Comment on: Gemifloxacin once daily for 5 days versus 7 days for the treatment of community-acquired pneumonia: a randomized, multicentre, double-blind study

make. However, they go on to suggest that these relatively low concentrations of ertapenem will be ineffective against this strain. It is with this issue that we are concerned. In our article, we indicate that the emergence and spread of resistance are greatly enhanced at subinhibitory concentrations; however, resistance emergence is also a dynamic process of the progressive concentration of resistant strains within the clinical population, often by the same or related strains. In the individual case that Brink et al. describe, ertapenem did not select carbapenem-resistant P. aeruginosa; however, we are concerned with the much broader picture for, in the face of thousands or even millions of such challenges, we maintain that selection of resistant strains is likely to occur and progressively increases their concentration in the clinical setting. With regard to their later comment, the widespread dissemination of extended-spectrum b-lactamases (ESBLs) and transferable ampC b-lactamases does demonstrate that we have not used cephalosporins optimally. We do, though, have the benefit of hindsight and different strategies; in particular, the use of cephalosporins in combination with a b-lactamase inhibitor might have suppressed the huge problem of cephalosporin resistance that we now face. It should be possible to learn from experience that we must not take individual treatment cases in isolation as a model for drug strategy because, unless we consider the broader potential impact of using weaker drugs, we may well face a new ‘ESBL’ problem, but this time it will be carbapenemases.