Integrating the Patient Sharing Network Among Healthcare Facilities into the Containment of Carbapenem-Resistant Enterobacterales in Tennessee

Abstract

Carbapenem-Resistant Enterobacterales (CRE) are a group of antimicrobial-resistant organisms (AROs) that primarily cause healthcare-associated infections. Patient movement across healthcare facilities has been shown as a means of CRE transmission. To understand the impact of incorporating the patient sharing network into the containment of CRE in Tennessee, this research aims to assess (1) the risk factors associated with re-hospitalizations of CRE patients, (2) the association between patient sharing network among hospitals and CRE prevalence, and (3) the impact of incorporating the patient sharing network in a coordinated containment approach in Tennessee. This research used a retrospective cohort of known CRE infections from the Tennessee surveillance data. A multivariable Cox model identified that CRE-infected patients re-hospitalized within 12 months had more comorbidities, higher sepsis rates, and more healthcare exposures than patients not re-admitted. These patient characteristics were used to select CRE surrogates, a subset of the general patient population with similar characteristics to CRE-infected patients. Each doubling of incoming direct and 12-month indirect transfers of CRE surrogates was independently associated with a 33% increase in hospital-level CRE cases/1,000 patient-days. A mathematical model was used in simulations of a coordinated containment strategy guided by a patient sharing network from the general patient transfers. The simulations predicted a 21% reduction in total CRE cases after three years. A containment that targeted intervention to facilities identified by the network from CRE surrogates resulted in a further decrease in CRE cases (26%) and required fewer interventions to targeted hospitals. Patient sharing network structure influences the impact of the regional CRE containment effort.