Reducing antimicrobial resistance in the community by restricting prescribing: can it be done?

The strategy of decreasing antimicrobial prescribing to reduce existing antimicrobial resistance appears attractive, but its effectiveness, particularly in the community setting, remains unclear. Contrasting results obtained from the relatively few studies in this area confuse matters further. Prescribing reductions have successfully reduced the prevalence of resistance among respiratory pathogens, but in these cases single bacterial clones dominated the resistant population. In contrast, this strategy has not succeeded in reducing plasmid-encoded resistance among Escherichia coli. The reasons why some prescribing restriction policies are more successful than others are complex, with the three key determinants being the fitness cost of resistance, the clonal structure of the resistant bacterial population and co-selection of resistant organisms by other antimicrobials. The resistant bacterial phenotypes that are likely to be the easiest to eliminate will be those composed of relatively clonal populations that bear a fitness cost of resistance and are not significantly subjected to co-selection by other antimicrobials. Plasmid-encoded resistance seldom meets these criteria and, hence, is likely to be the most difficult to reduce.