Prediction Models for COVID-19 Mortality Using Artificial Intelligence

Abstract

The coronavirus disease 2019 (COVID-19) pandemic has placed a great burden on healthcare systems worldwide. COVID-19 clinical prediction models are needed to relieve the burden of the pandemic on healthcare systems. In the absence of COVID-19 clinical prediction models, physicians' practices must depend on similar clinical cases or shared experiences of best practices. However, if accurate prediction models that combine parameters are introduced, they could provide the estimated risk of infection or experiencing a poor outcome following infection. The use of prediction models could assist medical staff in assigning patients when allocating limited healthcare resources and may enhance the prognosis of patients with COVID-19. Recently, several systematic reviews for COVID-19 have been published, some of which focus on prediction models that use artificial intelligence. We summarize the important messages of a systematic review titled "COVID Mortality Prediction with Machine Learning Methods: A Systematic Review and Critical Appraisal," published in this Special Issue.

Keywords: COVID-19; artificial intelligence; machine learning; mortality; prediction models; systematic review.