The use of PD to suppress the development of resistance is a key element. This can be achieved in two ways; by properly dosing the current drugs to maintain their utility for a long period of time and by developing new drugs at a dose and schedule that helps reduce emergence of resistance. Another key element of future PD use is optimization of patient outcome. This requires measurement of drug concentrations, and although logistically difficult, it has been used recently when modeling levofloxacin clearance to reliably identify patients at a high probability for having high drug clearance, and therefore at highest risk for failure. In such cases, an additional agent can be added to improve their likelihood of a good outcome. We hope to use PD data to choose a dose that is optimal for a large percentage of the population, without giving a toxic event, and then in individual patients, use modeling techniques to further optimize outcome for the individual.