Low frequency of community-acquired bacterial co-infection in patients hospitalized for COVID-19 based on clinical, radiological and microbiological criteria; a retrospective cohort study.

Background To define the frequency of respiratory community-acquired bacterial co-infection in patients with coronavirus disease 2019 (COVID-19) based on a complete clinical assessment, including prior antibiotic use, clinical characteristics, inflammatory markers, chest computed tomography (CT) results and microbiological test results. Methods This study was conducted within a cohort of prospectively included patients admitted for COVID-19 in our tertiary medical centres between 1-3-2020 and 1-6-2020. A multidisciplinary study team developed a diagnostic protocol to retrospectively categorize patients as unlikely, possible or probable bacterial co-infection based on clinical, radiological and microbiological parameters in the first 72 hours of admission. Within the three categories, we summarized patient characteristics and antibiotic consumption. Results Among 281 included COVID-19 patients, bacterial co-infection was classified as unlikely in 233 patients (82.9%), possible in 35 patients (12.4%) and probable in 3 patients (1.1%). Ten patients (3.6%) could not be classified due to inconclusive data. Within 72 hours of hospital admission, 81% of the total study population and 78% of patients classified as unlikely bacterial co-infection received antibiotics. Conclusions COVID-19 patients are unlikely to have a respiratory community-acquired bacterial co-infection. Prospective studies should define the safety of restrictive antibiotic use in COVID-19 patients.