Giampaglia, C.M.S.; Martins, M.C.; de Oliveira Vieira, G.B.; Vinhas, S.A.; da Silva Telles, M.A.; Palaci, M.; Marsico, A.G.; Hadad, D.J.; Mello, F.C.Q.; de Souza Fonseca, L.; Kritski, A.
SETTING: Three mycobacteria reference laboratories in the south-eastern part of Brazil.
OBJECTIVE: To evaluate the automated Mycobacteria Growth Indicator Tube (MGIT) for drug susceptibility testing of Mycobacterium tuberculosis.
DESIGN: Performance of the automated BACTEC MGIT 960 (M960) system for testing M. tuberculosis susceptibility to streptomycin (SM), isoniazid (INH), rifampicin (RMP) and ethambutol (EMB) was evaluated with 95 clinical isolates and compared to the results of the radiometric BACTEC 460TB (B460) system, the proportion method (PM), and the resistance ratio method (RRM). Judicial susceptibility profiles of 88 isolates were defined based on two or more concordant results among B460, PM and RRM, and used as a reference for comparison with M960 results.
RESULTS: Agreement rates between M960 and conventional methods were 95.2% with B460, 96.6% with the PM and 93.4% with the RRM. The lowest agreement rates were obtained for SM with the RRM and for EMB with B460. When comparing M960 with judicial susceptibility profiles, the agreement rate was 97.9%. The agreement rates obtained for INH and RMP were 99.2% and for SM and EMB they were 96.2% and 96.9%, respectively. The mean time to reporting the M960 results was 6.9 days.
CONCLUSION: M960 offers great improvements when compared to the proportion and resistance ratio methods and would benefit patient treatment.
Keywords: Mycobacterium tuberculosis; susceptibility; MGIT 960; proportion method; resistance ratio method; BACTEC 460