In their recent article, Meyer and colleagues [1] suggested that a 60% increase in ceftazidime use during the previous 2 years was a factor leading to the outbreak of resistant Klebsiella species. However, only 41% of patients from whom these resistant strains were isolated received this agent.

Their experience contrasts with ours, where, during the period from 1989 to 1990, ceftazidime use increased by 20%. Indeed, strains isolated from the first and last month showed a 2% decrease in resistance for Pseudomonas aeruginosa, a 15% decrease for Enterobacter aerogenes, and no change for isolates of Escherichia coli, Serratia marcescens, or Enterobacter cloacae.

The investigators used two simultaneous strategies to decrease further resistant isolates: ceftazidime restriction and barrier precautions for colonized and infected patients. Because only nine randomly selected organisms were investigated by plasmid analysis and not by chromosome analysis or serotyping, we do not know whether an identical strain was being passed from patient to patient, from patient to physician, or from physician to patient during the "outbreak." If so, infection control procedures alone would have aborted the epidemic.

The authors do not state which antibiotic was substituted for ceftazidime for general use during their restriction period or if any new toxicities occurred. They do say that in the last month (April 1990) four times as many approvals were made for treatment with imipenem-cilastatin as for ceftazidime and that the development of imipenem resistance to Acinetobacter strains was the price for changing antibiotic policies. We wonder if this outbreak of 432 isolates could have occurred if proper infection control procedures had been in place.