erm gene distribution among Norwegian Bacteroides isolates and evaluation of phenotypic tests to detect inducible clindamycin resistance in Bacteroides species.

The aims of this study were to describe the distribution of the most common erm genes in a collection of Norwegian Bacteroides isolates and to investigate whether the phenotypic tests for determining inducible clindamycin resistance among Bacteroides species recommended by EUCAST, NordicAST and the manufacturer of E-test, are effective. We investigated 175 unique Bacteroides isolates for the presence of erm(B), erm(F) and erm(G) genes, determined their minimum inhibitory concentrations (MICs) to clindamycin and categorised their susceptibility according to EUCAST breakpoints. 27 isolates were resistant to clindamycin. Furthermore, we investigated whether these recommended methods could detect inducible resistance in the Bacteroides isolates: 1) EUCAST recommendation: Dissociated resistance to erythromycin (clindamycin susceptible with erythromycin MIC > 32 mg/L), 2) NordicAST recommendation: Double disk diffusion test (DDD) or 3) Manufacturer of E-test's recommendation: prolonged incubation of clindamycin E-test for 48 h. erm genes were detected in 30 (17%, 95% CI 12%-23%) of 175 Bacteroides isolates with erm(F) as the dominating gene. There were six (4%, 95% CI 1%-7%) of 148 clindamycin susceptible isolates harbouring erm genes, they were considered inducibly resistant to clindamycin. None of the methods for phenotypic detection of inducible clindamycin resistance performed satisfactory with sensitivities of 33%, 17% and 0% and specificities of 90%, 99% and 97% for dissociated resistance, DDD and prolonged incubation of clindamycin E-test, respectively. In our view, the scientific basis for investigating every Bacteroides isolate for inducible resistance to clindamycin is weak. Molecular detection of erm genes may prove a better option than the phenotypic methods we evaluated.