured levels regarding the age group. In the case of deceased level, the risk is different by gender in the 31-55 years age group, and the differences decrease to be not significant for more younger and senior drivers by gender. The ratio between the proportions in Table 11 is M/F = 2.96 from 18 to 30, 3.94 from 31 to 55, and 1.40 in the interval of 56-75 years.
In the injury level, the contrast of the hypothesis test allows one to conclude that there is more risk to be injured within the sample of female drivers, and there are significant differences with males in two age groups encompassing from 31 to 75 years. The ratios indicate a larger proportion of females in these intervals (1.15 and 1.32, respectively). This analysis remarks the importance of ratios analysis, instead of absolute numbers, showing that conclusions can differ depending on gender and age.

Limitation of the Study
This study used the information on traffic accidents, available annually from INEC; in which, there was no disaggregation of variables, for example, types of collisions, road users, and did not record light and serious injuries as of 2015. From 2019, the data improved, but they are still unreliable.
The availability of disaggregated information for drivers in two years (2017-2018) made it possible to apply the contrast of hypotheses, leading to conclusions with significant nuances.

Conclusions
We conclude that from 2000 to 2014, the decrease in traffic collisions has not been constant; there are years when they increase. In addition, the number of deaths has increased in most years, making road safety a public health issue for Ecuador. The analysis of the results indicates that since the new traffic law came into force in Ecuador at the end of 2014, there has been a 31.12% decrease in the number of traffic accidents up to 2019. Despite this decrease, traffic fatalities show a worrying growth rate of 1.8%, and if this trend continues, Ecuador has not met the target 3.6 of the SDG 2020 and has to adopt its measures to comply with the new decade on road safety until 2030.
The analysis revealed that 41.76% of traffic accidents in Ecuador are caused by recklessness and carelessness of drivers, such as: driving inattentive to traffic conditions, driving while drowsy or in poor physical condition, and not keeping the minimum lateral safety distance between vehicles, which, in general, are risky behaviors on the road. In addition, the type of traffic accident with the highest incidence is a collision with an average of 45.29%. From 2015 to 2019, the automobile has a higher probability of collision with 35.74%, and motorcycles have a lower probability of collision with 16.55%. The predominant time of traffic collisions recorded from 2015 to 2019 arises in the afternoon hours between 12: 00 and 17:59 with 29.75%.
The analysis by sex shows that men are more likely to die in traffic accidents, with an average of approximately 2182 (80%) deaths, and women 551 (20%) deaths in the period studied, the most affected age range is between 20 and 29 years of age, followed by those between 30 and 39 years of age. From these data, it is not known whether the victims correspond to pedestrians, drivers, cyclists, and other road users, because it is not recorded by those responsible for road safety.
The hypothesis test analysis remarks the importance of considering ratios instead of absolute numbers, showing significant differences between males and females, depending on the age group and level of injury. In the case of deceased level, the risk is different for males and females in the 31-55 years age group, and the differences decrease to be not significant for more younger and senior drivers of both sexes. The ratio between the proportions in Table 11 is M/F = 2.96 from 18 to 30, 3.94 from 31 to 55, and 1.40 in the interval of 56-75 years.
For the injury level, the hypothesis test contrast allows one to conclude that there is a higher risk of being injured within the sample of female than men drivers, and there are significant differences between them in two age groups: from 31 to 75 years old. The ratios indicate a higher proportion of females in these intervals (1.15 and 1.32, respectively).
Finally, the data analysis remarks the recording process must be improved; this would contribute to aid future research to learn about the worried health problem and the influencing factors and would propose more concrete solutions to reduce traffic crashes in Ecuador.