Clinical evaluation of the microscopic observation drug susceptibility assay for detection of Mycobacterium tuberculosis resistance to isoniazid or rifampin.

This prospective study evaluated the performance of the microscopic observation drug susceptibility (MODS) assay for the direct detection of Mycobacterium tuberculosis drug resistance. MODS assay sensitivity, specificity, and positive and negative predictive values were 96.7% (95% confidence interval [95% CI], 92.1 to 98.8%), 78.4% (95% CI, 73.5 to 80.6%), 82.4% (95% CI, 78.4 to 84.2%), and 95.8% (95% CI, 89.9 to 98.5%), respectively, for isoniazid resistance and 96.0% (95% CI, 90.3 to 98.6%), 82.9% (95% CI, 78.8 to 84.7%), 80.0% (95% CI, 75.2 to 82.1%), and 96.7% (95% CI, 91.9 to 98.8%), respectively, for rifampin resistance. For both rifampin and isoniazid testing, the likelihood ratio for a negative test was < or =0.05, indicating that the MODS assay may be useful for ruling out drug resistance.

Clinical evaluation of the microscopic-observation drug-susceptibility assay for detection of tuberculosis.

BACKGROUND: There is an urgent need for low-cost methods for rapid, accurate detection of Mycobacterium tuberculosis in clinical specimens. The microscopic-observation drug-susceptibility (MODS) assay is a relatively low-cost and simple liquid culture method that has been proposed for use in resource-limited environments. METHODS: This prospective study evaluated the performance of the MODS assay for detection of M. tuberculosis in persons undergoing evaluation for pulmonary tuberculosis in Brazil and Honduras. Respiratory specimens were evaluated using smear microscopy, culture on Lowenstein-Jensen medium, and culture using the MODS assay. A subset of specimens was also cultured using the Mycobacterial Growth Indicator Tube (MGIT) 960 automated system (Becton Dickinson). A study subject was considered to have tuberculosis if at least 1 culture on Lowenstein-Jensen medium was positive for M. tuberculosis. FINDINGS: A total of 1639 respiratory specimens obtained from 854 study subjects were analyzed. On a per-subject basis, MODS sensitivity was 97.5% (95% confidence interval [CI], 95.7-98.6), and specificity was 94.4% (95% CI, 93.1-95.2). Median times to detection were 21 days (interquartile range [IQR], 17-25 days) and 7 days (IQR, 5-10) for culture on Lowenstein-Jensen medium and for the MODS assay, respectively (P<.01). For 64 specimens cultured using the MGIT 960 automated system, median time to growth was similar for the MODS assay (7 days; IQR, 7-10 days) and the MGIT 960 automated system (8 days; IQR, 6-11.5 days; P=.16). The percentage of contaminated cultures was lower for the MODS assay than for culture on Lowenstein-Jensen medium (3.8% vs. 5.8%; P<.01). CONCLUSIONS: The MODS assay is a relatively simple test whose good performance characteristics for detection of pulmonary tuberculosis may make it suitable for resource-limited environments.