785. Bacterial Co-infection and Empirical Antibiotic Therapy in Patients with COVID-19

BACKGROUND: Understanding the rate and composition of bacterial co-infection is important to determine antibiotic therapy in SARS-CoV-2 infection, but those vary according to healthcare settings and regional differences. We evaluated the rate of bacterial co-infection in hospitalized patients with COVID-19 in a single tertiary hospital in South Korea. METHODS: In this retrospective study, all the adult patients with COVID-19 hospitalized between Feb 2020 and Dec 2021 were included. Bacterial co-infection rate was assessed by results of sputum cultures, blood cultures, pneumococcal urinary antigen, Legionella urinary antigen, sputum Legionella pneumophilia PCR, and sputum multiplex PCR for Mycoplasma pneumoniae and Chlamydia pneumoniae. Characteristics and outcomes of patients were evaluated according to antibiotics exposure prior to hospitalization. RESULTS: Of 367 adult patients, 300 (81.7%) patients having sputum culture results were included in the analysis. Of these, 127 (42.3%) had a history of antibiotic exposure within 1 month before hospitalization. The coinfection rate within 48 hours of hospitalization was confirmed in 8.3% (25/300): 6.4% (11/163) of patients without prior antibiotic exposure and 11% (14/127) of patients with prior antibiotic exposure. In the group without prior antibiotic exposure, pathogens responsible for community-onset infections were isolated, whereas nosocomial pathogens were predominantly isolated in the antibiotic-exposed group. Empirical antibiotics were used in 144 (66%) of 275 patients without positive results for microbiological tests. Empirical antibiotic use in patients without positive results for microbiological tests was not significantly associated with 30-day mortality or in-hospital mortality after adjusting covariates including age, sex, comorbidity, anti-inflammatory treatment, and COVID-19 severity. [Figure: see text] [Figure: see text] CONCLUSION: In this study with a high rate of microbiological testing, bacterial co-infection was not frequent, and the results varied depended on previous exposure to antibiotics. Given the rarity of bacterial co-infection and the lack of potential benefits of empirical antibiotic therapy, the antibiotic use in patients with COVID-19 should be restricted as an important target of antibiotic stewardship. DISCLOSURES: All Authors: No reported disclosures.