Avaliação de estudos clínicos de tratamentos medicamentosos em pacientes acometidos por COVID-19 (SARS-CoV-2) : uma abordagem estatística crítica de ensaios clínicos fase III
Nenhuma Miniatura disponível
Data
2024-05-23
Autores
Santos, Rodolfo Rodrigo Pereira [UNIFESP]
Orientadores
Atallah, Álvaro Nagib [UNIFESP]
Tipo
Dissertação de mestrado
Título da Revista
ISSN da Revista
Título de Volume
Resumo
Contexto: A pandemia de COVID-19 motivou extensas pesquisas sobre tratamentos, incluindo ensaios clínicos com padrões flexíveis. Esta revisão avalia a qualidade dos ensaios clínicos para tratamentos e vacinas contra a COVID-19, com o objetivo de identificar lacunas e limitações. Objetivos: Analisar criticamente ensaios clínicos randomizados de Fase III para tratamentos de COVID-19 referentes a prevenção de infecções, agravamentos da doença e mortalidade. Métodos: Seguindo as diretrizes estabelecidas, foi realizada uma revisão dos ensaios de Fase III. Os estudos elegíveis envolveram adultos e foram pesquisados sem restrições de idioma ou data de publicação. Extração de dados, avaliação de risco de viés e avaliação da estrutura GRADE foram realizadas. As análises estatísticas incluíram poder de teste post-hoc e recálculos do tamanho da amostra. Resultados: A revisão incluiu 46 ensaios de Fase III com aproximadamente 64.000 participantes. Foram encontradas discrepâncias entre tamanhos amostrais empregados e tamanhos recalculados, destacando possíveis limitações. A avaliação do risco de viés identificou vieses significativos, particularmente no cegamento e no desempenho. A qualidade da evidência variou entre os tratamentos, principalmente com níveis moderados a baixos. Conclusão: As considerações metodológicas são cruciais em ensaios clínicos, enfatizando o planejamento cuidadoso, a definição de hipóteses e a determinação do tamanho da amostra. A comunicação transparente de metodologias estatísticas é essencial para a replicação dos resultados. Estudos piloto são propostos como alternativa ética. A ausência de análise de poder de teste e de justificativa para o tamanho da amostra levanta preocupações. A integração de análises estatísticas e perspectivas clínicas possibilita mais resultados confiáveis. O poder estatístico e os recálculos do tamanho da amostra são vitais para práticas baseadas em evidências durante emergências. A combinação do risco de viés e do GRADE facilitaram as avaliações críticas dos estudos encontrados.
Background: The COVID19 pandemic prompted extensive research on treatments, including clinical trials with flexible standards. This review evaluates the quality of clinical trials for COVID19 treatments and vaccines, aiming to identify gaps and limitations. Objective: To critically analyse Phase III randomized clinical trials for COVID19 treatments targeting infection prevention, disease severity, and mortality. Methods: Following established guidelines, a review was conducted on Phase III trials. Eligible studies involved adults and were searched without restrictions on language or publication date. Data extraction, risk of bias assessment, and GRADE framework evaluation were performed. Statistical analyses included post hoc power tests and sample size recalculations. Results: The review included 46 Phase III trials with approximately 64,000 participants. Discrepancies between sample sizes and recalculated sizes were found, highlighting potential limitations. Risk of bias assessment identified significant biases, particularly in blinding and performance. Evidence quality varied across treatments, mostly with moderate to low levels. Conclusion: Methodological considerations are crucial in clinical trials, emphasizing careful planning, hypothesis definition, and sample size determination. Transparent reporting of statistical methodologies is essential for result replication. Pilot studies are proposed as an ethical alternative. The absence of power analysis and sample size rationale raises concerns. Integration of statistical analysis and clinical perspectives ensures reliable findings. Statistical power and sample size recalculations are vital for evidencebased practices during emergencies. Combining risk of bias and GRADE was an alternative to bring more confidence focused on the critical issues in the studies found.
Background: The COVID19 pandemic prompted extensive research on treatments, including clinical trials with flexible standards. This review evaluates the quality of clinical trials for COVID19 treatments and vaccines, aiming to identify gaps and limitations. Objective: To critically analyse Phase III randomized clinical trials for COVID19 treatments targeting infection prevention, disease severity, and mortality. Methods: Following established guidelines, a review was conducted on Phase III trials. Eligible studies involved adults and were searched without restrictions on language or publication date. Data extraction, risk of bias assessment, and GRADE framework evaluation were performed. Statistical analyses included post hoc power tests and sample size recalculations. Results: The review included 46 Phase III trials with approximately 64,000 participants. Discrepancies between sample sizes and recalculated sizes were found, highlighting potential limitations. Risk of bias assessment identified significant biases, particularly in blinding and performance. Evidence quality varied across treatments, mostly with moderate to low levels. Conclusion: Methodological considerations are crucial in clinical trials, emphasizing careful planning, hypothesis definition, and sample size determination. Transparent reporting of statistical methodologies is essential for result replication. Pilot studies are proposed as an ethical alternative. The absence of power analysis and sample size rationale raises concerns. Integration of statistical analysis and clinical perspectives ensures reliable findings. Statistical power and sample size recalculations are vital for evidencebased practices during emergencies. Combining risk of bias and GRADE was an alternative to bring more confidence focused on the critical issues in the studies found.