Associations between hospital structure, infection control and incidence of hospital-acquired viral respiratory infections: a 10-year surveillance study.

Background: Protecting patients from hospital-acquired viral respiratory infections is a major challenge, but the hierarchy of measures to achieve this is not yet completely clear.

Aim: To describe the epidemiology of hospital-acquired viral respiratory infections and associations with structural hospital factors and adherence to infection control protocols.

Methods: Retrospective cohort study conducted over 10 consecutive years (2014-2023) within 27 hospital wards in a 900-bed university hospital in Paris, France. All hospitalized adult patients who were tested for at least one virus on a respiratory sample during their stay were included. Structural factors (percentage of double occupancy rooms) and adherence to infection control protocols by healthcare workers (measured by consumption of alcohol-based hand sanitizer and of facemasks) were included as predictors in the model.

Main outcome and measure: Incidence of hospital-acquired viral respiratory infections, defined by a positive PCR test for at least one respiratory virus, performed at least 5 days after hospital admission. Data were analyzed on ward-year aggregated data, with a linear mixed-effects model.

Findings: Overall, 183 994 viral PCR tests were performed over the study period. Incidence of hospital-acquired viral respiratory infections was 0.57/1000 hospital-days. After adjustment on other factors (mean length of stay, use of PCR testing), incidence of hospital-acquired viral respiratory infections in a given ward was significantly associated with: the incidence of community-acquired viral respiratory infections among patients admitted to the ward (+ 0.10/1000 hospital-days per each additional point of incidence; P < 0.001), the number of double-occupancy rooms (+ 0.04/1000 hospital-days per each 10%-increase of double-occupancy rooms; P = 0.03) and masks consumption (+ 0.33/1000 hospital-days per 10 additional masks used per day; P = 0.04). Similar results were found for double-occupancy rooms (+ 0.01/1000 hospital-days per each 10%-increase of double-occupancy rooms; P = 0.012) in the sub-group analysis of influenza cases.

Conclusion: In a given hospital ward, the incidence of community-acquired cases and the proportion of double-occupancy rooms are independently associated with the incidence hospital-acquired viral respiratory infections.