Comparison of three methods for identification of Enterobacteriaceae.

This study compares the ability of three commercial overnight methods, API 20E, Minitek and Enteric-Tek, to accurately and completely identify 368 clinically isolated Enterobacteriaceae without supplemental tests. Organisms included Escherichia coli (54 strains), Shigella spp. (7), Edwardsiella tarda (1), Salmonella enteritidis (10), Citrobacter spp. (30), Klebsiella spp. (55), Enterobacter spp. (68), Hafnia alvei (2), Serratia spp. (33), Proteus spp. (64), Morganella morganii (24), Providencia spp. (18), and Yersinia enterocolitica (2). Methods were those of the manufactures without supplemental tests. API 20E correctly identified 93.2% of strains to species and 3.3% to genus level only, with 3.0% as part of a spectrum of identifications, and 0.5% incorrect identifications. Minitek yielded 96.0% correct identifications to species and 0.5% to genus level only, with 2.5% spectrum identifications, and 1.0% incorrect identifications. Enteric-Tek correctly identified 97.0% of strains to species level with 3.0% spectrum identifications. API 20E identification of some Serratia and Citrobacter strains was to genus level only. Problem organisms for Minitek included Enterobacter agglomerans and Serratia marcescens. A comparison of these three commercial methods shows that all three have the ability to identify most clinically isolated Enterobacteriaceae without supplemental tests.