Abstract
Accurate assessment of antibiotic susceptibility is critical for treatment of antimicrobial resistant (AMR) infections. Here, we examine whether antimicrobial susceptibility testing in media more physiologically representative of in vivo conditions improves prediction of clinical outcome relative to standard bacteriologic medium. This analysis reveals that ∼15% of minimum inhibitory concentration (MIC) values obtained in physiologic media predicted a change in susceptibility that crossed a clinical breakpoint used to categorize patient isolates as susceptible or resistant. The activities of antibiotics having discrepant results in different media were evaluated in murine sepsis models. Testing in cell culture medium improves the accuracy by which MIC assays predict in vivo efficacy. This analysis identifies several antibiotics for treatment of AMR infections that standard testing failed to identify and those that are ineffective despite indicated use by standard testing. Methods with increased diagnostic accuracy mitigate the AMR crisis via utilizing existing agents and optimizing drug discovery.