Antibiotic Resistance in Intensive Care Units: Dynamics of Colonization

The dynamics of colonization of antibiotic-resistant bacteria in hospital settings are complex and depend on bacteria and healthcare worker related characteristics. Many factors influence colonization and in addition these factors interact with each other as well. Knowledge of local resistance epidemiology such as dominant colonization route (endogenous and cross-transmission), antibiotic use, contact rates, cohorting of healthcare workers (extent one-to-one care) and adherence to hand hygiene is important in the development of control strategies. In intervention studies these factors need to be carefully measured before and after implementation of a certain strategy. In addition, only one strategy should be implemented at a time so that the effect of this specific strategy can be evaluated. So far, most intervention studies did not or only partially comply with the above, which makes conclusions drawn from these studies difficult to interpret (chapter 1). During a ten-week period in an intensive care unit at Cook County Hospitals Chicago, microbiological surveillance and bacterial genotyping, without reporting results and subsequent isolation procedures, showed that the majority of colonizations with methicillin-susceptible and methicillin-resistant Staphylococcus aureus (MSSA and MRSA) were imported into the ICU and were not the result of cross-transmission. When culture results would have been reported and patients isolated subsequently, this probably would have led to the false conclusion that this strategy was indeed succesful in the prevention of cross-transmission in this unit (chapter 2). Measurement of contact rates, cohorting and adherence to hand hygiene showed that the magnitude of these factors varies for different healthcare workers. The risk of bacterial transmission for physicians and nurses was calculated using these measurements and it showed that physicians had a 1.6 times higher risk to transmit bacteria than nurses in that particular intensive care unit (chapter 3). The dominant colonization route can be determined by microbiological surveillance and bacterial genotyping. The latter is time-consuming, labour-intensive and costly. Mathematical modeling seems a promising tool to determine colonization routes based on minimal clinical and surveillance data (chapter 5). Although time-consuming and costly, combining different molecular typing techniques and epidemiological linkage of patients is relevant to gain insight in resistance epidemiology because it provides more detail about the origin of resistance determinants and underlying dynamics (chapter 4). A study to determine colonization dynamics of cephalosporin-resistant Enterobacteriaceae and to evaluate the effects of a change in antibiotic prescription (cycling vs restriction policy) was performed. Colonization was predominantly acquired endogenously with the use of amoxicillin-clavulanate as the only modifyable risk factor. A stepwise reduction in the use of amoxicillin-clavulanate, at the cost of quinolone use, failed to reduce acquisition rates of cephalosporin-resitant Enterobacteriaceae in this setting. Resistance to quinolones, however, increased during this period (chapter 8). Still, more well-designed studies are needed to gather knowledge about which strategies to use in the control of antibiotic resistance