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

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

To facilitate better control of MRSA in hospital settings, several critical questions need to be answered. First, what are the relative roles of nosocomial transmission and infection importation from the community? Public health officials require an accurate assessment of the current force of infection within and into hospitals in order to deploy appropriate containment measures; however, with the increasing prevalence of MRSA in the community, disentangling HAIs from infections imported from the community has become difficult. Second, how many patients are colonized, and who and where are these high-risk individuals? Effective control would benefit from accurate determination of asymptomatic colonization rates in the general population; failure to estimate and account for colonization may result in long-term control issues (Cooper et al., 2004a) . Although colonized patients can be identified using sequencing methods (Harris et al., 2010; Long et al., 2014) , the expense of these assays limits their application, particularly in underdeveloped countries where MRSA has become endemic. In light of this situation, mathematical modeling offers an alternative approach for locating individuals with a high probability of colonization and guiding the targeted deployment of laboratory testing (Grundmann and Hellriegel, 2006; van Kleef et al., 2013; Opatowski et al., 2011) . However, this inference problem is again complicated by the unobserved stealth transmission dynamics that occurs in the highly complex time-varying contact networks of the real world (Donker et al., 2010; Vanhems et al., 2013; Jarynowski and Liljeros, 2015; Obadia et al., 2015a; Obadia et al., 2015b; Rocha et al., 2016; Nekkab et al., 2017; Duval et al., 2018) .