Treatment-Emergent Cefiderocol Resistance in Critically Ill Patients: A Narrative Review of Incidence, Mechanisms, Detection Pitfalls and Implications for Surveillance and Stewardship

Cefiderocol is among the few agents reliably active against carbapenem-resistant Gram-negative pathogens, and its value depends on durability. Resistance emerging during therapy is documented across pathogens and settings, yet reported incidences span 1% to 40%. The question was last synthesised in 2022; the reviews published since address baseline non-susceptibility, clinical outcomes, testing methodology and resistance development across the novel agents as a class, and none revisits incidence. This review finds that spread is produced by definition, denominator, population, testing method and epidemiological setting rather than by biology: one dataset yields 7%, or between 1% and 2% according to the criterion applied, and the highest figures come from outbreak units where clonal transmission was counted as within-patient evolution. The frequently cited figure of under 2% is what methods of limited sensitivity detect, applied without a standard definition and usually without paired isolates; the true incidence remains unknown. Heteroresistance, present in about half of carbapenem-resistant Acinetobacter baumannii isolates, has not been shown to predict emergence or outcome. Supported measures include reference broth microdilution, storage of the baseline isolate, and repeat testing on triggers during therapy rather than afterwards. A consensus definition is the precondition for everything else.