244. Bacterial co-infection in COVID-19: an analysis of a National Cohort of Hospitalized U.S. Veterans

Abstract

Abstract Background Bacterial co-infection has been reported with COVID-19, but the extent of co-infection nationally is unclear. We sought to describe the temporal and spatial trends in bacterial co-infection across the US among COVID-19 positive admissions to Veterans Affairs (VA) hospitals. Methods This retrospective cohort study included patients admitted to VA hospitals from March 1, 2020 through May 31, 2022 with a positive COVID-19 test within the previous 14 days and up to 2 days after admission. We summarized temporal and spatial patterns of bacterial co-infection, defined as a positive clinical microbiology culture for the bacterial pathogens listed in Table 1, defined as either community-onset (COI, within 2 calendar days of admission), or hospital-onset (HOI, > 2 calendar days after admission). We performed a univariate analysis including facility and patient factors and generated descriptive statistics to describe the frequency of occurrence over time and space overall, and within each organism. Organisms Table 1:List of organisms included in our study Results By the end of June 2021, there were 35,299 hospitalizations observed from 33,383 patients admitted with positive COVID-19 tests in VA. Co-infection was detected among 7.4% of hospitalizations (2.9% for COI and 4.7% for HOI). VA patients older than 70, Asian or unknown race, higher Charlson Comorbidity Index were more likely to experience HOI and COI. Facility-level rates of HOI and COI over the study period presented substantial geographic variability, ranging from 0 to 45.5 per 1000 patient days and from 0 to 6.98 per 100 hospitalizations for HOI and COI, respectively [Fig 1]. Between March 2020 and June 2021, monthly facility-level rates of HOI and COI also varied substantially within and between facilities [Fig 2]. Average monthly co-infection rates increased in the first three months of the pandemic, with HOIs subsequently declining gradually and COIs remaining stable across VA. The correlation coefficients between hospital mortality and facility-level co-infection rates for HOI and COI ranged from –0.5 to 0.7 [Fig 3]. Spatial variation Fig 1:Plot for spatial distribution across VA facilities of HOIs and COIs, measured as Co-Infections per 100 patient hospitalizations Temporal variation Fig 2:Plot for temporal distribution across VA facilities of HOI and COI, measured as Co-Infections per 100 patient hospitalizations Mortality correlation Fig 3:Plot for spatial distribution of the correlation coefficients between hospital mortality and HOI and COI Conclusion Bacterial co-infection was infrequent during hospitalization with COVID-19 in the VA healthcare system, and has mild to moderate association with hospital mortality. However, substantial geographic and temporal variation was observed. Disclosures Karim Khader, PhD, BioFire Diagnostics: Grant/Research Support.