Comparison of BBL Crystal ANR ID Kit and API rapid ID 32 A for identification of anaerobic bacteria.

BBL Crystal ANR ID Kit and the API System rapid ID 32 A are miniaturized identification systems for anaerobes using enzymatic tests. The incubation period of both systems is 4 hours. A comparative evaluation of the BBL Crystal Identification System Anaerobe ID Kit (Becton Dickinson Microbiology Systems, Cockeysville, USA) with anaerobes grown on Columbia and Schaedler agar plates (Becton Dickinson Microbiology Systems, Cockeysville, USA) and the API System rapid ID 32 A (BioMérieux SA, Lyon, France) with bacteria grown on Columbia agar (Becton Dickinson Microbiology Systems, Cockeysville, USA) which is recommended by the manufacturer as cultivation medium, was performed with 207 mostly fresh clinical anaerobe isolates, including 104 gram-negative bacilli, 12 gram-negative cocci, 15 gram-positive cocci, 14 gram-positive sporeforming bacilli and 62 representatives of gram-positive non-sporeforming bacilli. With supplemental testing the Crystal system with inocula from Columbia and Schaedler agar and API inoculates from Columbia agar identified to genus level 144 (69.6%), 152 (73.4%) and 109 (52.7%) isolates, respectively. Misidentification to genus level was found by Crystal from Columbia and Schaedler agar and by API from Columbia agar in 17 (8.2%), 15 (7.3%) and 12 (5.8%) isolates, respectively. 36 isolates were not determined to species level by classical anaerobic methods or the systems only identified to genus level. 26 anaerobes were not included in the database of the Crystal or API system. From the remaining 145 clinical isolates with supplemental testing, Crystal from Columbia and Schaedler agar plates correctly identified 91 (62.8%) and 102 (70.3%), respectively, and API, 69 (47.6%) isolates. For the correct identification to genus and species level of the 207 clinical isolates tested, the Crystal system from Columbia and Schaedler agar and API system from Columbia agar required supplemental testing, as specified by the manufacturer, for 39 (27.1%), 34 (22.4%) and 14 (12.8%) isolates, respectively. Among the 207 clinical isolates tested, 27 had been frozen and 26 had been lyophilized. In a comparative evaluation, the fresh isolates showed a slightly to significantly better identification rate than the frozen or lyophilized specimens in all three tests. The individual reproducibility of the Crystal ANR ID, which had been tested before the accuracy study was performed, ranged from 90.8% to 100%. The overall reproducibility was determined to be 97.3%. Time consumption studies and cost analysis did not show a significant difference between both systems, but Crystal ANR ID was found to be easier to use than API rapid ID 32.