Community-acquired and hospital-acquired bacterial co-infections in patients hospitalized with Covid-19 or influenza: a retrospective cohort study

Abstract

Abstract Background Bacterial co-infections are believed to be less frequent in patients with Covid-19 than influenza, but frequencies varied between studies. Methods This single-center retrospective, propensity score-matched analysis included adult patients with Covid-19 or influenza admitted to normal-care wards between 02/2014 and 12/2021. Covid-19 cases were propensity score matched to influenza cases at a 2:1 ratio. Community-acquired and hospital-acquired bacterial co-infections were defined as positive blood or respiratory cultures ≤ 48 h or > 48 h after hospital admission, respectively. The primary outcome was comparison of community-acquired and hospital-acquired bacterial infections between patients with Covid-19 and influenza in the propensity score-matched cohort. Secondary outcomes included frequency of early and late microbiological testing. Results A total of 1337 patients were included in the overall analysis, of which 360 patients with Covid-19 were matched to 180 patients with influenza. Early (≤ 48 h) microbiological sampling was performed in 138 (38.3%) patients with Covid-19 and 75 (41.7%) patients with influenza. Community-acquired bacterial co-infections were found in 14 (3.9%) of 360 patients with Covid-19 and 7 (3.9%) of 180 patients with influenza (OR 1.0, 95% CI 0.3–2.7). Late (> 48 h) microbiological sampling was performed in 129 (35.8%) patients with Covid-19 and 74 (41.1%) patients with influenza. Hospital-acquired bacterial co-infections were found in 40 (11.1%) of 360 patients with Covid-19 and 20 (11.1%) of 180 patients with influenza (OR 1.0, 95% CI 0.5–1.8). Conclusion The rate of community-acquired and hospital-acquired bacterial co-infections was similar in hospitalized Covid-19 and influenza patients. These findings contrast previous literature reporting that bacterial co-infections are less common in Covid-19 than influenza.