Análise de uma década de acidentes graves nas rodovias federais do Ceará no período 2008-2017

Abstract

Traffic accidents are a social problem that must be faced by public power and society because of the enormous financial and social damage they cause. The traffic accidents on the highways and federal roads of Ceará in the period 2008-2017 were mapped according to several explanatory variables, among them Urbanized Location, which portrays the level of urbanization as a function of the frequency of accidents (dynamic criterion), Road Structure, Year, Beginning of the Month, Holidays, Day of the Week, Phase of the Day and Type of Accident. Due to the high cost, the fatal accidents stood out among the serious ones and led to the development of logistic models of spatial, temporal and specific nature. These models identified the variables that contributed, with a confidence level higher than 0.95, to the occurrence of fatality on highways and federal roads. Among the results obtained, it can be mentioned that the chance of fatality in a major accident occurred in the highways and federal roads of Ceará in the period 2008-2017 was: 3.9 times greater in the rural area compared to the large urban centers; 2.5 times higher at dawn compared to full day; 84% higher on straight lines compared to crosses; 36% higher in 2017 compared to 2016; 17% higher at the beginning of the month compared to the rest of the month; 21% lower on Thursdays compared to Sundays; 129% higher in accidents resulting from incompatible speed compared to disobedience to signaling; 60% higher in accidents due to undue oversight compared to lack of attention; 339% higher in person-to-person car rash compared to rear-end collisions; and 235% higher in frontal collisions compared to rear collisions. Others have also drawn attention, such as vacations and holidays, which did not contribute to the increase in fatality among major accidents, although the presence of holidays contributed to the increase in the rate of serious accidents per day. This finding clearly showed that the variables that best describe major accidents (in terms of absolute frequency) are not necessarily the ones that best describe fatal accidents involving fatality.