Proficiency testing of conventional drug susceptibility tests of Mycobacterium tuberculosis

Proficiency testing of indirect drug susceptibility tests of Mycobacterium tuberculosis was begun in 1985 by the Laboratory Centre for Disease Control (LCDC) with the participation of Provincial Public Health Laboratories in Canada. Comparable sets of 60 cultures of Mycobacterium tuberculosis representing 30 strains were distributed by LCDC to the participating laboratories to be tested for drug susceptibility against isoniazid, streptomycin, rifampin, and ethambutol using conventional methodologies. Intralaboratory agreement values determined by comparing results obtained on sets of duplicate cultures were high and were found to vary little from drug to drug and from laboratory to laboratory. Interlaboratory agreement was determined by comparing results reported by participating laboratories to those obtained by the Reference Laboratory. Agreement percentages were found to be lower for drug-resistant cultures than for drug-susceptible cultures. The reliability of drug susceptibility testing results was higher for isoniazid and rifampin, than for ethambutol and streptomycin. This study shows that the higher subsidiary drug concentrations do not compare well with main drug concentrations, especially in the case of streptomycin and ethambutol. The significance of the higher subsidiary concentrations in in vitro susceptibility testing is therefore in need of clarification. The proficiency testing results obtained in this study compare favorably with those reported in other developed countries despite the fact that a variety of testing procedures are used throughout the country.