Mathematical studies of the dynamics of antibiotic resistance

In Part I we discuss models for the spread of nosocomial antibioticresistant bacteria. We focus on pathogens for which re-admission of colonized individuals is important, i.e., the feedback-loop between the hospital and the extramural population. In Chapter 2 we will discuss an analytical model and in Chapter 3 we will focus on colonization with Methicillin-resistant Staphylococcus aureus (MRSA). We use both an analytical and a simulation model. Both models suggest that isolation of identified carriers of MRSA in hospitals combined with either screening on admission of high-risk patient or the screening of contact patients in case of the identification of an unexpected MRSA carrier in the hospital, may be sufficient to prevent high levels of MRSA in the hospitals. However, the so-called Dutch search and destroy policy in which both interventions are applied ensures that the current low prevalence level of MRSA in the Netherlands is far less sensitive to changes in the parameter values. In Part II we use real hospital data to draw conclusions for specific pathogens/diseases. In Chapter 4 we use a simple observation to disentangle the phenomena that patients who acquire an infection are likely to stay longer in a unit and that patients who stay longer in a unit are more likely to acquire an infection. In Chapter 5 we use likelihood methods in a Markov chain approach to distinguish between different infection routes on the basis of the fluctuations in the prevalence. This method is applied to data for colonization with two different pathogens. This method is also used to determine optimal culture frequencies