Abstract
BACKGROUND: Inflammation plays a major role in coronavirus disease (COVID-19). Factors that convey information about the status of inflammation could predict disease severity and help identify patients prone to clinical deterioration. Here, we aimed to evaluate the predictive value of inflammatory markers on the extent of lung involvement and survival of patients with COVID-19. MATERIALS AND METHODS: Eighty patients with confirmed COVID-19 were enrolled. Demographic, clinical, and laboratory data were collected at admission. All patients underwent chest computed tomography (CT); the extent of lung involvement was assessed by a scoring system. Patients were followed up until death or discharge occurred. Logistic regression analysis was performed to evaluate the association of investigated variables with COVID-19-related death. The association between different variables and CT score was assessed using linear regression model. Receiver operator characteristic curve analysis was applied to identify the predictive value of inflammatory markers and CT score on survival. RESULTS: The mean age of patients was 54.2 ± 15.2 years; 65% were male. Increased neutrophil-to-lymphocyte ratio (β =0.69, odds ratio [OR] =1.50), platelet-to-lymphocyte ratio (β =0.019, OR = 1.01), and decreased lymphocyte to C-reactive protein ratio (LCR) (β = -0.35, OR = 0.62) were significantly associated with a higher CT score and increased odds of death (P < 0.05). Lactate dehydrogenase level was also positively related with extensive lung involvement and death (β =1.15, OR = 1.52, P < 0.05). The LCR threshold for identifying survivors from nonsurvivors was 0.53 (area under curve [AUC] =0.82, 78% sensitivity and 74% specificity). Lung involvement ≥50% on chest CT was an excellent predictor of death (AUC = 0.83, 81% sensitivity and 79% specificity). CONCLUSION: Daily-performed laboratory tests that represent inflammation have great value for predicting the amount of disease burden and risk of mortality. Moreover, their cost-effectiveness and feasibility turn them into ideal prognostic markers.