Background: The COVID-19 pandemic has brought unprecedented changes to the way healthcare is delivered. While the impact of death and comorbidity due to infections with SARS-CoV-2 is a ubiquitous concern, it is unclear what effect this pandemic is having on hospitalized patients who are not infected with the SARS-CoV-2 virus. We seek to estimate the impact of the COVID-19 pandemic on patients without COVID-19 infection who are admitted to the hospital with sepsis.
Methods: We used our institution’s electronic health record to perform a retrospective, multicenter cohort study of adult patients hospitalized due to sepsis within a United States hospital system during March 1 – May 31 of 2019 and 2020. We excluded patients with a COVID-19 positive status. Independent variables included the presence of a COVID-19 test order, isolation status, and the timing of hospitalization in relation to the COVID-19 pandemic. Cohorts were based on the time-period during which patients were hospitalized (2019 or 2020). Subcohorts were derived from all possible combinations of the stated independent variables. Our primary outcome was in-hospital mortality. Chi-Square testing was performed on all possible pairs of subcohorts to test for associations with mortality.
Results: We identified 17,064 patients from the spring months of 2019 and 2020 that were hospitalized with sepsis and were without a COVID-19 positive status. Baseline characteristics are reported in Table 1. Mean initial qSOFA score was 1.35 for patients hospitalized in 2019 and 1.03 for patients hospitalized in 2020. Mortality was 19.9% for the 2019 cohort and 18.3% for the 2020 cohort (P=0.0096). During subcohort comparison analysis, a non-statistically significant difference in mortality was observed between the ‘2020 Isolation and Tested Negative’ subcohort (i.e. persons under investigation for COVID-19) and the ‘2020 No Isolation and Not Tested’ subcohort (19.5% vs 17.6%, P=0.1261; Table 2).
Conclusions: Among patients with sepsis that is not due to COVID-19, those hospitalized during the COVID-19-pandemic (2020 cohort) demonstrated a comparable mortality rate to patients hospitalized pre-pandemic; this is despite being characterized by a notably lower mean qSOFA score. Had the COVID-19 pandemic not contributed to mortality, we would have expected to observe a more preferable mortality rate in the 2020 cohort due to its lower mean qSOFA score. Our study found a non-statistically significant elevation in mortality for persons under investigation for COVID-19 when compared to persons not under investigation for COVID-19 during the pandemic, suggesting that certain processes (including, but not limited to, patient isolation and viral testing) may have affected inpatient outcomes. There remains a need to further elucidate the impacts that the COVID-19 pandemic has had on healthcare workers and systems so that we may better prepare for similar events. Future studies should incorporate regional analysis to help control for geographic variations in outbreak timelines, population dynamics, and regional operations.
