RESUMEN

The list of occurrences linked to significant climate change has grown in recent decades. These changes can be influenced by a set of covariates, such as temperature, location and period of the year. Analyzing the relation among elements and factors that influence the behavior of such events is extremely important for decision-making in order to minimize damages and losses. Exceedance analysis uses the tail of the distribution based on Extreme Value Theory (EVT). Extensions for these models have been proposed in literature, such as regression models for the tail parameters and a parametric or semi-parametric distribution for the part that comes before the tail (well known as bulk distribution). This work presents a new extension to exceedance model, in which the parameters for the bulk distribution capture the effect of covariates such as location and seasonality. We considered a Bayesian approach in the inference procedure. The estimation was done using MCMC -- Markov Chain Monte Carlo methods. Application results for modeling maximum and minimum temperature data showed an efficient estimation of extreme quantiles and a predictive advantage compared to models previously used in literature.

RESUMEN

Introduction: Occupational accidents represent a severe and complex public health issue. Objectives: To identify temporal trends in occupational mortality in Brazil from 2010 to 2019. Methods: This was an ecological study with time series analysis using data from the Brazilian Ministry of Health Mortality Information System (Ministério da Saúde/ Sistema de Informações sobre Mortalidade). The mortality rate was calculated using Prais-Winsten estimation. Results: In the study period, 34,683 work-related deaths were recorded in Brazil, with a higher occurrence among White (51.0%) men (94.3%) aged 20 to 39 years (44.8%). The highest proportion of deaths (16.5%) was identified in the state of São Paulo. Regarding sex, temporal trends were stable for both men (annual percentage change [APC] = -1.7; 95%CI -3.9 to 0.7) and women (APC = -0.8; 95%CI -1.8 to 0.2). The age groups up to 19 years (APC = -5.1; 95%CI -9.0 to l.l) and 20 to 39 years (APC = -3.3; 95%CI -6.0 to -0.5) showed a decreasing trend, while the remaining age groups showed a stable trend. Black race had a decreasing trend (APC = -8.1; 95%CI -10.7 to -5.5), while White (APC = -2.3; 95%CI -38.0 to 0.2) and mixed races (APC = -1.2; 95%CI -5.2 to 2.9) had a stable trend. Eight states showed a decreasing trend; only the state of Pará (APC = 2.1; 95%CI 0.8 to 3.4) showed an increasing trend, while the other states had a stable trend. Conclusions: Temporal trends in occupational mortality were stable for most of the indicators evaluated. There is a lack of measures contributing to occupational safety and health in Brazil.

Introdução: Acidentes de trabalho representam um grave e complexo problema de saúde pública. Objetivos: Analisar a tendência temporal da mortalidade por acidentes de trabalho no Brasil no período de 2010 a 2019. Métodos: Estudo ecológico de séries temporais, com dados do Sistema de Informações sobre Mortalidade. Calculou-se a taxa de mortalidade utilizando regressão de Prais-Winsten. Resultados: Foram registrados 34.683 óbitos decorrentes de acidentes de trabalho no país, com maior ocorrência no sexo masculino (94,3%), na faixa etária de 20 a 39 anos (44,8%) e na raça branca (51,0%). A maior proporção de óbitos foi identificada no estado de São Paulo (16,5%). A tendência temporal apresentou estabilidade no sexo masculino (variação percentual anual (VPA) = -1,7; IC95% -3,9 a 0,7) e feminino (VPA = -0,8; IC95% -1,8 a 0,2). A tendência revelou decréscimo nas faixas etárias de até 19 anos (VPA = -5,1; IC95% -9,0 a 1,1) e de 20 a 39 anos (VPA = -3,3; IC95% -6,0 a -0,5), enquanto as demais faixas apresentaram estabilidade. A raça preta apresentou decréscimo (VPA = -8,1; IC95% -10,7 a -5,5), enquanto a branca (VPA = -2,3; IC95% -38,0 a 0,2) e a parda (VPA = -1,2; IC95% -5,2 a 2,9) apresentaram estabilidade. Oito unidades da federação apresentaram decréscimo; apenas o estado do Pará (VPA = 2,1; IC95% 0,8 a 3,4) apresentou acréscimo, enquanto as demais unidades registraram estabilidade. Conclusões: A tendência temporal da taxa de mortalidade apresentou estabilidade na maior parte dos indicadores avaliados. Percebe-se a carência de discussões que possam contribuir com ações no campo da segurança e saúde no trabalho.

RESUMEN

Extreme Value Theory (EVT) aims to study the tails of probability distributions in order to measure and quantify extreme events of maximum and minimum. In river flow data, an extreme level of a river may be related to the level of a neighboring river that flows into it. In this type of data, it is very common for flooding of a location to have been caused by a very large flow from an affluent river that is tens or hundreds of kilometers from this location. In this sense, an interesting approach is to consider a conditional model for the estimation of a multivariate model. Inspired by this idea, we propose a Bayesian model to describe the dependence of exceedance between rivers, where we considered a conditionally independent structure. In this model, the dependence between rivers is captured by modeling the excess marginally of one river as a consequence of linear functions of the other rivers. The results showed that there is a strong and positive connection between excesses in one river caused by the excesses of the other rivers.