Based on review of the literature,14 , 27 , 28 several covariates were selected for analysis as potential confounding variables in our regression analyses. These included age, sex, comorbidities, disease severity, and lymphocyte count.