ABSTRACT
It is widely recognized that patterns of antimicrobial use affect prevailing rates of resistance. Less often noted is that, in some cases, resistance will affect prescribing patterns. The first relationship is driven by the biological process of natural selection, the second is driven by physicians' rational behavioral responses to a changing environment. When presented with a patient, a physician must estimate for every available antibiotic the probability that the patient's infection will be cured by the drug. These probabilities will depend on a number of factors, one of which is the rate of resistance to each antibiotic prevailing in the surrounding community. The higher the rate of resistance to a particular antibiotic is, the lower the ex ante probability that it will be effective. All else being equal, the physician will choose the antibiotic associated with the highest probability of cure. Resistance affects physicians' drug choice via these probabilities; once the rate of resistance to a particular antibiotic reaches a critical level, physicians will cease to use it, instead prescribing its closest substitute. The purpose of this chapter is to measure empirically the impact of resistance on physicians' drug choice. Documenting the relationship between resistance and drug choice is important because the drugs to which many pathogens have developed resistance are typically the least expensive antibiotics. Therefore, increasing levels of resistance will increase drug spending.
