Researchers develop a model that predicts the COVID-19 clinical evolution at the hospital admission point

  • A joint project from the UPC, the UB, the HUB, and IDIBELL studied the probability of suffering severe pneumonia, needing ventilation, or even dying during hospital admission due to COVID-19. As a result of the statistical models created, based on data from more than 5,000 patients admitted to five hospitals in different pandemic waves, they developed a web application for the early identification of potentially high-risk COVID-19 patients and the prediction of the disease course.
NO120 - App COVID19 Imatge noti

Predicting high-risk cases among COVID-19 patients when they are admitted to the hospital is key to providing them with adequate treatment and to optimizing healthcare resources, especially at times of high disease incidence and risk of hospital collapse. Therefore, it is necessary to have rigorous anticipation mechanisms based on complete and accurate data.

With this idea, the DIVINE project, an acronym for “DynamIc eValuation of COVID-19 cliNical statEs and their prognostic factors to improve intra-hospital patient management”, has been developed, financed by the Generalitat de Catalunya, within the framework of the Pandemics 2020 call, and that has been carried out over the last year and a half. The project is the result of the collaboration of the Biostatistics and Bioinformatics Research Group (GRBIO) −formed by researchers from the Polytechnic University of Catalonia – BarcelonaTech (UPC) and the University of Barcelona (UB)−, from the Bellvitge University Hospital ( HUB) and the Biostatistics Unit of the Bellvitge Biomedical Research Institute (IDIBELL), under the coordination of Guadalupe Gómez Melis, a GRBIO researcher, and professor in the Department of Statistics and Operations Research at the UPC.


Predictive models of patient evolution

As a result of the project, the MSMpred application has been developed. An online clinical prediction tool presents the probability of suffering severe pneumonia, needing mechanical ventilation, or even dying, as well as the transition time between these states.

To create this tool, the researchers have developed a multi-state statistical model, which analyzes the disease stages in each patient. Data from more than 5,000 COVID-19 patients admitted during the first five waves of the pandemic have been analyzed and combined. Specifically, sociodemographic data, known medical history and comorbidities, clinical symptoms, and detailed information on clinical management have been taken into account.

These data have been collected and coded by specialist doctors from five hospitals in the Southern Metropolitan Area: the Bellvitge University Hospital, the Moisès Broggi Hospital, the Viladecans Hospital, the Sant Boi de Llobregat Hospital, and the Garraf Hospital Consortium.


Therapeutic ceiling

A key factor in predicting the evolution of the patient has been the application or not of a therapeutic ceiling, that is, predetermining the highest level of intervention. At the height of the pandemic, decisions on the therapeutic ceiling had to be adapted to an emergency situation. This study is a pioneer in describing the clinical characteristics, mortality, and complications of patients with COVID-19 based on the therapeutic ceiling. The results of this work are collected in the scientific article “Characteristics and outcomes by ceiling of care of subjects hospitalized with COVID-19 during four waves of the pandemic in a metropolitan area: a multi-center cohort study”, published in the journal Infectious Diseases and Therapy for this month of December.


Incubation time and evolution of the patient

In addition to identifying the most clinically relevant factors that influence a better or worse evolution in a patient hospitalized for COVID-19, the DIVINE project has made it possible to estimate the incubation time of the disease, from infection to the onset of symptoms, in different phases of the pandemic. The results of this work are collected in the article “SARS-Cov-2 incubation period according to vaccination status during the fifth COVID-19 wave in a tertiary-care center in Spain: a cohort study”, published in the journal BMC Infectious Diseases last November. In that phase of the project, the researchers developed a statistical framework, which would be applicable to future pandemics arising from other infectious diseases.

On the other hand, the data collected in the DIVINE project have also made it possible to assess how the profile of patients hospitalized for COVID-19 has evolved during the different waves of the pandemic, based on different indicators. It has been observed that as the pandemic progressed, hospitalized patients tended to be younger and with fewer comorbidities and, consequently, had a better prognosis.

In addition to Guadalupe Gómez Melis, principal investigator, GRBIO researchers Mireia Besalú, from the University of Barcelona, and UPC researchers Erik Cobo, Jordi Cortés, Daniel Fernández, Leire Garmendia, Klaus L. angohr, Nuria Pérez, and Xavier Piulachs. On behalf of IDIBELL and the HUB, the project has been led by the researcher Cristian Tebé, head of the Biostatistics Unit of this institute, and has had the participation of doctors and nurses from the HUB’s Infectious Diseases Service and researchers Jordi Carratalà, Carlota Gudiol, Gabriela Abelenda, Alexander Rombauts, Natàlia Pallarès, Gemma Molist, Pilar Hereu, and Sebastià Videla.

Scroll to Top