Determination of species- and laboratory-related interpretive breakpoints for doxycycline susceptibility testing using single-strain regression analysis.

Abstract

Single-strain Regression Analysis (SRA) was performed for doxycycline on a total of 68 bacterial strains representing 16 different species. Species- and laboratory-related zone diameter breakpoints were determined and compared with histograms of zone diameter values obtained from 942 routine susceptibility tests. Calculated breakpoints were similar within individual species. When considering the homogeneity of susceptibility groups within bacterial species, the calculated breakpoints gave rise to relatively few interpretive errors. In contrast, general breakpoints for doxycycline as recommended by the Swedish Reference Group (SRG) (R less than = 20 mm and S greater than = 26 mm) would give rise to a high proportion of false interpretations in the present laboratory. Fifty-three per cent of H.influenzae strains would have been assigned to the wrong susceptibility group. For E.coli and K.pneumoniae, 22 and 35 per cent, respectively, would have been erroneously categorized using SRG breakpoints. E.cloacae and E.aerogenes would have been assigned another category in 39 and 50 per cent, respectively. The procedure for setting species-specific and laboratory-related interpretive breakpoints is described. Determination of species- and laboratory-related interpretive breakpoints using SRA provides a new approach towards improved accuracy of disc-diffusion susceptibility testing.