On the basis of a large outbreak of vancomycin-resistant Enterococcus faecium in a German university hospital, we estimated costs (≈1 million Euros) that could have been avoided by early detection of the imminent outbreak. For this purpose, we demonstrate an easy-to-use statistical method. R ecently, vancomycin-resistant Enterococcus faecium (VRE) has been detected with increasing frequency in Germany (1). Although some hospitals have reported only sporadic fi ndings, others have been faced with extended outbreaks (2-5). Apart from threatening patient health, these pathogens have an unfavorable economic impact on resources for healthcare in general. Control of VRE has proven to be costly and time-consuming (6). We therefore examined to what extent early implementation of control measures could have prevented these undesirable consequences in an outbreak of VRE at a university hospital in southwestern Germany.At this hospital ≈68,000 inpatients and 220,000 outpatients are treated each year. In 2003, 5 VRE-colonized patients were identifi ed. In 2004, VRE were fi rst detected in April. While at most 3 cases per month were observed through July, 8 colonized patients were found in August (Figure 1, panel A). The number of colonized patients remained relatively high in the following months. By the end of December, the cumulative number of patients with VRE was 48 (Figure 1, panel B). Although medical microbiologists were concerned about a possible outbreak as early as August, decision makers were reluctant to acknowledge a situation that needed action to be taken until January 2005. At that time, an infection control program was implemented. It included VRE screening in stool and anal/rectal swab samples from patients exhibiting an increased risk for VRE carriage (2). This program resulted in a sharp increase of detected cases to a total of 105 patients for February and March 2005 (Figure 1, panel A).Retrospectively, we examined whether comprehensible proof of an imminent outbreak could have been given early enough to have convinced decision makers of the need for control measures. For early outbreak detection, we have chosen a simple approach that can be easily adopted by infection control nurses dealing with multiresistant patho-