Covid-19 Outcome Prediction Model by Using Radio-Diagnostic Methods

Abstract

Research goal: The goal of this research is to create a prediction model for a possible outcome (death or survival) of COVID-19, which model incorporates the easily available radio diagnostic methods such as classical radiology and the disease scoring system. Part of the goal of this study is to define the chances and probabilities of occurrence of death as a result of the primary disease and to identify the risk factors that have the highest influence on the final outcome of COVID-19. Methodology: The methodological approach used in this study is the binary logistic regression which is part of the group of generalized linear statistical methods. Results: Results show that patients with complications and comorbidities have the highest chances of death from COVID-19 (OR 16,53 with CI 8,21 - 33,25 and 4,08 with CI 1,34 - 12,38). Men are also exposed to higher but insignificant mortality risk with OR 1,55 with CI 0,86 - 2,80. Every additional year of age increases the mortality risks by 1,06 times (CI 1,03 - 1,09), while every additional score of the primary disease leads to increased chances of unwanted outcome by 1,24 (CI 1,04 - 1,47). Conclusion: The mortality outcome of COVID-19 is not an exclusive consequence of the primary disease but it is usually determined in correlation with different comorbidities and existing complications as well as other standard influencing factors such as age and gender. Contribution and significance of the research: The primary importance of this research is the fact that it allows for an improved precision and upgrade to the basic model of standard factors by using new predictors, specifically secondary complications from the radio-graphic picture and scoring of the primary disease, which leads to higher utilization of cheap and easily available radio-diagnostic methods.