ArticleRutherford I, Moody V, Gavan TL, Ayers LW, Taylor DL.
J Clin Microbiol. 1977 Apr;5(4):458-64.
Three separate hospital clinical microbiology laboratories using three different identification systems participated in the identification of Enterobactericeae from a central pool of 'unknown" clinical isolates. With conventional tubed media, API-20E (Anlytab Products Inc.) and R/B tube (Corning Diagnostics) systems, there was a 91.1% agreement in the species designation. No significant differences at the 95% confidence level were found among the systems. Evaluation of individual tests within the systems used revealed lysine decarboxylase of the conventional and citrate of the API-20E system to be significantly different from the same test within the other two systems. The lysine decarboxylase of the conventional system had species relatedness, whereas the differences in citrate of the API-20E system were not related to a particular species. These individual test variations did not affect final organism identification. Reproducibility, evaluated as the system's ability to designate the same identification on two separate occasions, was 92 to 94% for each system. Exact duplication of selected sets of reactions was 60% for conventional, 45% for API-20E, and 61% for R/B. The variations in sets of reactions differed with the system and with the organism involved. The findings suggest equivalency among the three systems in ability to identify common clinical isolates of Enterobacteriaceae and point out the limited usefulness of these systems for biochemical biotyping.