Abstract
Objectives: The clinical course of COVID-19 ranges from mild to severe. The predictability of clinical outcomes gains importance in managing the disease. In this retrospective cohort study, we aimed to investigate the relationship between biomarker levels and the clinical severity of COVID-19.
Methods: COVID-19 patients (n=618) admitted to a tertiary care hospital in Istanbul, Turkey were classified according to their clinical status using a scoring system designed by WHO. Laboratory parameters such as D-dimer, ferritin, and lymphocyte count levels were evaluated. In order to find out the relation between laboratory biomarkers and the severity of COVID-19, univariable and multivariable logistic regression analyses were used.
Results: A positive correlation was found when WHO Score was compared with D-dimer levels (r=.508, p<0.01), and ferritin levels (r=.391, p<0.01), whereas a negative correlation was observed between WHO Score and lymphocyte count levels (r=-.381, p<0.01). The cut-off values for D-dimer and ferritin were found as 0.86 ng/mL (70.9% sensitivity and 87.1% specificity) and 92.74 ng/mL (78.5% sensitivity and 52.2% specificity) respectively. According to the multivariable logistic regression model, lymphocyte count (β=- 0.305, p<0.001) and D-dimer levels (β=1.326, p<0.001) are the statistically significant regressors for hospitalization need.
Conclusion: Patients with higher D-dimer and ferritin levels were likely to have more severe disease, whereas patients with higher lymphocyte counts overcame the disease mildly. Unlike other studies, evaluating D-dimer and lymphocyte count will likely give more detailed information on COVID-19 clinical outcomes. J Microbiol Infect Dis 2022; 12(3):89-96.

Key words: COVID-19, ferritin; D-dimer, absolute lymphocyte count, multivariable logistic regression analysis