Individuals with cystic fibrosis may have a variety of nonfermentative gram-negative (NFGN) bacteria isolated from their sputum. The failure to accurately identify Burkholderia cepacia among these, or to misidentify other NFGN species as B. cepacia, can have grave medical, epidemiological, and psychological consequences. In this study, 66 isolates of multiply antibiotic resistant NFGN bacteria cultured from sputum from 66 cystic fibrosis patients were received from 34 clinical microbiology laboratories (referring labs). The majority (74%) were identified by the referring labs as B. cepacia, while the remaining isolates were either identified as belonging to one of 7 other NFGN species or not identified. All isolates were blindly re-analyzed (a) phenotypically by biochemical profile and growth on a novel B. cepacia selective medium, and (b) genotypically by using B. cepacia species-specific polymerase chain reaction. Species identification by these means disagreed with the referring labs' identification for 17 (26%) of the 66 isolates. For 9 of these 17, identification by both phenotype and genotype disagreed with the referring labs' identification; these were considered definitely misidentified by the referring lab. For the other 8 isolates, identification by either phenotype or genotype differed from the referring labs' identification; these were considered possibly misidentified. Of the 49 isolates identified as B. cepacia by the referring labs, 5(10%) were definitely and an additional 7 were possibly misidentified. Conversely, our analysis identified 2, and possibly 3,B. cepacia isolates that were either not identified or identified as other NFGN species by the referring labs. These data demonstrate that current clinical microbiology laboratory technologies may not be capable of accurately identifying B. cepacia and that taxonomic ambiguities may confound identification of some strains.
