The role of hemogram parameters and C-reactive protein in predicting mortality in COVID-19 infection

This study aimed to investigate hemogram parameters and C-reactive protein (CRP) that can be used in clinical practice to predict mortality in hospitalized patients with a diagnosis of COVID-19.

This cohort study was conducted at University Hospital, which is a designated hospital for COVID-19 patients. Adult patients who were admitted to our hospital emergency department with suspected COVID-19 and who were hospitalized in our institution with a COVID-19 diagnosis were analysed.

There were 148 patients hospitalized with COVID-19. All-cause mortality of follow-up was 12.8%. There were statistically significant results between the two groups (survivors and nonsurvivors), which were classified based on hospital mortality rates, in terms of the lymphocyte to C-reactive protein ratio (LCRP), systemic immune inflammation index (SII), neutrophil-to-lymphocyte ratio (NLR), platelet-to-lymphocyte ratio (PLR), CRP concentration and comorbid disease. In a receiver operating characteristic (ROC), curve analysis, LCRP, NLR, PLR and SII area under the curve (AUC) for in-hospital mortality were 0.817, 0.816, 0.733 and 0.742, respectively. Based on an LCRP value of 1 for in-hospital mortality, the sensitivity and specificity rates were 100% and 86.8%, respectively. Based on the average SII of 2699 for in-hospital mortality, the sensitivity, specificity and accuracy rates were 68.4%, 77.5% and 76.3%, respectively. A total of 19 patients died during hospitalization. All of these patients had an LCRP level ≤ 1; 14 had an NLR level ≤ 10.8; 13 had an SII ≥ 2699 (Fisher's exact test, P = .000). Independent predictors of in-hospital mortality rates were LCRP < 1, PLR, SII ≥ 2699, white blood cell count, CRP, age, comorbidities, and ICU stay.

We concluded that inflammatory parameters, such as LRCP, SII and NLR, were associated with disease severity and could be used as potentially important risk factors for COVID-19 progression.