Conferências
Conferência 1 - Conferência de abertura - 29/05 (11:30 – 12:30) – Anfiteatro
Palestrante: Tarylee Reddy – Biostatistics Research Unit, South African Medical Research Council, Durban, South Africa – School of Mathematics, Statistics and Actuarial Science, University of Kwa-Zulu Natal, Westville, South Africa
In longitudinal studies of biomarkers, an outcome of interest is the time at which a biomarker reaches a particular threshold. Due to the inherent variability of several studies have applied persistence criteria, designating the outcome as the time to the occurrence of two consecutive measurements less (or greater) than the threshold. In this presentation, we discuss a method to estimate the time to attainment of two consecutive measurements less than a meaningful threshold, which takes into account the patient-specific trajectory and measurement error. An expression for the expected time to threshold has been presented, which is a function of the fixed effects, random effects and residual variance. While the initial approach was motivated by continuous biomarkers, we present extensions of the methodology to accommodate censored observations as well as ordinal outcomes. We present a range of specific applications to HIV, cardiology, SARS-CoV2 and schizophrenia demonstrating the relevance of the methodology. Through these applications we demonstrate that the method proposed is computationally efficient, robust, and offers more flexibility than existing frameworks.
Conferência 2 – 30/05 (10:30 – 11:30) – Anfiteatro
Palestrante: Elisângela Aparecida da Silva Lizzi – Universidade Tecnólogica Federal do Paraná (UTFPR) – campus de Cornélio Procópio
O uso de modelos estatísticos desempenha um papel fundamental na vigilância em saúde e saúde pública, abrangendo uma ampla gama de aplicações em diversos campos. Desde estudos de epidemiologia de campo, pesquisa clínica, políticas públicas e até mesmo áreas emergentes como DNA e nanotecnologia como modelos que usam estatística associado a bioinformática, a pesquisa quantitativa desempenha um papel essencial na tomada de decisões. Durante a pandemia de COVID-19 em 2020, vimos a importância dos métodos quantitativos em saúde pública. Eles foram usados para rastrear a evolução dos casos, a taxa de transmissão, o número de óbitos e projetar novos casos, com o uso de gráficos, mapas e tabelas. Essas métricas se tornaram parte integrante dos noticiários, mas também precisam ser compreendidas pela população em geral para estabelecer protocolos de biossegurança em benefício de todos. Pode-se citar modelos de séries temporais e modelos espaço-temporais no contexto de política macro. No entanto, é crucial entender que os números por si só não fornecem respostas completas, eles precisam ser interpretados dentro do contexto em que estão inseridos. Devemos desmistificar o uso de métodos quantitativos e modelos estatísticos, entendendo suas implicações e limitações em saúde, ainda mais na era dos termos como Data Analitics, Big Data, Data Miner, Machine Learning e Inteligência Artificial e entre outros. Em alguns casos, interpretações errôneas podem levar a conclusões equivocadas, embasadas em números distorcidos, com isso a política implementada será obsoleta. O uso de estatística em saúde pública busca promover uma interação entre a pesquisa quantitativa, a prática clínica, epidemiologia, etiologia das doenças, conhecimento qualitativo, medicina baseada em evidências, além de contexto histórico e social. Ela visa transformar os valores numéricos em ações tangíveis, orientando a formulação de políticas e a tomada de decisões que afetam o cotidiano das pessoas. Portanto, a análise estatística desempenha um papel crucial na promoção do bem-estar social e na definição de diretrizes em situações críticas que afetam a saúde pública.
Conferência 3 – 30/05 (11:30 – 12:30) – Anfiteatro
Palestrante: Victor Hugo Lachos – Connecticut University, USA
HIV RNA viral load measures are often subjected to some upper or lower detection limit, depending on the quantification assays. Hence, the responses are either left- or right-censored. Censored mixed-effects models are routinely used to analyze this type of data and are based on normality assumptions for the random terms. However, those assumptions might not provide robust inference in the presence of atypical observations. In this work, we develop a Bayesian analysis of censored linear models replacing the Gaussian assumptions with the flexible class of scale mixture of normal (SMN) distributions while accounting for within-subject serial correlation through useful dependence structures and taking advantage of the No-U-Turn sampler (NUTS) to obtain posterior simulations. The SMN is an attractive class of symmetric heavy-tailed distributions that includes the normal distribution, the Student-t, slash, and the contaminated normal distributions as special cases. To illustrate the flexibility and applicability of the proposed model, an HIV AIDS study on viral loads dataset will be analyzed.
Conferência 4 - 31/05 (11:30 – 12:30) – Anfiteatro
Palestrante: Rafael Izbicki, Universidade Federal de São Carlos – UFSCar
While Null Hypothesis Significance Testing (NHST) remains a widely used statistical tool, it suffers from several shortcomings, such as conflating statistical and practical significance, sensitivity to sample size, and the inability to distinguish between accepting the null hypothesis and failing to reject it. Recent efforts have focused on developing alternatives to NHST to address these issues. Despite these efforts, conventional NHST remains dominant in scientific research due to its simplicity and perceived ease of interpretation. Our work presents a novel alternative to NHST that is just as accessible and intuitive: REACT. It not only tackles the shortcomings of NHST but also offers additional advantages over existing alternatives. For instance, REACT is easily applicable to multiparametric hypotheses and does not require stringent significance-level corrections when conducting multiple tests. We illustrate the practical utility of REACT through real-world data examples, using criteria aligned with common research practices to distinguish between the absence of evidence and evidence of absence.
Conferência 5 - Conferência de Encerramento - 31/05 (12:30 – 13:30) – Anfiteatro
Palestrantes: John Hinde, University of Galway, Ireland
Chris Brien, Australian Plant Phenomics Facility, University of Adelaide & UniSA STEM, University of South Australia
ESALQ is currently celebrating 60 years of postgraduate education in statistics that
started with the formation of the first Masters programme in agricultural statistics in
Brazil. In many other ways the Department and some key individuals have been very
influential in the development of biometry in Brasil. Of course, there is the teaching
and training of students from all over Brasil, but also the associated research and
scholarship reinforced by the international links that have been made over the years,
The Department has also been heavily involved in consultancy, within ESALQ and more
widely, and this joint applied work forms the backbone of the extensive collaborative
work that has been undertaken with colleagues across the world. From the IBCs in
Campinas, SP, in 1955, Guarujá, SP, in 1979, and Florianópolis, SC, in 2010 there have
been strong links with the International Biometric Society both in organization,
participation, and leadership. This talk will seek to highlight some of the key moments
and contributions of ESALQ mirroring the development of biometry over the last 60
years, with some speculation as to what the future may hold.
