Potential confusion regarding the term 'resistance' in epidemiological surveys

Sir, We have previously pointed out the need to avoid confusion between a clinical breakpoint (CBP) and an epidemiological cut-off value (ECV) and that the term ‘breakpoint’ should be used to indicate clinical resistance.1 The terms (CBP and ECV) have been clearly defined by Kahlmeter et al.2 and by EUCAST (http://www.srga.org/Eucastwt/eucastdefinitions.htm), which also states that ‘clinical breakpoints should be used in everyday clinical laboratory work to advise on therapy in the patient’ whereas ECV ‘can be used as the most sensitive measure of resistance development for measuring resistance development in hospitals and the community…’ (http://www.srga.org/eucastwt/MICTAB/index.html). Arguably, the term ‘resistance development’ is unfortunate and misleading in this context and would better be described as ‘emerging reduced susceptibility’. Organisms whose MICs exceed the ECV show reduced susceptibility compared with the wild-type population and may yet respond to clinical treatment as their MIC may lie below the CBP. Turnidge and Paterson,3 in their review of breakpoint setting, posed the question ‘Should strains possessing an acquired resistance mechanism but having MICs below the PK/PD breakpoint be considered resistant and reported as such, regardless of the MIC?’. Using the term ‘resistant’, in this context, can indeed be ambiguous or downright confusing if applied to both those organisms exceeding the ECV and to those organisms which exceed the CBP. This becomes crucial in antibiotic susceptibility monitoring programmes that are of great value both to clinicians and to epidemiologists. Clinicians are interested in whether empirical treatment will result in a favourable outcome and as such they need to know the percentage of clinically resistant organisms. The epidemiologist, in contrast, is looking for deviation in the susceptibility profile from that of a wild-type population, i.e. early emergence of a population with reduced susceptibility based on ECVs. Both are valid, although the clinician's requirement is perhaps the more pressing, as therapy represents the final objective for antimicrobial use. The potential for confusion is illustrated by a recent report from The Netherlands.4 The CBP criterion for ‘resistance’ used in the 2002–04 MARAN reports has been replaced with ECV in the recent 2005 report. As the authors state: ‘This [use of ECV instead of CBP] may result in a change in resistance percentages in comparison with previous years’. Although trends were indeed retrospectively analysed by applying ECV to data from earlier years, nevertheless, a clinician consulting the report could be forgiven for assuming that the ‘resistant’ percentages referred to clinical resistance. Susceptibility data should primarily be described on the basis of CBPs (where possible harmonized internationally2), as clinical efficacy is the objective for the clinician as end-user. However, there is also a need to recognize and respond to those changes in susceptibility that are reflected by the ECV and which may not be revealed by the use of CBP. To this end, we propose that future surveillance reports should summarize susceptibility data with separate reference to both CBP (for clinical resistance) and ECV (for reduced susceptibility). The term ‘resistant’ should, however, be limited to the former. None to declare.