1424. Modeling the Impact of Multi-Drug Resistant Organism (MDRO) Regional Containment Strategies in a Tennessee Patient Transfer Network

Abstract Background Centers for Disease Control and Prevention (CDC) guidance recommends coordinated multifacility interventions to contain selected MDRO, such as carbapenem-resistant Enterobacterales (CRE). A mathematical model using data from a northeastern state estimated these interventions would lead to 76% relative reduction in CRE prevalence in a hospital network connected through patient transfers, but whether results would be similar using data from another state is unclear. Methods We used Tennessee (TN) surveillance and hospital discharge data to estimate CRE transmissibility and patient transfers in a deterministic compartmental model to simulate regional spread of CRE. Simulations were initialized with the first clinical detection of CRE in the hospital with the most outgoing transfers. Interventions were initiated in facilities that detected CRE and those sending or receiving the largest numbers of patients to/from these facilities, including: 1) biweekly point prevalence surveys; and 2) enhanced infection control implemented immediately on CRE detection. We assumed interventions reduced intrafacility transmission by 20%. Surveys stopped after 2 consecutive negative rounds in a facility (Figure 1). We ran simulations using 2 patient transfer networks: 1) all hospitalized patients; and 2) CRE surrogates (patients clinically similar to CRE-infected patients) (Figure 2).Figure 1.Methods: Multifacility Susceptible-Infectious-Susceptible model (a) and simulations schematics (b). In the multi-facility SIS model, we assume hospitals are at constant occupancy. Each bed was occupied by either an infectious (I) or susceptible patient (S). (b) Interventions outlined in the simulations followed the recommendations of the Interim Guidance for a Public Health Response to Contain Novel or Targeted Multidrug-resistant Organisms (MDROs). Initial detection at the facility with the highest number of outgoing transfers followed by immediate containment efforts and reduced transmissibility in intervened hospitals by 20% by day 30. Simulations were run 35 times. Each run randomly selected a transmissibility value derived from the distribution in our estimates derived from CRE surveillance data in TN.Figure 2.Methods: the patient transfer network of Tennessee (TN) hospitals and communitiesThe network of Tennessee (TN) hospitals and communities, represented by Health Referral Regions (HRRs) according to the Dartmouth Atlas of Healthcare (https://www.dartmouthatlas.o