Abstract
OBJECTIVE: To describe antibiotic prescribing and microbiological findings in patients admitted to two London hospitals with COVID-19. METHODS: This is a retrospective review of confirmed SARS-CoV-2 infected adults admitted between 9th February and 10th May 2020. Demographics, critical care unit (CCU) admission, antibiotic prescribing and microbiology results within 10 days of COVID-19 diagnosis were analysed. RESULTS: 1155 patients were identified. 32.9% (380) died. 12.4% (143) had positive microbiology. After excluding likely contaminants, 6.9% (80) had clinically significant microbiology. The most common organisms isolated from blood cultures were Escherichia coli 9.5% (7), Klebsiella pneumoniae 4.0% (3), and Staphylococcus aureus 2.7% (2). A high percentage of blood cultures yielded coagulase negative staphylococci (51/74, 68.9%) and likely represented contamination. Organisms isolated from lower respiratory tract samples included Candida albicans 44.4% (12), Staphylococcus aureus 22.2% (6), Klebsiella species 11.0% (3), Pseudomonas aeruginosa 11.0% (3), and Citrobacter species 11% (3). Legionella and pneumococcal urinary antigen tests were positive in 0/117 and 2/71 (2.8%) samples. 91% (1051) of patients received antibiotics. Clarithromycin (24.2% total antibiotic use) and amoxicillin (21%) were most frequently used, followed by piperacillin-tazobactam (12.6%), gentamicin (10.6%), co-amoxiclav (9.3%) and meropenem (3.2%). Piperacillin-tazobactam or meropenem use was associated with a higher length of stay and mortality. CONCLUSIONS: Positive microbiology in COVID-19 patients is uncommon. Antibiotic use was widespread, despite lack of microbiological evidence of co-infection. When present, positive microbiology was more likely due to gram negative bacteria. Current local clinical and antimicrobial guidelines have incorporated these findings and recommend against routine antibiotic use in COVID-19 patients.