Inference and control of the nosocomial transmission of methicillin-resistant Staphylococcus aureus

Pei, Sen; Morone, Flaviano; Liljeros, Fredrik; Makse, Hernán; Shaman, Jeffrey L

To address these issues, here we develop an agent-based network model-Bayesian inference system for estimating unobserved colonization and importation rates from simple incidence records. We use this system to infer the transmission dynamics of the most commonly diagnosed MRSA strain, UK EMRSA-15 Das et al., 2013) , from multiple Swedish hospitals (Materials and methods). Key features estimated include the number of infections acquired in hospital and imported from outside, as well as the locations of individuals with a high colonization probability. Such information is crucial for designing cost-effective control measures (Cooper et al., 2003; eLife digest Antibiotic-resistant bacteria like the Methicillin-resistant Staphylococcus aureus (MRSA) can live in people for many years without making them sick. During this time, the bacteria can spread to others who come in contact with the MRSA-infected person. The number of people with stealth MRSA infections living in the community has been increasing. As a result, hospitals may not only be dealing with MRSA infections that originated onsite, but also cases imported from the community. That makes tracking and controlling MRSA infections in hospitals difficult. Now, Pei et al. show that computer modeling can help identify the role MRSA infections from the community play in hospital outbreaks and test ways to control them. In the experiments, data from an MRSA outbreak that occurred at 66 Swedish hospitals over 6 years were analyzed using statistical methods and computer modeling. This helped to identify patients who were likely colonized with MRSA within the hospital and those who had acquired it in the community. Next, Pei et al. used computer modeling to test what would have happened if these high-risk individuals had received interventions to prevent them from spreading MRSA in the hospital. This showed that targeting individuals at high-risk of a MRSA infection could reduce the spread of MRSA in the hospital.