Institution-specific versus published in vitro antimicrobial susceptibility data in making formulary decisions.

The results of susceptibility testing of 549 isolates of gram-negative organisms to 17 antimicrobial agents were compared with published reports of the sensitivity of those organisms to those agents. All gram-negative bacilli isolated from cultures obtained from hospitalized patients during a three-month period were preserved for antimicrobial sensitivity testing. Standard Kirby-Bauer disk diffusion susceptibility tests were performed using 17 broad-spectrum antimicrobial agents that either were included in the hospital formulary or were being considered for inclusion. Organisms were recorded as being sensitive or resistant to each drug, and the results were compared with the published results of in vitro sensitivity studies. When the results of actual antimicrobial sensitivity testing varied from published results, the discordant results were assigned a ranking of 1 to 4 based on the percentage difference. In 34 of 77 drug-organism pairs tested, the results of susceptibility testing differed by more than 10% cumulative susceptibility from published values; 26 of these represented instances in which the results of actual testing were at least 10% less than published values. For seven of the antimicrobial agents that were being considered for inclusion in the hospital formulary, results indicating unexpectedly suboptimal activity against institutional pathogens were a determinant in eliminating the agents from further consideration. In vitro testing of antimicrobial susceptibility of local pathogens can be a better method of predicting the susceptibility of such pathogens to new antimicrobial agents than relying on published susceptibility data. Pharmacy and therapeutics committees should consider testing prevalent institutional pathogens for susceptibility to all antimicrobial agents that are proposed additions to the formulary.