ABSTRACT
Background. Antibiotic use among patients with COVID-19 is common, exceeds the prevalence of probable bacterial co-infection, and promotes development of resistant organisms. Lack of diagnostic microbiological data may prolong empiric broad-spectrum therapy. Here we evaluate the use of the BioFire FilmArray pneumonia panel (PP), a novel rapid diagnostic test, and antibiotic decisions among intensive care unit (ICU) patients with COVID-19. Methods. We conducted a retrospective review of adult ICU patients admitted with COVID-19 between January 2020 and May 2021 at an academic medical center. ICU patients who underwent bronchoscopy/bronchoalveolar lavage (BAL) with PP (PP group) were matched by age (< 65 or ≥65), BMI (< 30 or ≥30), and BAL date (within 60 days) to ICU patients who did not undergo BAL (no-BAL group). PP patients were matched by age and BMI to ICU patients who underwent BAL without PP (no-PP group). Antibiotic use was compared between groups. Chi squared analysis, t-test, and ANOVA were used for comparisons as appropriate. Results. 65 patients were included;the majority were male (65%), < 65 years (86%), and had BMI ≥30 (54%) (Table 1). Only 17 no-PP matches were identified for PP patients due to infrequent BALs. Similar proportion of patients in PP and no-PP groups had organisms identified from BAL (54% vs. 47%, p=0.65). Among PP patients with a detected organism, all (n=13) had subsequent changes in antibiotic regimen ≤72 hours after BAL;10/13 (77%) had a change targeted to detected organism and 5/13 (39%) had antibiotic narrowing. Among PP patients with no detected organism, only 4/11 (36%) had antibiotic narrowing or maintenance off antibiotics. In all groups, average antibiotic use exceeded 70% of admission duration. Table 1. Patient characteristics and antibiotic management. Abbreviations: BAL - bronchoalveolar lavage Conclusion. Rapid, highly sensitive diagnostic tests have potential to guide clinical decisions and promote antibiotic stewardship among patients with severe viral pneumonia and suspected bacterial co-infection. In this descriptive analysis, antibiotic management did not differ significantly with use of PP. While most patients with detected organism on PP had targeted antibiotic changes, a negative PP did not appear to influence antibiotic narrowing. Larger studies and provider education are needed to evaluate potential of the PP for antibiotic stewardship.
ABSTRACT
Background: Patients with COVID-19 may be at increased risk for secondary bacterial infections. At our quaternary care hospital in New York City, the rapid escalation of COVID-19 cases was accompanied by a massive surge in the need for hospital and critical care capacity. During this time, we noted a increase in infections caused by carbapenemase-producing Enterobacterales (CPE). Methods: We retrospectively assessed microbiology data to identify patients with positive testing for SARS-CoV-2 who had clinical cultures with meropenem-resistant and/or carbapenemase gene-positive Enterobacterales. We obtained microbiological and clinical data by manual chart review. Available clinical isolates underwent long-range genomic sequencing using the MinION (Oxford) for rapid genotyping, resistance gene detection, and phylogenetic analysis. Results: From March 1 to May 18, we identified 33 CPE isolates from 13 patients, including 29 Klebsiella pneumonia and four Enterobacter cloacae. Most patients (11/13) had a positive respiratory culture, and 7/13 developed bacteremia. All patients had prolonged, complex hospitalizations with extensive antibiotic exposure. We performed long-range sequencing on 19 isolates from 12 patients. 15/16 K. pneumoniae isolates belonged to sequence type (ST) 258 encoding KPC (14 KPC-2;1 KPC-3);one ST70 isolate encoded KPC-2. All four E. cloacae isolates belonged to ST270 and encoded NDM-1. Phylogenetic analysis of ST258 isolates including historical isolates from our hospital revealed a distinct lineage of isolates from COVID-19 patients (72% bootstrap support), with expected clustering of isolates from the same patient and patients that were cohorted together. Conclusion: While CPE have declined substantially in New York City in recent years, increased detection in patients with COVID-19 may signal a reemergence of these highly resistant pathogens in the wake of the global pandemic. System-level factors, such as the rapid scale-up of critical care capacity, while clearly needed to address the unprecedented reach of COVID-19, may have contributed to isolate clustering in these patients. Increased surveillance and antimicrobial stewardship efforts will be needed to mitigate the impact of CPE in the future.