Describing the monthly variability of hospital-onset Clostridioides difficile during early coronavirus disease 2019 (COVID-19) using electronic health record data

Abstract

Objective: To assess the relative risk of hospital-onset Clostridioides difficile (HO-CDI) during each month of the early coronavirus disease 2019 (COVID-19) pandemic and to compare it with historical expectation based on patient characteristics.

Design: This study used a retrospective cohort design. We collected secondary data from the institution's electronic health record (EHR).

Setting: The Ohio State University Wexner Medical Center, Ohio, a large tertiary healthcare system in the Midwest.

Patients or participants: All adult patients admitted to the inpatient setting between January 2018 and May 2021 were eligible for the study. Prisoners, children, individuals presenting with Clostridioides difficile on admission, and patients with <4 days of inpatient stay were excluded from the study.

Results: After controlling for patient characteristics, the observed numbers of HO-CDI cases were not significantly different than expected. However, during 3 months of the pandemic period, the observed numbers of cases were significantly different from what would be expected based on patient characteristics. Of these 3 months, 2 months had more cases than expected and 1 month had fewer.

Conclusions: Variations in HO-CDI incidence seemed to trend with COVID-19 incidence but were not fully explained by our case mix. Other factors contributing to the variability in HO-CDI incidence beyond listed patient characteristics need to be explored.

Figure 1.

Percentage of total admissions each month that were associated with a COVID-19 diagnosis or hospital-onset Clostridioides difficile infection (HO-CDI).

Figure 2.

Posterior mean log standardized incidence ratio (SIR) and 90% credible interval for hospital-onset Clostridioides difficile infection (HO-CDI) by month from January 2018 to April 2021.