ABSTRACT
Background: Reports indicate that COVID-19 patients have more bloodstream infections (BSI) on the intensive care unit (ICU) potentially due to lapses in infection control practice or other factors.Methods: Retrospective single-site study of ICU-BSIs in mechanically ventilated (MV) COVID-19 patients during the first pandemic year. Clinical, demographic and laboratory data including targeted pathogen genome sequencing was analysed during first (March 13th - May 31st 2020) and second (October 1st 2020 - March 15th 2021) pandemic waves.Findings: There were 305 MV-ICU patients in wave one and 440 in wave two with peak occupancy of 113 and 155 patients, respectively. The BSI rate was higher during both waves than pre-pandemic, but more in wave-two than wave-one, particularly during the first 28 days on ICU (14.1 vs. 9.4/1000 bed days; p=0.03) and with E. faecium (3.57 vs. 0.47/1000 bed days; p=0.0067). 22/28 (77%) of E. faecium BSIs had no microbiologically definable focus and the high wave-two rate could not be explained by transmission. Wave-two BSI-patients received more corticosteroids and tocilizumab and had higher crude hospital mortality compared with non-BSI patients (41% vs. 21% p<0.0001), a phenomenon not seen in wave one (32% vs. 27% p=0.551). Interpretation: MV-COVID-19 patients had a greater BSI-burden in wave two associated with the emergence of primary endogenous E. faecium. Identifying mechanisms and causal links between admission COVID-19 disease severity, immunomodulation, BSI and death could help identify new approaches to improving outcomes for COVID-19 patients.