Objective: To describe the methodological characteristics and good research practices of the intervention studies for COVID-19 developed in Brazil in the first months of the pandemic. Method: A review in the CONEP-COVID bulletin (05/28/2020) and the International Clinical Trials Registry Platform, ClinicalTrials.gov, ReBEC was conducted to identify drug-type, biological therapy or vaccine intervention studies registered in Brazil. The studies were evaluated for methodological characteristics and power for different magnitudes of effect. Results: 62 studies were included, 55 identified on the CONEP website and seven on a registration database. Several interventions are being tested: chloroquine/hydroxychloroquine, azithromycin, plasma convalescent, tocilizumab, sarilumab, eculizumab, vaccine, corticoids, anticoagulants, n-acetylcysteine, nitazoxanide, ivermectin, lopinavir/ritonavir, etc. By May 2020, 22 research protocols were published in a protocol registry database, 82% were randomized clinical trials and 59% had adequate control group. However, 59% were not masked and only 24% included patients with positive test with diagnostic accuracy. Most of the studies would have power >80% just to identify large effect sizes. In a prospective follow-up, until July 21st/2020, 60% of the studies available at CONEP were not registered in the ICTRP/ReBEC/ClinicalTrials platforms. Conclusion: The interventions evaluated during the Brazilian research response reflect international initiatives, but with a different distribution, a larger proportion of studies asseessed hydroxychloroquine/chloroquine. Limitations in methodological design and sample planning represent challenges that could affect the research outreach.
Objetivo: Descrever as características metodológicas e de boas práticas em pesquisa dos estudos de intervenção para COVID-19 desenvolvidos no Brasil nos primeiros meses da pandemia. Método: Revisamos o boletim da CONEP-COVID (28/05/2020) e as bases International Clinical Trials Registry Platform, ClinicalTrials.gov e ReBEC para identificar estudos registrados no Brasil, avaliando intervenções de tipo medicamento, terapia biológica ou vacinas. Descrevemos as características metodológicas e calculamos o poder para diferentes magnitudes de efeito. Resultados: Foram incluídos 62 estudos, 55 identificados no site da CONEP e mais sete nas bases de registro. As intervenções medicamentosas mais frequentemente testadas nesses estudos foram: cloroquina/hidroxicloroquina, azitromicina, plasma convalescente, tocilizumabe, sarilumabe, eculizumabe, vacina, corticoides, anticoagulantes, n-acetilcisteína, nitazoxanida, ivermectina e lopinavir/ritonavir. De 22 protocolos, publicados até maio de 2020 nas bases de registro, 18 (82%) eram ensaios clínicos randomizados e 13 (59%) tinham grupo controle adequado. Entretanto, nove (41%) eram mascarados e somente cinco (24%) incluía pacientes diagnosticados com teste de laboratório específico (e.g. RT-PCR). A maioria desses trabalhos teria poder >80% apenas para identificar grandes tamanhos de efeito. Em seguimento prospectivo, observamos que 60% dos estudos, disponíveis na CONEP até maio de 2020, não estava em nenhuma das plataformas de registro (ICTRP/ReBEC/ClinicalTrials) até o dia 21/07/2020.Conclusão: As intervenções avaliadas durante a resposta brasileira em pesquisa refletem iniciativas internacionais, porém com uma distribuição diferente, tendo um número elevado de estudos avaliando hidroxicloroquina/ cloroquina. Limitações no delineamento metodológico e planejamento amostral representam desafios que poderiam afetar o alcance dos trabalhos.