RESUMO

A collaborative study involving seven laboratories was undertaken to evaluate the reproducibility and the reliability of the broth disk elution test against anaerobic bacteria by comparing with the reference agar dilution method. A two breakpoint broth test was also evaluated. Assays were performed using the same testing conditions (i.e. medium, temperature, atmosphere and incubation time). One hundred Gram-negative and Gram-positive clinical isolates were initially studied. Overall agreement of 98.5% and 97.5%, were found for disk elution and the two breakpoint tests, respectively. In order to assess the reliability of the disk elution test, two different lots (LOT1 and LOT2) of disks of piperacillin and clindamycin were selected, to obtain two final concentrations after dilution (10 and 60 mg/mL; 1 and 4 mg/mL, respectively). Two hundred and eighty assays were performed against one strain of both Bacteroides fragilis(piperacillin MIC, 8.0 mg/mL; clindamycin MIC, <0.5 mg/mL) and Bacteroides thetaiotaomicron(piperacillin MIC, 16.0 mg/mL; clindamycin MIC, <0.5 mg/mL). With LOT 1, considering both species and both antibiotics, the agreement among six laboratories ranged from 85% to 100% (P > 0.05) with the higher concentration. Overall agreement among all laboratories was 91%. No optimal agreement (>90%) for clindamycin-Bacteroides thetaiotaomicron using the LOT1 (77%) was found. Since this finding was not observed with LOT2 (100% agreement), discrepancies were attributed to variation between lots. Overall agreement with LOT2 was 100% for all centres. The present study indicates that the broth disk elution method proved to be a reliable and suitable alternative for routine susceptibility testing for anaerobic bacteria, as a resistance screening method for clinical purposes.