Discordance without error: why antimicrobial susceptibility testing methods give different results and which result should we report?

Antimicrobial susceptibility testing (AST) methods may yield different results even when testing is technically correct and quality control is acceptable. This narrative review examines why disk diffusion, broth microdilution, gradient diffusion, commercial microdilution, and automated systems can produce discordant MICs, zone diameters, or interpretive categories. Method-dependent differences arise from broth-versus-agar environments, antimicrobial diffusion and stability, incubation conditions, inoculum effects, endpoint detection, concentration ranges, breakpoint proximity, heteroresistance, and delayed or inducible resistance expression. Importantly, method correctness does not guarantee method equivalence or suitability for every organism-antimicrobial combination. Current CLSI, EUCAST, and ISO principles are integrated with comparative evidence to determine which result should guide reporting when valid methods disagree. We propose a nine-step AST Discordance Resolution Framework that prioritizes pair-specific method suitability, quantitative result review, breakpoint proximity, known method limitations, and appropriate reference or confirmatory testing. Discordance should therefore prompt structured evidence-based resolution rather than automatic attribution to analytical laboratory error.