The primary method for preventing HAIs is to reduce colonization because colonized individuals are at higher risk of infection and death. The U.S. Centers for Disease Control and Prevention (CDC) advocates several strategies for lowering MDRO colonization through reduced transmission, including improving hand hygiene, cohorting and enforcing contact precautions for colonized patients, educating healthcare workers (HCWs), minimizing medical device use, improving surveillance, screening, and notification processes, streamlining communication between healthcare facilities, improving antimicrobial stewardship, increasing rates and effectiveness of decolonization, and cleaning the hospital environment. However, gaps remain in understanding how pathogens are transmitted, which limits the ability to attribute the effects of an intervention. Additionally, hospitals differ in their operations (e.g., staffing ratios, equipment management, cleaning practices) and patient composition, limiting the generalizability of the results of intervention from one hospital to another.
Modeling and simulation studies of healthcare delivery can be useful in developing an understanding of transmission pathways. These studies are also valuable at determining which elements of an intervention drive rates of HAIs, both of which can aid understanding of the potential effectiveness of an intervention. In addition, by assessing the relative importance of different aspects of hospital operations, models can be flexibly adapted to determine the potential effectiveness of prevention and control strategies at a more hospital-specific level, which is essential in a resource-constrained environment. A framework for understanding how local operations and patient populations influence the effectiveness of interventions is needed to guide to hospitals in implementing such measures to reduce the incidence of HAIs.
The goals of this proposal are to provide a framework for hospitals to identify patients at risk for colonization and infection during their stay using data on patient network connections and evaluating more effective and targeted intervention options to control the spread of MDROs and reduce HAIs. The specific goals of this project are:
- Assess the impact of operational practices, including HCW staffing and mobile equipment movement, on the transmission of MDROs.
- Develop a predictive model of patients likely to become colonized or infected with an MDRO.
- Assess the potential impact of interventions to reduce colonization and infection in identified high-risk patients.
- Assess the relative importance of potential pathways of transmission.
- Evaluate the potential effect of operational changes to reduce connectivity between susceptible and colonized patients.
- Evaluate the impact, cost, and cost-effectiveness of IPC interventions at the hospital level.
A significant challenge in reducing transmission, colonization, and infection is successfully identifying patients at risk. Using data from electronic health records (EHRs), we will develop prediction models of colonization risk to aid interventions and reduce transmission. Healthcare facilities can implement many interventions to curb transmission and colonization, which can be applied to individual patients or, more broadly, across a facility. Identifying the direct benefit of a single intervention is difficult in practice as often, many interventions are combined. However, modeling facilitates the deconstruction of the marginal effect of an additional intervention in terms of infections and costs.
- Analysis of EHR data for hospitalized patients can be used to develop a digital twin within a hospital.
- Development of models of MDRO spread within a hospital, which has been used for multiple studies on evaluating different kinds of interventions for diseases such as MRSA.
- Strong prediction of MRSA risk for patients in a hospital setting.
- Theoretical studies on testing and the inference of HAIs in hospitals.
Greg Madden, UVA School of Medicine
Costi Sifri, UVA School of Medicine
Eili Klein, One Health Trust, JHU