Predictors of mortality in hospitalized patients with COVID-19

Aim. To determine clinical and laboratory factors associated with a severe course and lethality in hospitalized patients with COVID-19. Materials and methods. A retrospective comparative study included data of 745 adult patients hospitalized with COVID-19 from 16.05.2020 to 30.09.2020 (Tomsk, Russia). The intergroup comparison of indices, ROC analysis, and determination of odds ratio to assess the association between risk factors and the outcome were performed. Results. Age > 62 years, pneumonia within a year before COVID-19, and the presence of >= 3 comorbidities were associated with a fatal outcome (FO). Negative predictors of the outcome at the time of hospitalization included dyspnea, diastolic blood pressure <= 80 and pulse pressure > 48 mmHg, SpO(2) < 94% (and/or a decrease to <= 89% throughout hospitalization). Laboratory predictors of FO at admission were platelets <= 183 x 10(9) / l, neutrophils > 4.57 x 10(9) / l, lymphocytes <= 1.08 x 10(9) / l, neutrophil-to-lymphocyte ratio > 4.8, aspartate aminotransferase > 39 U / l, urea > 6.75 mmol / l, lactate dehydrogenase > 219 U / l, blood albumin <= 38 g / l, C-reactive protein (CRP) > 47 mg / l. When threshold values were reached during any of the hospitalization periods, FO was associated with CRP > 38 mg / l, ferritin > 648.6 mu g / l, D-dimer > 731.11 ng / ml, white blood cells > 14.27 x 109 / l, lymphocytes <= 0.73x 10(9) / l, duration of oxygen therapy > 3 days, need for non-invasive and invasive ventilation > 1 day, need for glucocorticoid administration > 1 day, reaching a total course dose > 6 mg for dexamethasone. Conclusion. The factors associated with FO in hospitalized patients with COVID-19 were identified.