RESULTS
We evaluated multiple approaches to develop mathematical models using
parameters that can predict the progression of disease. Candidate
parameters were selected from mechanistic understanding of the process
of pathogenesis of COVID-19 to evaluate their possible impact on the
clinical outcome. Regression models utilize data to build predictive
models. Hypotheses are examined and confirmed with pre-determined
statistical confidence and inferential power. These models incorporate
all the experimental variability in the data set. Since the models
contained numeric factors and numeric ordinal outcomes, we utilized
methods of Multiple Linear Regression [41]. In this approach, we
used the simulated data set from COVID-19 affected subjects, organized,
and analyzed it to understand the variability of each of the parameters.