Effect of the source of Mueller-Hinton agar and resistance frequency on the detection of methicillin-resistant Staphylococcus aureus

Inconsistencies in the results of disk diffusion tests of oxacillin against Staphylococcus aureus that occurred when using commercially prepared Mueller-Hinton agar from different sources led us to evaluate the ability of media from different sources to detect resistance to oxacillin, methicillin, and nafcillin in S. aureus. Mueller-Hinton agar from five manufacturers was prepared in our laboratory and used for standard disk diffusion and agar dilution tests. Ten oxacillin-resistant S. aureus isolates, of which three were definitive-resistant and seven were occult-resistant, were examined. All definitive-resistant strains were resistant to all three antimicrobial agents on four out of five agars. The occult-resistant strains were consistently detected as resistant on only one of the agars. With only slight differences, oxacillin, methicillin, and nafcillin resistance was more readily detected by disk diffusion and agar dilution when initially incubated at 30 degrees C, and extended incubation improved the detection. The frequency of resistance within a population of occult-resistant cells was low compared with the frequency within a population of definitively resistant cells. The heterogeneity of colony morphology and apparent growth rates within a population of occult-resistant cells contributed to the problem of detecting some resistant isolates. Definitive-resistant isolates were characterized by a very high and stable frequency of resistance. Occult-resistant strains were characterized by a lower frequency of resistance, although the true frequency of resistance may be difficult to ascertain because of heterogeneity in growth rates.