Factors associated with mortality in patients hospitalized for COVID-19 in Spain. Data from the RERFAR registry.

Objective: To determine the baseline characteristics associated with higher  mortality at 42 days in patients hospitalized for COVID-19 in Spain.

Method: The study analyzed a prospective cohort of hospitalized COVID-19  patients. The dependent variable was 42-day mortality. Data on the subjects'  demographic and clinical characteristics, comorbidities, usual therapy and  supportive interventions and treatments was collected within 48 hours from  admission. To determine the potential association of the data with mortality, a  multivariate analysis was performed using logistic regression.

Results: 15,628 patients were included, 18.2% of whom (n = 2,806) died  during the study period. According to the multivariate analysis, the variables  that were significantly associated (p < 0.05) with mortality upon admission  were: being referred from a nursing home (OR 1.9); having a high respiratory  rate (OR 1,5); having moderate (OR 1.7) or severe (OR 2.9) pneumonia  (CURB-65); aspartate aminotransferase transaminase ≥ 100 IU/l (OR 2.1);  lactate dehydrogenase ≥ 360 IU/L (OR 1.6); procalcitonin > 0.5 ng/mL (OR  1.8); creatine kinase ≥ 294 U/L (OR 1.5); D-dimer > 3,000 ng/mL (OR 1.5);  hemoglobin < 11.6 g/dL (OR 1.4) and C-reactive protein > 120 mg/L (OR 1.2;  requiring respiratory support within the first 48 hours (oxygen therapy [OR  2.0], non-invasive ventilation [OR 2.8], and mechanical ventilation [OR 3.5]);  and being treated with interferon-beta (OR 1.5). On the contrary, being under  80 years of age was associated with lower mortality.

Conclusions: The analysis, based on the data in the RERFAR registry, showed that the factors associated with poorer prognosis were older age,  assessed using the CURB-65 scale, level of respiratory support required,  severe pneumonia (CURB-65), hypertransaminasemia, elevated creatine  kinase, lactate dehydrogenase, and D-dimer levels, anemia, and elevated  respiratory rate.