Predictive accuracy of blood inflammatory markers on COVID-19 mortality

Abstract

Background: The impact of COVID-19 may be more severe in developing countries. Our study aims to analyze the accuracy of several inflammatory biomarkers in predicting COVID-19 mortality, providing information about the most suitable markers for developing countries.
 Methods: A retrospective cohort study was conducted at Dr. Soetomo General Hospital, Indonesia, from March to June 2020. White Blood Cells (WBC) count, Neutrophil-Lymphocyte Ratio (NLR), Procalcitonin (PCT), D-Dimer, and C-Reactive Protein (CRP) have been collected from the electronic medical records. We performed survival analysis to provide the hazard ratio and Receiver Operating Characteristic (ROC) curve analysis to test for accuracy for each parameter.
 Results: A total of 423 patients who met the criteria for participating had a median age of 54 (IQR 45-61) years. Patients in the death group are characterized by older age and shorter length of hospitalization. The WBC, NLR, PCT, D-Dimer, and CRP are found significantly higher in the death group (P=0.000). The WBC, NLR, PCT, D-Dimer, and CRP have an Area Under the Curve (AUC) of 0.709, 0.773, 0.738, 0.721, and 0.769, respectively moderate accuracy in predicting COVID-19 patient mortality. We found that NLR is significantly more accurate than the age parameter (Z=3.527; P=0.000) but has equal accuracy with other laboratory parameters. 
 Conclusions: Since NLR obtained the highest accuracy, we still recommend routine complete blood count tests as prognostic biomarkers with the highest feasibility to be performed in developing countries.