Disc testing of meningococci.

Abstract

In a recent article (1), Block et al. compared the disc diffusion test with MIC determination by the E test for studying penicillin susceptibility of Neisseria meningitidis, concluding that the disc diffusion test was unreliable. Although the authors used the definitions of reduced susceptibility proposed by us (2, 3), it is not clear whether they also followed our disc susceptibility procedure; if this procedure is not folllowed, our zone diameter breakpoints cannot be applied. They mention that the “disc diffusion testing was performed by using 2-U penicillin and 1-μg oxacillin discs … according to National Committee for Clinical Laboratory Standards recommendations” (1). To the best of our knowledge, the National Committee for Clinical Laboratory Standards (NCCLS) has not recommended such a procedure. To support their conclusions, the authors refer to a previous study (8) in which our breakpoints, generated in Mueller-Hinton agar (MH) plus 5% blood, were found to be of little use. They do not mention that in that previous study, another growth medium was used: chocolate agar with a GC base (8), a much richer medium for which we never proposed breakpoints. In our opinion, the data reported by Block et al. (1) lack both internal and external validation. The authors apparently did not use internal quality control procedures, which are mandatory, such as reference strains, blinded determinations by several independent observers, or checking internal reproducibility of the E test and the disc test. The E test is a commercial method, not a “gold standard,” for MIC determination; strains for which MICs are between 0.064 and 0.094 μg/ml may already be low-level resistant. The rare zone diameter distribution Block et al. observed could be explained by inconsistency of the low-charge disc potency, but this factor was neither taken into account nor ruled out. No attempt was made to study ampicillin susceptibility, although the oxacillin disc was clearly described as detecting both the Penr Ampr and the Pens Ampr phenotypes (3). None of the strains, particularly the ones giving conflicting results, were sent to a reference laboratory to obtain an authoritative external validation. Distinguishing between penicillin-susceptible and low-level penicillin-resistant strains can be very difficult, even for experienced laboratories (5); the NCCLS had to modify their initial criteria for determining MICs for meningococci by recommending addition of 5% blood to MH (7). We obtained consistent results by combining susceptibility data, penicillin PBP2 affinity data, and data on penA gene modifications (4, 6, 9).