Four-year prospective study of STAPH-IDENT system and conventional method for reference identification of Staphylococcus, Stomatococcus, and Micrococcus spp.

A 4-year prospective study compared the accuracy of the STAPH-IDENT system (bioMérieux Vitek, Inc., Hazelwood, Mo.) with that of the reference procedure of the Centers for Disease Control and Prevention for the identification of Staphylococcus species, Stomatococcus mucilaginosus, and Micrococcus species. The study compared the results from 1,106 cultures (500 eye cultures, 217 strains submitted for reference identification, and 389 known stock strains) representing 21 species of the family Micrococcaceae. The overall agreement of genus and species identifications was 81.1%. The percent agreement for the five most common clinical isolates was as follows: Staphylococcus epidermidis, 97.1% (517 isolates); Staphylococcus hominis, 82.5% (57 isolates); Staphylococcus aureus, 77.2% (162 isolates); Staphylococcus haemolyticus, 75.8% (61 isolates); and Staphylococcus warneri, 64.1% (39 isolates). The lowest percent agreement was with Staphylococcus cohnii (11.1%; (9 isolates). Of the 217 isolates sent to the Centers for Disease Control and Prevention for identification, 60.4% (131) were correctly identified by the STAPH-IDENT system. Of these, S. epidermidis accounted for 23.9%, S. aureus accounted for 15.6%, S. warneri accounted for 6.9%, Staphylococcus lugdunensis accounted for 6.5%, S. haemolyticus accounted for 5.5%, and S. hominis accounted for 4.1%. The STAPH-IDENT system did not perform adequately when dealing with commonly encountered organisms and is unsuitable for identifying uncommon isolates.