Schroeder K, Küchler R
Forschungsinstitut Borstel, Germany.
Zentralbl Bakteriol. 1992 Dec;277(4):454-66. doi: 10.1016/s0934-8840(11)80470-6.
The frequency of isolation of environmental mycobacteria from clinical specimens has increased significantly in the course of the last years. Assessment of the medical importance of such isolates always depends on a correct identification of any of these strains. In a multicentric study, the efficacy of a standardized, computer-aided system for identification has been evaluated. 13 Laboratories received 24 strains of mycobacteria each and were asked to test all these strains for 16 different features. The results were coded in a standardized way. For the evaluation of these tests, the laboratories were supplied with a list containing all possible codes, which might be indicative of one or another identification. The results proved the logic of the approach. However, the application of the system lead to a correct identification in only 74% of all attempts. It is concluded that such a standardized approach is not useful for mycobacteria, because it requires too much working time and material. It will be more practical to concentrate on a smaller number of tests in the peripheral laboratories, which might allow the identification of M. tuberculosis and a few other species and to leave the definitive identification to reference laboratories. The development of faster, simpler and more reliable methods remains highly desirable.