ABSTRACT
This paper deals with the problem of modeling counterfactual reasoning in scenarios where, apart from the observed endogenous variables, we have a latent variable that affects the outcomes and, consequently, the results of counterfactuals queries. This is a common setup in healthcare problems, including mental health. We propose a new framework where the aforementioned problem is modeled as a multivariate regression and the counterfactual model accounts for both observed and a latent variable, where the latter represents what we call the patient individuality factor ( φ ). In mental health, focusing on individuals is paramount, as past experiences can change how people see or deal with situations, but individuality cannot be directly measured. To the best of our knowledge, this is the first counterfactual approach that considers both observational and latent variables to provide deterministic answers to counterfactual queries, such as: what if I change the social support of a patient, to what extent can I change his/her anxiety? The framework combines concepts from deep representation learning and causal inference to infer the value of φ and capture both non-linear and multiplicative effects of causal variables. Experiments are performed with both synthetic and real-world datasets, where we predict how changes in people's actions may lead to different outcomes in terms of symptoms of mental illness and quality of life. Results show the model learns the individually factor with errors lower than 0.05 and answers counterfactual queries that are supported by the medical literature. The model has the potential to recommend small changes in people's lives that may completely change their relationship with mental illness.
ABSTRACT
Resumo: Introdução: Durante a última década, o modelo tradicional de ensino médico foi amplamente debatido, e, há pouco tempo, a pandemia da Covid-19 impôs novamente mudanças, iniciadas em caráter de urgência em todo o mundo, trazendo novos desafios à formação médica no que concerne ao ensino e à aprendizagem. A estratégia de ensino conhecida como Worked Example é uma ferramenta instrucional na qual um especialista mostra a solução de um determinado problema para um aprendiz de forma pormenorizada. Objetivo: Este estudo teve como objetivo comparar as aulas expositivas em vídeo com a técnica Worked Example para ensino de tomografia de tórax. Método: Trata-se de um estudo experimental realizado, por meio de uma intervenção educacional, com alunos do curso de Medicina em fases iniciais e finais da prática clínica. Resultado: A análise de variância de medidas repetidas (ANOVA) foi usada na análise estatística. Houve diferença significativa entre as notas antes e depois do treinamento (F1; 74 = 46,008; p < 0,001) e entre as fases do curso (F2; 148 = 19,452; p < 0,001). Não houve diferença estatisticamente significativa entre os grupos (F2; 74 = 1,401; p = 0,240). Não houve diferença significativa no esforço mental referido na comparação entre grupos (F1; 69 = 0,092; p = 0,762), porém os alunos do segundo ano apresentaram um escore de esforço significativamente maior. Conclusão: O Worked Example, uma técnica com boa aplicabilidade para estudantes da graduação e adequada para formatos digitais, mostrou-se igualmente eficaz a aula expositiva, técnica consagrada no ensino de radiologia.
Abstract: Introduction: During the last decade, the traditional model of medical education was widely debated, and recently, the Covid-19 pandemic added new challenges in teaching and learning in medical education. The Worked Example is an instructional tool in which an expert shows a detailed problem solution for a learner. Objective: This study aimed to compare, in an experimental study with medical students, lectures in online format with Worked Example for teaching chest tomography. Method: Experimental study through an educational intervention with medical students in the initial and final stages of clinical practice. Result: Analysis of variance of repeated measures (ANOVA) was used in the statistical analysis. There was a significant difference between the grades before and after training (F1; 74 = 46.008; p < 0.001) and between the course phases studied (F2; 148 = 19.452; p < 0.001). There was no statistically significant difference between the groups (F2; 74 = 1.401; p = 0.240). There was no significant difference in mental effort reported in the comparison between groups (F1; 69 = 0.092; p = 0.762), but students in the 2th year had a significantly higher effort score. Conclusion: Worked Example, which is a good technique for students and suitable for digital formats, was equally effective as lecture, a well-known instructional method for teaching radiology.
ABSTRACT
INTRODUÇÃO: A depressão pós-parto (DPP) tem sido alvo de inúmeras investigações científicas devido à sua alta prevalência e gravidade e também por suas repercussões negativas na puérpera e na criança. OBJETIVO:Comparar mulheres com e sem DPP em um grupo de puérperas selecionadas aleatoriamente a partir dos partos ocorridos em uma maternidade de Belo Horizonte (MG). MÉTODOS: Selecionamos 245 mulheres que tiveram parto em uma maternidade de Belo Horizonte. Aplicamos uma entrevista semiestruturada para a obtenção de dados psicossociais e demográficos e uma entrevista estruturada (Mini Neuropsychiatric Interview, MINI-Plus) para o diagnóstico de depressão maior segundo critérios do Diagnostic and Statistical Manual of Mental Disorders, 4th edition (DSM-IV). RESULTADOS:26,9 por cento das mulheres tinham diagnóstico de DPP. Não encontramos diferenças entre as características sociodemográficas das mulheres com (n = 66, 26,9 por cento) e sem diagnóstico de DPP (n = 179, 73,1 por cento). Entretanto, diversas variáveis clínicas e psicossociais se mostraram significativamente diferentes entre os dois grupos, tais como história de depressão, vivência de estresse ou presença de sintomas depressivos ou ansiosos durante a gravidez, ocorrência de complicações maternas ou na criança no pós-parto e insuficiência de suporte nos cuidados pós-natais. CONCLUSÃO: A identificação de fatores associados à DPP é importante para a compreensão de sua etiopatogenia e para o estabelecimento de estratégias de prevenção e tratamento precoce dessa grave doença.
INTRODUCTION: Postpartum depression has been the object of extensive research both because of its high prevalence and its repercussions on the mother and the infant. OBJECTIVE:To compare women with and without depression in a sample of postpartum women randomly selected at a maternity in the city of Belo Horizonte, Brazil. METHODS: A total of 245 women who gave birth at a private maternity hospital in Belo Horizonte were selected. A semi-structured interview was used for the collection of psychosocial and demographic data. Diagnosis of major depression was established using and a structured interview (Mini Neuropsychiatric Interview, MINI-Plus), based on criteria set forth in the Diagnostic and Statistical Manual of Mental Disorders, 4th edition (DSM-IV). RESULTS: Of the total sample, 26.9 percent had a diagnosis of postpartum depression. No differences were observed in the social and demographic characteristics of depressed women (n = 66, 26,9 percent) and of those without depression (n = 179, 73,1 percent). Nevertheless, several clinical and psychosocial variables showed significant differences between the groups and were therefore associated with postpartum depression, namely previous history of depression, presence of stress or depressive/anxiety symptoms during pregnancy, postpartum complications affecting the mother or the infant, and lack of support and care in the postpartum period. CONCLUSION: The identification of factors associated with postpartum depression is important for an improved understanding of the pathophysiology of this disease and for the establishment of strategies aimed at prevention and early diagnosis.
