The microplate nitrate reductase assay (MNRA) and the rezasurin microtitre assay (REMA) were used for the susceptibility testing of 73 clinical isolates and the results were compared with those that were obtained using the Bactec 460 TB and Bactec MGIT 960 systems. The REMA and the MNRA were performed in 96-well plates. For the REMA, the concentrations of isoniazid (INH) and rifampicin (RIF) ranged from 1.0-0.01 µg/mL and 2.0-0.03 µg/mL, respectively. For the MNRA, the INH concentration was between 1.0-0.03 µg/mL and the RIF concentration was between 2.0-0.06 µg/mL. For the MNRA, the sensitivity, specificity, positive predictive value, negative predictive value and INH/RIF agreement were 100/95.6, 97.6/100, 96.8/100, 100/98 and 98.6/98.6, respectively, and for the REMA, they were 100/91.3, 90.4/100, 88.5/100, 100/96.1 and 94.5/97.2, respectively. Our data suggest that these two rapid, low-cost methods may be inexpensive, alternative assays for the rapid detection of multidrug resistant tuberculosis in low-income countries.

The susceptibility of 49 Mycobacterium tuberculosis clinical isolates to isoniazid (INH) and rifampisin (RIF) (28 multi-drug resistant-tuberculosis samples) was determined by a nitrate reductase assay (NRA) on blood agar. Agreement between the NRA and other testing methods was found to be 93.8% for both INH and RIF. The sensitivity, specificity, positive predictive value and negative predictive value for INH were 92.8%, 94.2%, 86.6% and 97%, respectively. The sensitivity, specificity, positive predictive value and negative predictive value for RIF were 90.4%, 96.4%, 95% and 93.1%. In conclusion, we show here that blood agar can be used effectively for the NRA test.