COVID-19 mortality prediction model, 3C-M, built for use in resource limited settings - understanding the relevance of neutrophilic leukocytosis in predicting disease severity and mortality

In this study, a combination of clinical and hematological information, collected on day of presentation to the hospital with pneumonia, was evaluated for its ability to predict severity and mortality outcomes in COVID-19. Ours is a retrospective, observational study of 203 hospitalized COVID-19 patients. All of them were confirmed RT-PCR positive cases. We used simple hematological parameters (total leukocyte count, absolute neutrophil count, absolute lymphocyte count, neutrophil to lymphocyte ration and platelet to lymphocyte ratio); and a severity classification of pneumonia (mild, moderate and severe) based on a single clinical parameter, the percentage saturation of oxygen at room air, to predict the outcome in these cases. The results show that a high absolute neutrophil count on day of onset of pneumonia symptoms correlated strongly with both severity and survival in COVID-19. In addition, it was the primary driver of an initial high neutrophil-to-lymphocyte ratio (NLR) observed in patients with severe disease. The effect of low lymphocyte count was not found to be very significant in our cohort. Multivariate logistic regression was done using Python 3.7 to assess whether these parameters can adequately predict survival. We found that clinical severity and a high neutrophil count on day of presentation of pneumonia symptoms could predict the outcome with 86% precision. This model is undergoing further evaluation at our centre for validation using data collected during the second wave of COVID-19. We present the relevance of an elevated neutrophil count in COVID-19 pneumonia and review the advances in research which focus on neutrophils as an important effector cell of COVID-19 inflammation.