Status and trends in research on clinical prediction models for severity risk stratification in confirmed Covid-19 patients

• Autor(a) principal: Monzón-Peréz, Maicel
• Data de Publicação: 2023
• Outros Autores: Sanchez-Valdés, Lizet, Lage-Dávila , Agustín
• Tipo de documento: preprint
• Idioma: spa
• Título da fonte: SciELO Preprints
• Texto Completo: https://preprints.scielo.org/index.php/scielo/preprint/view/5795
• topic: prognosis; triage; theoretical models; covid-19; sars cov-2; pandemic; pronóstico; triaje; modelos teóricos; covid-19; sars cov-2; pandemia

Introduction: Previous knowledge in the scientific literature on clinical prediction models in patients with Covid-19 may be useful for the development of new research. Objective: We describe the sources, authors, documents and key issues that are part of the research front. Identify which models, outcome variables, predictors and algorithms have been relevant. We Identify to what extent the available models could meet the quality attributes and what characteristics they must have to be applicablein the Cuban context. Methods: A review and scientometric analysis was carried out on the research in development and validation of clinical predictive models for Covid-19. The scientometric indicators were used and a thematic map was made for the analysisof the conceptual structure of the subject.Results: The subject was of great interest with papers published in the highest level journals. It is possible to distinguish a context of low and high risk application according to the primary and secondary health levels. The systematic reviewpublished by Wynants et al. was the publication with the greatest impact and an important source for the identification of models, main components, as well as possible causes of bias.Conclusions: The literature recognizes that most of the published models are not recommended for general use in clinical practice, so it is an open research front. However, the data obtained could be useful for the development and validation ofCuban models.