Abstract
Accurate characterization of respiratory bacterial co-infection is critical for guiding empirical antibiotic treatment for hospitalised patients with coronavirus disease 2019 (COVID-19). We retrospectively assessed the clinical and analytical predictors of respiratory bacterial co-infection and described the empirical use of antibiotics in COVID-19 hospitalised patients. Respiratory bacterial co-infection was documented in 6.9% (80/1157) of the patients. The predominant bacteria isolates were Haemophilus influenzae, followed by Streptococcus pneumoniae and Pseudomonas aeruginosa. Respiratory bacterial co-infection was associated with having had a positive culture for a respiratory pathogen in the last year (OR = 25.89), dyslipidaemia (OR = 2.52), heart failure (OR = 7.68), ferritin levels < 402 ng/mL (OR = 2.28), leukocyte count > 8.7 × 10(9)/L (OR = 2.4), and patients with chronic obstructive pulmonary disease treated with inhaled corticosteroids (OR = 12.94). Empirical antibiotic treatment was administered in 42.33% of patients, although it declined across the distinct study periods (p < 0.001). Patients admitted to intensive care units harbouring co-infection exhibited worse outcomes and more bacterial secondary infections. In conclusion, respiratory bacterial co-infection prevalence was low, although it could lead to unfavourable outcomes. Moreover, the percentage of empirical antibiotic treatment remained high. The study's findings allowed the identification of several predictors for respiratory bacterial co-infection and could help implement adequate antibiotic stewardship measures.