Association of Hematologic biomarkers and their combinations with disease severity and mortality in COVID-19- an Indian perspective

Am J Blood Res. 2021 Apr 15;11(2):180-190. eCollection 2021.

Abstract

Background: COVID-19 is a systemic viral infection with a significant impact on the hematopoietic system, hemostasis as well as immune system. It would be of utmost importance to explore if the most routinely used tests could serve as an aid in determining patient's clinical status or predicting severity of the disease.

Methods: A prospective cross-sectional study was conducted on 506 Covid-19 positive patients and 200 controls over a period of two months (June and July 2020). The cases were sub-classified based on disease severity into mild to moderate (n=337), severe (n=118) and very severe (n=51) and based on survivor status into survivors (n=473) and non-survivors (n=33).

Results: There were statistically significant differences in WBC count, Absolute neutrophil count (ANC), Absolute lymphocyte count (ALC), absolute monocyte count (AMC), neutrophil-lymphocyte ratio (NLR), platelet-lymphocyte ratio (PLR), monocyte-lymphocyte ratio (MLR) Red blood cell distribution width (RDW-SD) and RDW CV between covid cases vs controls; among the clinical subgroups and among the survivors and non-survivors. There was a significant strong positive correlation between various parameters, that is, NLR and MLR (r: 0.852, P=0), MPV and PDW (r: 0.912, P=0), MPV and PLCR (r: 0.956, P=0), PDW and PLCR (r: 0.893, P=0). NLR (AUC: 0.676, P=0) was the best single parameter and NLR+RDW-CV was best combination parameter as per area under curve (0.871) of ROC to distinguish severe from mild to moderate disease.

Conclusions: Leucocytosis, neutrophilia, lymphopenia and monocytosis were characteristic findings in covid cases while NLR and NLR+RDW-CV emerged as the most effective single and combination CBC parameters in distinguishing mild to moderate and severe cases respectively.

Keywords: COVID-19; MLR; NLR; PLR; RDW-CV; RDW-SD; hematological indices.