A evolução do eletrocardiograma na doença de Chagas: usos no diagnóstico, prognóstico e no seguimento de idosos

• Ano de defesa: 2022
• Autor(a) principal: Bruno Oliveira de Figueiredo Brito
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Minas Gerais; Brasil; MEDICINA - FACULDADE DE MEDICINA; Programa de Pós-Graduação em Ciências da Saúde - Infectologia e Medicina Tropical; UFMG
• Palavras-chave em Português: doença de Chagas; Idoso; Eletrocardiografia; Inteligência artificial; Insuficiência cardíaca; Doença de Chagas; Inteligência Artificial; Insuficiência Cardíaca; Dissertação Acadêmica
• Link de acesso: http://hdl.handle.net/1843/51544

Introduction: The natural history of Chagas disease (ChD) in the elderly population is unknown, and it is controversial whether the disease continues to progress in old age. When it progresses to the cardiac form, the heart failure is one of the leading causes of death. An artificial intelligence (AI) algorithm has shown excellent accuracy for detecting left ventricular systolic dysfunction (LVSD) using the electrocardiogram (ECG), but it has not been evaluated in ChD. Objective 1: To investigate the evolution of ECG changes in elderly people with chronic ChD compared to non-infected elderly (NChD) and how this affects the survival of the population of the elderly cohort of Bambuí in a 14-year follow-up. Methods 1: A 12-lead ECG of each subject was obtained in 1997, 2002, and 2008, and abnormalities were classified by the Minnesota Code. The influence of ChD on the ECG evolution was evaluated through semi-competitive risks. A survival analysis was performed from a 5.5-year Landmark; individuals of the ChD and NChD groups were compared separately for the development of major ECG abnormalities between 1997 and 2002. Objective 2: To analyze the AI-ECG's ability to recognize LVSD in patients with ChD from the SaMi-Trop cohort, defined as left ventricular ejection fraction determined by Echocardiogram ≤ 40%. Methods 2: Cross-sectional study of ECG obtained from the cohort of patients with ChD named SaMi-Trop. The digital ECGs of the participants were submitted to the analysis of the trained machine to detect LVSD. The diagnostic performance of the AI- ECG to detect LVSD was evaluated, and the echocardiogram was the gold standard. The model was enriched with NT-proBNP plasma levels, male sex, and QRS ≥ 120ms. Results 1: Among the 1,462 participants in the Bambuí Elderly Cohort, 557 had CDh (median age: 68 years for ChD and 67 years for NChD). ChD increases the risk of developing a new ECG abnormality when compared to NChD [HR: 2.89 (95% CI 2.28 – 3.67)]. Developing a new ECG abnormality in ChD increases the risk of death compared to those who maintain a normal ECG [HR: 1.93 (95% CI 1.02 – 3.65)]. Results 2: Among the 1,304 participants in the SaMi-Trop study, 7.1% of subjects have LVSD and 59.5% have major ECG abnormalities. The AI algorithm identified LVSD with OR= 63.3 (95% CI 32.3-128.9), sensitivity of 73%, specificity of 83%, accuracy of 83% and a negative predictive value of 97%; the AUC was 0.839. The model adjusted for the male sex and QRS ≥ 120ms improved the AUC to 0.859. The model adjusted for the male sex and elevated NT-proBNP had a higher accuracy of 0.89 and an AUC of 0.874. Conclusions: ChD is associated with a higher risk of progression to cardiomyopathy in the elderly. The occurrence of a new abnormality on the ECG increases the risk of death. AI - ECG of patients with ChD can be turned into a powerful tool for the recognition of LVSD, thus, contributing to the treatment with low-cost drugs that can improve symptoms and reduce mortality.