Traditional definition of healthcare-associated influenza underestimates cases associated with other healthcare exposures in a population-based surveillance system

Abstract

Objective: To provide comprehensive population-level estimates of the burden of healthcare-associated influenza. Design: Retrospective cross-sectional study. Setting: US Influenza Hospitalization Surveillance Network (FluSurv-NET) during 2012–2013 through 2018–2019 influenza seasons. Patients: Laboratory-confirmed influenza-related hospitalizations in an 8-county catchment area in Tennessee. Methods: The incidence of healthcare-associated influenza was determined using the traditional definition (ie, positive influenza test after hospital day 3) in addition to often underrecognized cases associated with recent post-acute care facility admission or a recent acute care hospitalization for a noninfluenza illness in the preceding 7 days. Results: Among the 5,904 laboratory-confirmed influenza-related hospitalizations, 147 (2.5%) had traditionally defined healthcare-associated influenza. When we included patients with a positive influenza test obtained in the first 3 days of hospitalization and who were either transferred to the hospital directly from a post-acute care facility or who were recently discharged from an acute care facility for a noninfluenza illness in the preceding 7 days, we identified an additional 1,031 cases (17.5% of all influenza-related hospitalizations). Conclusions: Including influenza cases associated with preadmission healthcare exposures with traditionally defined cases resulted in an 8-fold higher incidence of healthcare-associated influenza. These results emphasize the importance of capturing other healthcare exposures that may serve as the initial site of viral transmission to provide more comprehensive estimates of the burden of healthcare-associated influenza and to inform improved infection prevention strategies.