LAUSR

Abstract Background CRE are a world-wide public health challenge with extremely limited treatment options. Aminoglycosides have variable susceptibility against these organisms. At our institution, amikacin has been active against these isolates and useful as part of a combination regimen for CRE treatment. In this study, we compared the susceptibility results for 3 aminoglycosides between an automated susceptibility system (Phoenix) and broth microdilution (BMD). Methods Gentamicin, tobramycin, and amikacin susceptibility were determined in our academic medical center microbiology laboratory using an automated susceptibility system (Phoenix) and broth microdilution (BMD) according to CLSI guidelines against 120 recent CRE clinical isolates. Categorical agreement was defined between methods as classification of isolates in the same susceptibility category using CLSI breakpoints. Minor, major and very major error rates were calculated for each aminoglycoside. Results The primary CRE was K. pneumoniae (46%), followed by Enterobacter spp. (32%), and E. coli (6%). The categorical agreement ranged 58% (gentamicin) to 68% (tobramycin). Automated susceptibility system provided significantly higher susceptibility from 14% (gentamicin) to 30% (tobramycin and amikacin). Aminoglycoside Broth Microdilution Automated Susceptibility System %S MIC50 MIC90 MIC range %S MIC50 MIC90 MIC range Amikacin 63% 16 32 1 to >512 93% ≤8 16 ≤8 to >32 Tobramycin 12% 16 128 0.25 to >128 42% >8 >8 ≤2 to >8 Gentamicin 17% 16 128 0.25 to >128 31% >8 >8 ≤2 to >8 Aminoglycoside Categorical agreement Major error rate Very major error rate Minor error rate Amikacin 62.5% 0.8% 5.8% 30.8% Tobramycin 67.5% 0.8% 9.2% 22.5% Gentamicin 58.3% 4.2% 25.8% 11.7% Conclusion Automated susceptibility system over predicts the true susceptibility of CRE against all 3 aminoglycosides. This could be a major impact on the potential utility of the aminoglycosides especially amikacin for CRE infections. Disclosures All authors: No reported disclosures.