Hospital-Acquired COVID-19 among Patients of Two Acute-Care Hospitals: Implications for Surveillance.

OBJECTIVES: We quantified hospital-acquired COVID-19 during the early phases of the pandemic, and we evaluated solely temporal determinations of hospital acquisition. DESIGN: Retrospective observational study during early phases of the COVID-19 pandemic, March 1-November 30, 2020. We identified laboratory-detected SARS-CoV-2 from 30 days before admission through discharge. All episodes detected after hospital day 5 were categorized by chart review as community or unlikely hospital-acquired, or possible or probable hospital-acquired. SETTING: Two acute-care hospitals in Chicago, IL. PATIENTS: All hospitalized patients including an inpatient rehabilitation unit. INTERVENTIONS: Each hospital implemented infection-control precautions soon after identifying COVID-19 cases, including patient- and staff-cohorting, universal masking, and restricted visitation policies. RESULTS: Among 2,667 patients with SARS-CoV-2, detection before hospital day six was most common (n=2,612; 98%); days 6-14 uncommon (n=43; 1.6%); and, after day 14, rare (n=16; 0.6%). By chart review, most episodes after day 5 were categorized as community-acquired, usually because SARS-CoV-2 had been detected at a prior healthcare facility (68% of cases on days 6-14; 53% of cases after day 14). Incidence for possible and probable hospital-acquired cases, per 10,000 patient-days, was similar for ICU- and non-ICU patients at Hospitals A (1.2 vs 1.3, difference = 0.1; 95% CI, -2.8 to 3.0) and B (2.8 vs 1.2, difference = 1.6; 95% CI, -0.1 to 4.0). CONCLUSIONS: Most patients were protected by early and sustained application of infection-control precautions, modified to reduce COVID-19 transmission. Using solely temporal criteria to discriminate hospital- vs community-acquisition would have misclassified many "late-onset" SARS-CoV-2 positive episodes.

Computing antimicrobial use/antimicrobial resistance ratios: A novel way to assess inpatient antimicrobial utilization using current National Healthcare Safety Network metrics.

BACKGROUND: Current methods for benchmarking inpatient antimicrobial use (AU) could benefit from combining AU with antimicrobial resistance (AR) information to provide metrics benchmarked to microbiological data; this may yield more instructive and better risk-adjusted measurements than AU and AR in isolation. METHODS: In this retrospective single-center study, we computed facility-wide AU/AR ratios from 2019 to 2020 for specific antimicrobial agents and corresponding AR events, and compared median monthly AU/AR ratios between March 2019 through December 2019 (pre-COVID period) and March 2020 through December 2020 (COVID period). Aggregate AU was expressed as a ratio to aggregate AR events for antimicrobials that typically have activity against the AR organism and are frequently used to treat the AR organism in clinical practice. We also computed AU/AR ratios in our surgical intensive care unit in the pre-COVID period. RESULTS: High-median facility-wide monthly AU/AR ratios were observed for intravenous vancomycin/methicillin-resistant Staphylococcus aureus, with 130.0 in the pre-COVID period and 121.3 in the COVID period (p =.520). Decreases in facility-wide median monthly AU/AR ratios were observed between periods for meropenem/ESBL Enterobacterales (20.9 vs. 7.9, p < .001), linezolid/vancomycin-resistant Enterococcus (48.5 vs. 15.8, p =.004), and daptomycin/vancomycin-resistant Enterococcus (32.2 vs. 4.8, p = .002). Increases in facility-wide median monthly AU/AR ratios were observed between periods for ceftazidime-avibactam/carbapenem-resistant Enterobacterales (0.0 vs. 3.2, p = .020) and ceftazidime-avibactam/multidrug-resistant Pseudomonas aeruginosa (0.0 vs. 4.0, p = .017). The AU/AR ratio for intravenous vancomycin/methicillin-resistant S. aureus in the surgical intensive care unit was 191.5 in the pre-COVID period. CONCLUSIONS: AU/AR ratios may be used to supplement current AU and AR metrics. Future directions should include the development of more AU metrics benchmarked to microbiological information. AU metrics more specific to transplant infectious diseases should be developed.

Epidemiology of COVID-19 vs. influenza: Differential failure of COVID-19 mitigation among Hispanics, Cook County Health, Illinois.

BACKGROUND: During the early phases of the COVID-19 pandemic in the U.S., African-American or Hispanic communities were disproportionately impacted. To better understand the epidemiology and relative effects of COVID-19 among hospitalized Hispanic patients, we compared individual and census-tract level characteristics of patients diagnosed with COVID-19 to those diagnosed with influenza, another viral infection with respiratory transmission. We evaluated temporal changes in epidemiology related to a shelter-in-place mandate. METHODS: We evaluated patients hospitalized at Cook County Health, the safety-net health system for the Chicago metropolitan area. Among self-identified hospitalized Hispanic patients, we compared those with influenza (2019-2020 season) to COVID-19 infection during March 16, 2020-May 11, 2020. We used multivariable analysis to identify differences in individual and census-tract level characteristics between the two groups. RESULTS: Relative to non-Hispanic blacks and whites, COVID-19 rapidly increased among Hispanics during promotion of social-distancing policies. Whereas non-Hispanic blacks were more likely to be hospitalized for influenza, Hispanic patients predominated among COVID-19 infections (40% relative increase compared to influenza). In the comparative analysis of influenza and COVID-19, Hispanic patients with COVID-19 were more likely to reside in census tracts with higher proportions of residents with the following characteristics: Hispanic; no high school diploma; non-US citizen; limited English speaking ability; employed in manufacturing or construction; and overcrowding. By multivariable analysis, Hispanic patients hospitalized with COVID-19 compared to those with influenza were more likely to be male (adjusted OR = 1.8; 95% CI 1.1 to 2.9), obese (aOR = 2.5; 95% CI 1.5 to 4.2), or reside in a census tract with ≥40% of residents without a high-school diploma (aOR = 2.5; 95% CI 1.3 to 4.8). CONCLUSIONS: The rapid and disproportionate increase in COVID-19 hospitalizations among Hispanics after the shelter-in-place mandate indicates that public health strategies were inadequate in protecting this population-in particular, for those residing in neighborhoods with lower levels of educational attainment.