391. Outcomes of Empiric Antimicrobial Therapy in COVID-19 Positive Patients

Abstract

Background: The COVID-19 pandemic has revealed new challenges for antimicrobial stewardship. Optimal medical treatment is not completely understood at this time. The epidemiology and outcomes of bacterial co-infections are not well-established; however, empiric antibiotic (abx) use is anecdotally common. The purpose of this study is to characterize empiric antimicrobial drug selection and timing in COVID-19 and evaluate the impact on patient outcomes. Methods: Cross-sectional cohort study for COVID-19 positive inpatients from March 1, 2020 to June 1, 2020 at an academic medical center and 4 community hospitals. Inclusion: patients with a documented positive COVID-19 PCR nasopharyngeal swab. Exclusion: patients less than 18 years; deceased or transitioned to hospice within 24 hours of admission. Primary endpoint: empiric abx drug, initiation, duration and indication. Additional data collected: severity of illness, co-infection diagnosis, microbiology, and adverse drug effects (ADE). Clinical outcomes included time to recovery by COVID-19 ordinal outcome, clinical status at day 15, and readmission. Results: 400 patients were included with 27% from the ICU. COVID symptom category included mild (23.8%), moderate (53%), severe (15%), and critical (8.3%).within 1 day of admission. Most common documented indication community-acquired pneumonia (69%). Identified 43 (10.8%) microbiologically confirmed co-infections, including 5 MRSA and 7 Pseudomonas. Median duration of initial abx 4 days. 54/322 (16.8%) had abx restarted after discontinuation. Median days to recovery without abx was 10 days (7 - 14) and 14 days (9 - 20) with abx. Patient characteristics and outcomes described in table 1. 74 abx related ADE were identified: gastrointestinal 37 and renal 22. Conclusion: It's difficult to distinguish bacterial and Covid-19 in coinfections in patients ill enough to be hospitalized. Longer courses of empiric abx therapy were prevalent as the severity of illness increased. However, the low frequency of microbiologically confirmed bacterial co-infections results in potentially unnecessary abx exposure. This exposure increases risk of abx ADE and may not improve clinical outcome.