Performance evaluation of antimicrobial susceptibility testing devices across CLSI, EUCAST, ISO, FDA, and EMA frameworks: implications for breakpoint application

ABSTRACT Broth microdilution (BMD) remains the reference standard for determining minimum inhibitory concentrations (MICs) in antimicrobial susceptibility testing (AST). However, most routine laboratories rely on surrogate methods, including disk diffusion and commercial MIC-generating systems, whose results must correlate with those of reference BMD prior to clinical use. Evaluation criteria differ across standards development organizations (SDOs) and regulatory authorities, potentially affecting clinical interpretation, particularly in low- and middle-income countries (LMICs). This mini-review compares evaluation frameworks from major SDOs—the Clinical and Laboratory Standards Institute (CLSI), the European Committee on Antimicrobial Susceptibility Testing (EUCAST), and the International Organization for Standardization (ISO)—and regulatory frameworks, including the U.S. Food and Drug Administration (FDA) and the European In Vitro Diagnostic Regulation (IVDR). Key performance metrics, including essential agreement, categorical agreement, error rates, reproducibility, and bias, were analyzed, alongside differences in breakpoint setting and device approval pathways. Implications for clinical microbiology and antimicrobial stewardship were considered, with explicit emphasis on challenges faced in LMIC settings. Substantial heterogeneity exists across frameworks. CLSI and FDA emphasize breakpoint-dependent categorical agreement and error rates, whereas ISO prioritizes analytical measures such as essential agreement and bias without categorical errors. EUCAST broadly aligns with ISO analytically while separately maintaining responsibility for breakpoint setting. These differences can yield discordant device assessments, especially after breakpoint revisions. Surrogate MICs often correlate with BMD near breakpoints but may show systematic bias across MIC ranges, not fully captured by essential agreement alone, with important implications for LMIC laboratories. Divergent frameworks influence susceptibility interpretation, resistance surveillance, and antimicrobial stewardship, particularly in LMICs. Greater harmonization and awareness of methodological limitations are essential to ensure reliable AST reporting and support global antimicrobial resistance efforts across LMIC healthcare systems.