Sensitivity and Specificity of TB Gene Xpert
The Xpert MTB/RIF test demonstrates a pooled sensitivity of 89% and specificity of 99% when used as an initial diagnostic test for pulmonary tuberculosis in adults, substantially outperforming traditional sputum smear microscopy. 1
Overall Diagnostic Performance
For pulmonary TB detection as an initial test:
- Pooled sensitivity: 89% (95% CrI 85-92%) 2, 1
- Pooled specificity: 99% (95% CrI 98-99%) 2, 1
- The test increases TB detection among culture-confirmed cases by 23% compared to smear microscopy alone 2, 1
Performance by Clinical Context
In smear-positive, culture-positive patients:
- Sensitivity reaches 98%, making a negative Xpert result highly reliable with false-negative results occurring only 4% of the time 1, 3
In smear-negative patients (as an add-on test):
- Sensitivity drops substantially to 67% (95% CrI 60-74%) 3
- Specificity remains high at 99% 3
- Clinical suspicion remains critical in this population, as false-negative results are more common in paucibacillary disease 1
In people living with HIV:
- Pooled sensitivity is lower at 79% (95% CrI 70-86%) 2, 1
- This represents a 10% reduction compared to HIV-negative patients (86% sensitivity) 3
Rifampicin Resistance Detection
For detecting rifampicin resistance as a surrogate for MDR-TB:
- Pooled sensitivity: 95-96% (95% CrI 90-97%) 1, 3
- Pooled specificity: 98% (95% CrI 97-99%) 1, 3
- The positive predictive value may be lower in populations with low prevalence of drug resistance 1
Clinical Translation
In a hypothetical cohort of 1,000 patients where 10% have TB:
- Xpert will correctly diagnose 88 cases and miss 12 cases 3
- Smear microscopy will diagnose only 65 cases and miss 35 cases 3
- This represents a 35% improvement in case detection 3
Important Caveats and Limitations
False-negative results are more likely in:
- Paucibacillary disease (minimal bacterial burden) 1, 4
- Smear-negative cases, where sensitivity drops to 28-67% 5, 4, 3
- Patients with minimal or early disease presentation 4
Setting-specific performance:
- In low-incidence, high-resource settings with less extensive disease, sensitivity may be as low as 46% overall and 28% in smear-negative cases 4
- The test performs optimally in high-burden settings with more advanced disease 4
Nontuberculous mycobacteria (NTM):
- Xpert was positive in only 1 of 180 specimens with NTM, demonstrating excellent specificity for M. tuberculosis 3
Implementation Recommendations
Xpert MTB/RIF can replace smear microscopy as the initial diagnostic test in both low and high TB prevalence settings 2, 1
Despite higher costs compared to microscopy, the 23% increase in case detection justifies implementation as part of comprehensive TB control programs 2, 1
In smear-negative patients with high clinical suspicion, follow-up with culture and clinical monitoring remains essential while awaiting definitive results 5