Objectives: The objective of this multicentre study was to define the accuracy and reproducibility of the NCCLS disc diffusion method for Haemophilus influenzae against ampicillin and co-amoxiclav in Finnish clinical microbiology laboratories. Special attention was paid to the ability of the laboratories to detect beta-lactamase-negative ampicillin-resistant (BLNAR) strains.
Methods: Three BLNAR and two beta-lactamase-negative ampicillin-susceptible isolates (BLNAS)-originating from the American Type Culture Collection (ATCC) and UK National External Quality Assessment (UKNEQAS) schemes-were included in this study. Susceptibility tests for these isolates were performed in 26 clinical microbiology laboratories, in accordance with NCCLS guidelines. Additionally, low-strength discs for ampicillin (2 microg) and co-amoxiclav (3 microg) were tested.
Results: The low-strength discs for ampicillin and co-amoxiclav categorized more accurately BLNAR and BLNAS H. influenzae isolates than did the high-strength discs recommended by the NCCLS. In addition, the high-strength discs produced more major errors than the low-strength discs (22 versus six for ampicillin and 40 versus seven for co-amoxiclav). Great variation occurred in the method regardless of the antibiotic concentration of the discs.
Conclusions: The use of low-content ampicillin and co-amoxiclav discs is recommended for the susceptibility testing of H. influenzae. Interpretative criteria of S > or = 17 mm and R < or = 13 mm for both discs are suggested.