Comparative evaluation of the Eiken and API 20E systems and conventional methods for identification of members of the family Enterobacteriaceae

To evaluate the accuracy and utility of the Eiken Systek No. 1 (Eiken system; Eiken Chemical Co., Ltd., Tokyo, Japan), we conducted a clinical comparison, with 345 Enterobacteriaceae isolates, of the Eiken System with API 20E (Analytab Products, Inc., Plainview, N.Y.) and conventional methods. The Eiken system is a 21-biochemical-test battery tray stored at 25 degrees C and inoculated in one step. It is similar to the API 20E except that the Eiken system contains malonate, adonitol, and maltose; lacks gelatin, sucrose, melibiose, amygdalin, and arabinose; and uses reagent strips instead of liquid reagents. The API 20E and Eiken systems correctly identified 339 (97.7%) and 276 (79.5%), respectively, and misidentified 3 (0.9%) and 13 (3.7%), respectively, of the isolates. There were no identification codes for 5 (1.4%) organisms with the API 20E and 58 (16.7%) organisms with the Eiken system; of these latter unidentified organisms, 42 were identified as Proteus spp., Morganella sp., and Providencia rettgeri by conventional methods. There was no significant difference between the two rapid systems in total time required for inoculation and reading. Modifications for interpretation of decarboxylase and oxidase tests were needed for the Eiken system, and manipulation of reagent strips required considerable dexterity. However, the Eiken system was easier to inoculate than the API 20E, and, with minor increases in the data base to include more of the Proteus and Morganella spp. and P. rettgeri, the system should be reliable for identification of members of the family Enterobacteriaceae.