Reinterpretation of Susceptibility Data Using Current NCCLS Breakpoint Criteria

A common problem encountered by those who conduct and publish antimicrobial surveillance studies is that MIC breakpoints for some organism-antimicrobial combinations have been recently updated, which can lead to discrepancies in the published literature. This is particularly so for Streptococcus pneumoniae, for which amoxicillin, amoxicillin-clavulanate, cefepime, cefotaxime, and ceftriaxone MIC interpretive breakpoints defined by the National Committee for Clinical Laboratory Standards (NCCLS) for nonmeningitis isolates have recently been published and are higher than previous breakpoints intended for use in testing both meningitis and nonmeningitis isolates. In 2000, the following distinct NCCLS amoxicillin MIC interpretive breakpoints were published for nonmeningitis isolates: ≤2 μg/ml, susceptible; 4 μg/ml, intermediate; ≥8 μg/ml, resistant (2). At the same time, amoxicillin-clavulanate MIC interpretive breakpoints for nonmeningitis isolates, but not meningitis isolates, increased as follows: the susceptibility breakpoint increased from 1 to ≤2 μg/ml, the intermediate breakpoint increased from 2 to 4 μg/ml, and the resistance breakpoint increased from 4 to ≥8 μg/ml (2). In 2002, cefepime, cefotaxime, and ceftriaxone MIC interpretive breakpoints for nonmeningitis isolates of S. pneumoniae were set as follows: ≤1 μg/ml, susceptible; 2 μg/ml, intermediate; ≥4 μg/ml, resistant (3); these breakpoints are one doubling dilution higher than those for meningitis isolates and breakpoints published in previous years (2, 4). We recently reported data for 4,940 isolates of S. pneumoniae (98.9% nonmeningitis isolates, 1.1% meningitis isolates) from a surveillance study undertaken in the United States during 1999 (1) in which susceptibility data for these drugs were interpreted by using both breakpoints in existence at the time of the study and more-current, revised breakpoint standards (2, 3, 4). However, data presented in the abstract of reference 1 utilized only 1999 MIC interpretive breakpoints (4) in place at the time of the study, which, in the context of new revised breakpoint standards for these compounds, reported falsely high rates of resistance to amoxicillin-clavulanate (12.2%) and ceftriaxone (3.6%). We wish to clarify this. Reinterpretation using current NCCLS-defined breakpoint standards for these drugs (3) show rates of resistance to all compounds tested as follows: penicillin, 16.2%; amoxicillin-clavulanate, 2.0%; cefuroxime, 28.1%; ceftriaxone, 1.5%; trimethoprim-sulfamethoxazole, 30.3%; azithromycin, 21.4%; levofloxacin, 0.6%; moxifloxacin, 0.1%.

In this time of increased resistance, perceptions of falsely high resistance rates can further limit the choice of available agents, encouraging the use of inappropriate drugs. To avoid this irrespective of when the studies are undertaken, interpretive criteria defined at the time of publication should always be used.