Analyzing the Diagnostic Value of Laboratory Parameters with ROV Curve in Covid-19
Abstract
Objective: Since the emergence of Coronavirus Disease 2019 (Covid-19) in December 2019, it has turned into a pandemic. Covid-19 infection; can trigger pulmonary inflammation and also cause acute lung injury. In cases with severe pneumonia, long hospital stays and high mortality rates were detected. Therefore, there is a need for biomarkers to diagnose cases quickly and early. In our study, we aimed to evaluate the laboratory parameters and radiological findings used in Covid-19 patients with ROC curve analysis.
Material and Method: Our study was planned retrospectively in adult patients diagnosed with Covid- 19 in our hospital between May 2020 and July 2020. Data were obtained by scanning the results of routine biochemical and hematological laboratory tests of the patients. Laboratory data and radiological findings were compared using receiver operating characteristic ROC curve and Cox regression analysis.
Results: One thousand cases (54% males) with a mean age of 47.3 ± 17.65 (range: 19-78) years were analyzed. RT-PCR result for Covid-19 was positive in all cases. Significantly, in patients with advanced pulmonary involvement urea (p < 0.0001), glomerular filtration rate (p < 0.0001), lactate dehydrogenase (LDH) (p < 0.0001), aspartate aminotransferase (AST) (p < 0.0001), Neutrophil-Lymphocyte Ratio (p < 0.0001), Platelet-Lymphocyte Ratio and C-reactive protein (CRP) (p < 0.0001) were significantly different. In addition, in patients with advanced lung involvement, hemoglobin (Hb) (p < 0.0001), white blood cell count (WBC) (p < 0.0001), serum albumin (Alb) level (p < 0.0001) and lymphocyte count (p) < 0.0001) was lower, compared to others. WBC and CRP were significant in the multivariate model, p = 0.002 and p < 0.0001, respectively. Alb (AUC = 0.77), D-dimer (AUC = 0.68), LDH (AUC = 0.69) and CRP (AUC = 0.77) had very good accuracy in predicting cases with advanced lung involvement for Covid-19.
Conclusions: In order to make the diagnosis in Covid patients; indicators are needed to classify patients and detect disease progression. Our findings show that specifically Alb, D-dimer, LDH and CRP levels can be used to predict the severity of lung involvement in Covid-19 patients.