We want to be able to use information about the traffic flows at road junctions and covariates describing those junctions to predict the number of accidents occurring there. We develop here a Bayesian predictive approach. Initially we considered three simpler but related problems to assess the efficiency of some approximation techniques, namely: (I) Given a treatment with an effect that can be described mathematically as of a multiplicative form, we record Poisson countings before and after the treatment is applied. Then, given a new individual with a known counting before the treatment is used, we want to predict the outcome on that individual after the treatment is applied. (II) After observing the value on an individual before any treatment is applied, we decide, based on that value, which of two treatments to apply, and then register the post- treatment outcome. Given a new individual, with an observed value before he receives any treatment, we aim to derive the predictive distribution for the outcome after one of the treatments is used. (This problem is also considered when several possible treatments are available). (III) We compare the effects of two treatments, through a two-period crossover design. We assume that both the treatment effect and the period effect are of multiplicative forms. Estimative and approximation methods are developed for each of these problems. We use the Gibbs sampling approach, normal asymptotic approximations for the posterior distributions and the Laplace approximations. Examples are presented to compare the efficiency and performance of the different methods. We find that the Laplace method performs well, and has computational advantages over the other methods. Using the knowledge obtained solving these simpler problems we develop solutions for the traffic accidents problem and analyse a real data set. Stepwise procedures for the incorporation of the covariates through the use of Kullback-Leibler measure of divergence are developed. We also consider the three simpler problems assuming that the observations are exponentially and binomially distributed.

Made available in DSpace on 2016-08-10T10:55:26Z (GMT). No. of bitstreams: 1 Gelcio Sisterolli de Carvalho.pdf: 703636 bytes, checksum: ba6a5ab20cd9da7785e509791de705ec (MD5) Previous issue date: 2004-04-20 / The morbidity and mortality caused by traffic accidents is related to the process of urbanization and with the natural history of disease supported by multifactorial theory with coorte in reserved prognostics and, for the gravity of the events, determinant of the medical-sanitary, familiar, social and economic variable. Goiânia, from 1996 to 2002, experienced an increase of 36,18% in the number of deaths by traffic accidents. The rate of proportional mortality in the chapter of the external causes presented a median of 38,8% and sex distribution of 84,65% male and 15,35% female. Running overs, cyclists, motorcyclists, events involving cars had presented increases in the period of, respectively, 168,29%, 840%, 1.050% and 450%. This study it demonstrated that, of all the deaths occurred in Goiânia by traffic accidents, approximately, 40% had involved adolescents and adults up to 29 years. The information available through the Mortality Data System, based on death certificates (DC) issued by the Institute of Forensic Medicine (IFM), presents deficiencies caused by bad death notification form fulfilling. Evaluating 527 deaths by traffic accidents, in 2002, 97,53% had no information on the degree of instruction of the victim; 62,04% did not mention the presence of medical assistance; 35,83% classified the events as "other terrestrial transportation accidents", not characterizing the victim and the vehicle; 96,39% were not reported as work accidents; and not a single form had the location of the accident. The objective of obtaining a reliable geographic distribution of deaths, using a Geographical Information System, was compromised because of the impossibility to establish, with statistical significance the location of deaths. The information used was retrieved from the databases of the Criminology Institute (only 16,51% of the total death available) and of the Integrated System of Attendance to the Trauma and Emergencies of Goiás State (only 20,68%of the total death available). Traffic accidents is an event where the human being assumes the condition of both victim and villain of the process, therefore the epidemiologic control of this important cause of death will be possible with an interdisciplinary approach, with professionals from different sectors, through readings involving ethics and citizenship, consolidating efficient measures to develop Public Health in the urban space of Goiânia. / A morbi-mortalidade patrocinada pelos acidentes de trânsito está relacionada com o processo de urbanização das polis e com a história natural da doença sustentada na teoria multifatorial com coorte de prognóstico reservado e, pela gravidade dos eventos, determinante para o surgimento de variáveis médico-sanitárias, familiares, sociais e econômicas. Goiânia, no período de 1996 a 2002, experimentou um crescimento da mortalidade pelos acidentes de trânsito da ordem de 36,18%. O índice de mortalidade proporcional no Capítulo das Causas Externas apresentou uma mediana de 38,8% e, por sexo, índices de 84,65% no masculino contra 15,35%, no feminino. Os atropelamentos, eventos envolvendo ciclistas, motociclistas e carros apresentaram aumentos no período de, respectivamente, 168,29%, 840%, 1.050% e 450%. Este estudo demonstrou que, de todos os óbitos ocorridos em Goiânia por acidentes de trânsito, aproximadamente, 40% envolveram adolescentes e adultos até 29 anos. Com relação às informações geradas através do Sistema de Informações sobre Mortalidade, a partir das Declarações de Óbitos emitidas pelo Instituto Médico Legal de Goiânia, apresentam deficiências na qualidade do preenchimento do instrumento de notificação. Na avaliação de 527 óbitos por acidentes de trânsito, no ano de 2002: 97,53% ignoraram o grau de instrução da vítima; 62,04% não mencionaram a presença de assistência médica; 35,83% classificaram os eventos como outros acidentes de transporte terrestre , sem a caracterização da vítima e do veículo; 96,39% não foram relacionados ao trabalho da vítima; e nenhuma declaração localizou o logradouro do acidente. A tentativa de georreferenciar em um mapa, utilizando um Sistema de Informação Geográfico, foi comprometida em função da impossibilidade de estabelecer, com dados significativos estatisticamente, o resgate de informações utilizando as bases de dados do Instituto de Criminalística (16,51%) e do Sistema Integrado de Atendimento ao Trauma e Emergências do Estado de Goiás (20,68%). Por se tratar de causas evitáveis onde o homem assume a condição de vítima e vilão do próprio processo, o controle epidemiológico desta importante causa de óbito será possível com medidas interdisciplinares, multiprofissionais e intersetoriais, através das leituras envolvendo a ética e a cidadania, consolidando medidas eficazes para desenvolver a Saúde Pública no espaço urbano de Goiânia.

Thesis (Ph. D.)--University of North Texas, Aug., 2007. / Title from title page display. Includes bibliographical references.