What are the sensitivity and specificity of GeneXpert (Xpert Mycobacterium tuberculosis/Rifampicin) for detecting Mycobacterium tuberculosis and rifampicin resistance in sputum, broncho‑alveolar lavage, cerebrospinal fluid, urine, stool, and tissue specimens?

GeneXpert MTB/RIF Sensitivity and Specificity Across Different Specimen Types

For pulmonary TB, GeneXpert demonstrates excellent performance on sputum (sensitivity 85-96%, specificity 98-99%), but performance drops substantially for extrapulmonary specimens, with sensitivity ranging from 33% for non-respiratory samples to 60-89% for specific extrapulmonary sites. 1, 2, 3

Respiratory Specimens

Sputum (Standard Specimen)

• AFB smear-positive sputum: Sensitivity 96-100%, specificity 85-99% 1, 4, 3
• AFB smear-negative sputum: Sensitivity 66-77%, specificity 80-98% 1, 5, 3
• Overall pooled sensitivity: 85-95%, specificity 98% 1, 4, 2
• Sputum provides the highest diagnostic yield and should be the preferred specimen type 2

Other Respiratory Specimens

• Gastric aspirate/juice: Sensitivity 94% (95% CI 0.84-0.99), comparable to sputum 2
• Bronchoalveolar lavage (BAL): Performance similar to sputum but specific data limited in guidelines 1
• These specimens are particularly useful in children or patients unable to produce sputum 1

Extrapulmonary Specimens

Overall Performance

• Non-respiratory specimens (pooled): Sensitivity 33-89%, specificity 95-99.7% 3, 6
• The dramatic drop in sensitivity compared to respiratory specimens makes negative results unreliable for excluding TB 3, 6

Specific Extrapulmonary Sites

Cerebrospinal Fluid (CSF)

• Sensitivity 56-62%, specificity approximately 98% 5
• Critical limitation: False-negative rate of 40-50% in TB meningitis 5
• A negative result cannot exclude TB meningitis; empiric treatment should not be withheld based on negative GeneXpert alone 5

Pleural Fluid

• Sensitivity 56-62%, specificity approximately 98% 5
• Performance is suboptimal for tuberculous pleuritis 2

Lymph Node Tissue/Biopsy

• Sensitivity 60-63%, specificity 33% 3
• Major pitfall: Extremely low specificity makes this the worst-performing specimen type 3
• Care must be taken when interpreting GeneXpert results from lymph node tissue 3

Bone and Joint Specimens

• Both biopsy and joint fluid demonstrate high accuracy for bone and joint TB 2
• Specific sensitivity/specificity values not provided in guidelines but performance described as satisfactory 2

Urine and Stool

• Limited data in guidelines; not recommended as primary diagnostic specimens 5
• May be useful in disseminated/miliary TB as part of multi-specimen strategy 5

Rifampicin Resistance Detection

• Sensitivity for rifampicin resistance: >97% when performed on single specimen, increases to >97% on three specimens 1
• Specificity for rifampicin resistance: >97-99% 1, 4
• Critical caveat: Positive predictive value is low in populations with low prevalence of drug resistance (<2% MDR-TB prevalence) 1
• All rifampicin resistance results must be confirmed by phenotypic culture-based drug susceptibility testing 5, 7, 6

Special Populations

HIV-Positive Patients

• Pooled sensitivity: 79% (lower than HIV-negative patients) 4
• Performance remains acceptable but false-negative rate is higher 8, 4

Miliary/Disseminated TB

• Sensitivity: Only 50-60% across all specimen types 5
• False-negative rate: 40-50% 5
• Critical management point: Negative GeneXpert cannot exclude miliary TB; empiric treatment should be initiated based on clinical suspicion 5

Clinical Interpretation Algorithm

When GeneXpert is POSITIVE:

• Presume TB and initiate treatment immediately (PPV >95% for respiratory specimens) 5, 4
• Rifampicin resistance results should guide initial therapy but require phenotypic confirmation 7, 6

When GeneXpert is NEGATIVE:

• AFB smear-positive but GeneXpert negative: Check for inhibitors, collect new specimen 1, 5
• Smear-negative, GeneXpert negative, high clinical suspicion: Do NOT exclude TB; proceed with empiric treatment while awaiting culture 1, 5
• Extrapulmonary specimens: A negative result is particularly unreliable; culture remains essential 5, 3, 6

Critical Pitfalls to Avoid

• Never rely on GeneXpert alone for extrapulmonary TB: Sensitivity is too low (33-89%) to rule out disease 3, 6
• Never withhold treatment for miliary TB based on negative GeneXpert: False-negative rate approaches 50% 5
• Never start MDR-TB regimen based solely on GeneXpert rifampicin resistance: Phenotypic confirmation is mandatory 7
• Always send specimens for culture regardless of GeneXpert result: Culture remains the gold standard and provides comprehensive drug susceptibility data 1, 5, 3
• Lymph node tissue has unacceptably low specificity (33%): Interpret positive results with extreme caution 3

Specimen Collection Strategy

• Collect multiple specimens (≥3 sputum samples on different days) to maximize sensitivity 7
• For suspected extrapulmonary TB, obtain specimens from multiple sites (respiratory, urine, blood, relevant body fluids) 5
• Reserve sufficient specimen volume for culture before performing GeneXpert 5
• Results should be available within 48 hours of collection 5