190 Correspondence

growth. For example, subinhibitory antibiotic concentrations can affect bacterial mor-phology, ultrastructure, physiology, metab-olism and virulence (Zhanel, Hoban & Harding, 1992). However, the existing defi-nition of PAE simply represents the return of the organism to wild-type growth rates. It should be stated that the classical definition of the PAE is bacterial growth suppression following complete removal of extracellular antibiotic. It was the intention of our study to use the classical definition of PAE and thus to assess the effect on bacterial growth. Our purpose was not to study the bacterial effects of residual or subinhibitory antibiotic concen-trations, and thus study antibiotic induced bacterial effects on morphology, ultrastructure, physiology, metabolism and virulence. We are not discounting the critical importance that subinhibitory antibiotic concentrations con-tribute in affecting the myriad of bacterial physiological processes. Finally, it should be pointed out that the growing study of antibiotic pharmacodynamics involves rate and extent of bacterial killing, effects of subinhibitory antibiotic concentrations, and various other factors of which the PAE is only one important factor