Interlaboratory comparison of drug susceptibility and identification tests for mycobacterium tuberculosis complex
Jackson, K.; Sievers, A.; Dawson, D.; Anargyros, P.; Blacklock, Z.; Chew, W.; Dagnia, H.; Gow, B.; Bettelheim, K.; Dwyer, B.
Australian Journal of Medical Laboratory Science 10(3): 65-69
1989
Accession: 007476292
The five Australian state Mycobacterium Reference Laboratories participated in a comparative study of drug susceptibility and identification tests for Mycobacterium tuberculosis complex. Fifty-five cultures representing 9 M. tuberculosis strains, 1 M. bovis and 1 M. bovis Bacillie Calmette-Guerin (BCG) isolate were tested for drug susceptibility to streptomycin, isolated, ethambutol, rifampicin, thiophen-2-carboxylic acid hydrazide and pyrazinamide by the laboratories' routine methods. The cultures were also tested for nitrate reduction and niacin production. The result that was obtained by the majority of respondents for a particular test/organism combination was called the "correct" result and a discrepancy was therefore a deviation from this. The total discrepancy rate between laboratories for the drug susceptibility tests was 5%, and for the identification tests 7.6%. One strain in particular caused more discrepancies than others. After exclusion of the results of this strain the total discrepancy rate for the drug susceptibility tests was 2.7% and 7.2% for the identification tests. The overall agreement rate between laboratories was 93.7%, or 95% after exclusion of the aberrant strain. The results indicate that there is a high degree of unanimity between Australian laboratories in spite of the wide variety of methods used, and that the results reported are at an acceptable level of reliability and accuracy.