Análise de sobrevivência: teoria e aplicações em saúde / Survival analysis: theory and applications in health

O objetivo apoiar o desenvolvimento metodológico em análise de dados longitudinais, através da aplicação, teste e disseminação de métodos estatísticos inovadores no contexto da saúde, particularmente os voltados para a análise de sobrevida

[Risk factors for respiratory morbidity at 12 to 36 months in very low birth weight premature infants previously admitted to a public neonatal intensive care unit]. / Fatores associados à morbidade respiratória entre 12 e 36 meses de vida de crianças nascidas de muito baixo peso oriundas de uma UTI neonatal pública.

The aim of this paper was to estimate respiratory morbidity and its determinants for premature infants aged 12 to 36 months. The population comprised 84 infants from a cohort of very low birth weight premature infants. The outcome was the respiratory morbidity incidence rate. The relationship between the independent variables and respiratory morbidity was estimated using a Poisson regression model. From 12 to 24 months of age, 56.3% of children had experienced at least one episode of respiratory disease. >From 24 to 36 months, 38.1% of children were affected. Variables significantly associated with respiratory morbidity were bronchopulmonary dysplasia (RR = 1.9; 95%CI: 1.2-2.9), abnormal lung compliance (RR = 1.6; 95%CI: 1.1-2.3), neonatal pneumonia (RR = 2.8; 95%CI: 1.9-4.0), patent ductus arteriosus (RR = 1.6; 95%CI: 1.1-2.5), and respiratory morbidity in the first year of life (RR = 1.7; 95%CI: 1.2-2.5). The incidence of respiratory morbidity remains high in this group of high-risk infants, which calls for regular follow-up and effective interventions to prevent respiratory disease and to improve the quality of life of these children and their families.

Fatores associados à morbidade respiratória entre 12 e 36 meses de vida de crianças nascidas de muito baixo peso oriundas de uma UTI neonatal pública / Risk factors for respiratory morbidity at 12 to 36 months in very low birth weight premature infants previously admitted to a public neonatal intensive care unit

O objetivo do estudo foi estimar a morbidade respiratória entre 12 e 36 meses em crianças prematuras e identificar os fatores associados. A população compreendeu 84 crianças de uma coorte de prematuros de muito baixo peso. O desfecho foi a taxa de incidência de morbidade respiratória. A associação entre as variáveis independentes e morbidade respiratória foi verificada por modelo linear generalizado. Entre 12 e 24 meses, 56,3 por cento das crianças apresentaram morbidade respiratória. Entre 24 e 36 meses, 38,1 por cento das crianças foram acometidas. As variáveis associadas à morbidade respiratória foram: displasia broncopulmonar (RT = 1,9; IC95 por cento: 1,2-2,9), complacência pulmonar alterada (RT = 1,6; IC95 por cento: 1,1-2,2), pneumonia neonatal (RT = 2,8; IC95 por cento: 2,0-4,0), persistência do canal arterial (RT = 1,6; IC95 por cento: 1,1-2,4) e morbidade respiratória no primeiro ano de vida (RT = 1,8; IC95 por cento: 1,3-2,6). A incidência de morbidade respiratória entre 12 e 36 meses se manteve elevada neste grupo de crianças de alto risco, o que reforça a necessidade de acompanhamento e de intervenções efetivas na prevenção do adoecimento e na melhora da qualidade de vida destas crianças e suas famílias.

The aim of this paper was to estimate respiratory morbidity and its determinants for premature infants aged 12 to 36 months. The population comprised 84 infants from a cohort of very low birth weight premature infants. The outcome was the respiratory morbidity incidence rate. The relationship between the independent variables and respiratory morbidity was estimated using a Poisson regression model. From 12 to 24 months of age, 56.3 percent of children had experienced at least one episode of respiratory disease. >From 24 to 36 months, 38.1 percent of children were affected. Variables significantly associated with respiratory morbidity were bronchopulmonary dysplasia (RR = 1.9; 95 percentCI: 1.2-2.9), abnormal lung compliance (RR = 1.6; 95 percentCI: 1.1-2.3), neonatal pneumonia (RR = 2.8; 95 percentCI: 1.9-4.0), patent ductus arteriosus (RR = 1.6; 95 percentCI: 1.1-2.5), and respiratory morbidity in the first year of life (RR = 1.7; 95 percentCI: 1.2-2.5). The incidence of respiratory morbidity remains high in this group of high-risk infants, which calls for regular follow-up and effective interventions to prevent respiratory disease and to improve the quality of life of these children and their families.

Alternatives in modeling of body mass index as a continuous response variable and relevance of residual analysis.

This article presents alternatives for modeling body mass index (BMI) as a continuous variable and the role of residual analysis. We sought strategies for the application of generalized linear models with appropriate statistical adjustment and easy interpretation of results. The analysis included 2,060 participants in Phase 1 of a longitudinal study (Pró-Saúde Study) with complete data on weight, height, age, race, family income, and schooling. In our study, the residual analysis of models estimated by maximum likelihood methods yielded inadequate adjustment. The transformed response variable resulted in a good fit but did not lead to estimates with straightforward interpretation. The best alternative was to apply quasi-likelihood as the estimation method, presenting a better adjustment and constant variance. In epidemiological data modeling, researchers should always take trade-offs into account between adequate statistical techniques and interpretability of results.

Alternatives in modeling of body mass index as a continuous response variable and relevance of residual analysis

This article presents alternatives for modeling body mass index (BMI) as a continuous variable and the role of residual analysis. We sought strategies for the application of generalized linear models with appropriate statistical adjustment and easy interpretation of results. The analysis included 2,060 participants in Phase 1 of a longitudinal study (Pró-Saúde Study) with complete data on weight, height, age, race, family income, and schooling. In our study, the residual analysis of models estimated by maximum likelihood methods yielded inadequate adjustment. The transformed response variable resulted in a good fit but did not lead to estimates with straightforward interpretation. The best alternative was to apply quasi-likelihood as the estimation method, presenting a better adjustment and constant variance. In epidemiological data modeling, researchers should always take trade-offs into account between adequate statistical techniques and interpretability of results.

Neste artigo, discutem-se alternativas de modelagem do índice de massa corporal (IMC), analisado como variável contínua, e a análise de resíduos. Buscaram-se estratégias de aplicação dos modelos lineares generalizados adequadas tanto do ponto de vista do ajuste estatístico quanto da facilidade de interpretação dos resultados. Nestas análises, foram incluídos dados relativos a 2.060 participantes da Fase 1 de estudo longitudinal (Estudo Pró-Saúde), com informação completa de peso, estatura, idade, raça/cor, renda familiar e escolaridade. Em nosso estudo, a análise de resíduos dos modelos estimados pelo método da máxima verossimilhança, amplamente utilizado, não possibilitou ajuste adequado dos modelos aos dados. A transformação da variável resposta, apesar de resultar em um bom ajuste, não conduziu a estimativas de fácil interpretação. Considerou-se como melhor alternativa a mudança do método de estimação para quase-verossimilhança. Assim, melhor ajuste foi alcançado e a variância permaneceu constante. Na modelagem de dados epidemiológicos, cabe aos pesquisadores buscarem o melhor equilíbrio entre a aplicação adequada de técnicas estatísticas e a facilidade de interpretação dos dados.

Modelling over week patterns of alcohol consumption.

AIMS: This study aims to analyze alcohol consumption patterns throughout a week, controlled by socio-demographic characteristics, and to discuss the adequacy of the complex models employed. METHODS: The sample included 496 participants, from both sexes, > or =40 years old and with 7-day dietary records. Bayesian generalized additive mixed models (GAMM) were applied using two approaches: a multinomial model, with three categories of alcohol consumption behaviour including; non-drinkers, alcohol during meals only and alcohol at any time; and a gamma model for drinkers which considered the total amount of alcohol ingested per day. RESULTS: The multinomial model captured two different patterns of alcohol consumption: a sharp increase in consumption on weekends for mealtime only drinkers, the dominant behaviour among drinkers and a linear increase from Monday towards Sunday for those who drank at anytime. The effect of higher education changed from slightly protective for mealtime only drinkers to risky for anytime drinkers. The amount of alcohol consumed presents a pattern similar to the meals-only drinking. CONCLUSIONS: Alcohol consumption increased during the week. Two different alcohol consumption patterns were identified according to drinking behaviours. The methodological approach utilized was essential in uncovering these patterns.

[Associations between schooling, income, and body mass index among public employees at an university in Rio de Janeiro, Brazil: the Pró-Saúde Study]. / Associações entre escolaridade, renda e Indice de Massa Corporal em funcionários de uma universidade no Rio de Janeiro, Brasil: estudo Pró-Saúde.

This study focuses on associations between schooling, income, and body mass index (BMI). The analyses are based on data from 3,963 public university employees in Rio de Janeiro, Brazil, participating in phase 1 of a longitudinal study (the Pró-Saúde Study). ANOVA, Wald test, and generalized linear models were used to analyze differences between subgroups. Obesity prevalence was inversely associated with schooling, especially among women (p < 0.001). In regression analyses, schooling and per capita income were not associated with BMI among men. In women, schooling but not income was significantly and inversely associated with BMI (p < 0.001). Thus, low schooling plays an important role in the social determination of obesity, especially among women.

Associações entre escolaridade, renda e índice de Massa Corporal em funcionários de uma universidade no Rio de Janeiro, Brasil: estudo Pró-Saúde / Associations between schooling, income, and body mass index among public employees at an university in Rio de Janeiro, Brazil: the Pró-Saúde Study

O objetivo deste estudo foi investigar a associação entre escolaridade e renda com o índice de Massa Corporal (IMC). Nestas análises, foram estudados 3.963 funcionários de uma universidade no Rio de Janeiro, Brasil, participantes da fase 1 de um estudo longitudinal (Estudo Pró-Saúde). Para testar as diferenças entre os subgrupos, utilizaram-se análise de variâncias, teste de Wald e modelos lineares generalizados. A prevalência de obesidade variou inversamente com o nível educacional, especialmente entre as mulheres (p < 0,001). Nas análises de regressão múltipla, observou-se que, entre os homens, a educação e a renda familiar per capita não foram associadas com maior IMC. Entre as mulheres, a educação, mas não a renda, foi significativa e inversamente associada com IMC mais elevado (p < 0,001). Assim, a baixa escolaridade exerce um papel importante na determinação social da obesidade, principalmente entre mulheres.

This study focuses on associations between schooling, income, and body mass index (BMI). The analyses are based on data from 3,963 public university employees in Rio de Janeiro, Brazil, participating in phase 1 of a longitudinal study (the Pró-Saúde Study). ANOVA, Wald test, and generalized linear models were used to analyze differences between subgroups. Obesity prevalence was inversely associated with schooling, especially among women (p < 0.001). In regression analyses, schooling and per capita income were not associated with BMI among men. In women, schooling but not income was significantly and inversely associated with BMI (p < 0.001). Thus, low schooling plays an important role in the social determination of obesity, especially among women.

Análise de sobrevida: teoria e aplicações em saúde

Esta obra contém exercícios de modelagem baseados em problemas e investigações reais de análise de sobrevida aplicados em quase três anos de trabalho das autoras. Apresenta modernas técnicas de modelagem estatística, entre elas o software livre R. A publicação é um instrumento acessível e completo para quem trabalha tanto na área da epidemiologia geral quanto da clínica e em áreas das ciências biológicas que utilizem a técnica de análise de sobrevida.

Time series analysis of deaths due to diarrhoea in children in Rio de Janeiro, Brazil, 1980-1998.

Diarrhoea is still a considerable public-health problem in developing countries, especially among children aged less than five years. The well-known relationship between seasonal variation and aetiological agents of diarrhoeal diseases helps inform the decisions about the prevention and control measures. The aim of this study was to identify the temporal patterns of deaths due to diarrhoea in children, aged less than five years, in Rio de Janeiro State from 1980 to 1998. The study analyzed data on monthly deaths due to diarrhoea; the data were supplied by the Natural Mortality Information System (NMIS, Ministry of Health of Brazil). Auto Regressive Integrated Moving Average modelling was applied to the data. Using this model, the huge decline and the winter peaks could be highlighted, and they may express the predominance of rotavirus as the aetiology of diarrhoeal deaths, suggesting that an effective vaccine is the main measure for the prevention and control of severe diarrhoea.