[Results of two methods for testing bacterial susceptibility to antibiotics: disk method and a new semi-automatic method in liquid medium (ABACR procedure) and comparison with the minimal inhibitory concentration (author's transl)].

A performance analysis was established between three methods for testing of bacterial susceptibility to sixteen antibiotics: agar dilution or minimal inhibitory concentration (MIC) method, agar diffusion or DISK method, and semi-automatic dilution in liquid medium or ABACR method. Precision of the three methods was determined using three reference strains in repetition experiments (9 repetitions for MIC, 60 repetitions for both DISK and ABAC). The results show that MIC method was the most precise, and that the precision of DISK and ABAC methods was not sifnificantly different. Fidelity of both DISK and ABAC methods was compared in taking as reference the MIC of 200 wild strains of Gram-negative bacteria. Omitting the data concerning trimethoprim-sulfamethozoaxol, the results were found concordant between DISK and MIC methods in 68.6 percent of the cases, between ABAC and MIC methods in 73.6 percent, and between ABAC and DISK methods in 68.9 percent. With regard to MIC, ABAC method gave results slightly better than those of DISK method, especially for five antibiotics: gentamicin, minocyclin, doxycylin, chloramphenicol and polymyxin B. So, the ABAC method looks very interesting for a semi-automatic routine method in testing bacterial susceptibility to antibiotics.