This project's principal focus has been to investigate the in vitro susceptibility of ampicillin-resistant Escherichia coli to amoxicillin/clavulanate and ampicillin/sulbactam. These two antibiotic-beta lactamase inhibitor combinations have been stated to be comparable in their clinical efficacy and therefore should be comparable in their in vitro susceptibility results. A lack of correlation of routine susceptibility test results between these two agents was noted in clinical isolates of ampicillin-resistant E. coli from National Institutes of Health patients. The reproducibility of the results was assessed by repeating the sensitivity test procedure on the isolates and the validity was further investigated by performing susceptibility testing by other methods. Susceptibility results of 182 ampicillin-resistant isolates revealed that amoxicillin/ clavulanate and ampicillin/sulbactam had a low degree of correlation by the initial microdilution susceptibility testing procedure. This discrepancy was confirmed by retesting representative isolates, and the validity of the discrepancy was confirmed by another susceptibility test procedure. Of the isolates, 75.3 percent were amoxicillin/clavulanate sensitive and ampicillin/sulbactam intermediate or resistant. Plasmid fingerprinting was performed to demonstrate that the E. coli isolates represented different strains; therefore, the discrepant results were not limited to one or a few unusual E. coli isolates. Given the results of clinical trials that suggest the in vivo equivalence of both agents, it seems probable that the discrepancies observed are a consequence of inappropriate break points established by the National Committee for Clinical Laboratory Standards. The assessment of all ampicillin-resistant E. coli isolated during a 1- year period is in progress.
