What are the sensitivity and specificity of sputum acid-fast bacillus (AFB) smear microscopy for diagnosing tuberculosis (TB)?

Sensitivity and Specificity of Sputum AFB Smear Microscopy for Tuberculosis

For pulmonary TB, sputum AFB smear microscopy has a sensitivity of approximately 54% for a single specimen, increasing to 70% with three specimens, and a specificity of ≥90%. 1

Pulmonary Tuberculosis Performance

Sensitivity

• Single sputum specimen: 53.8% sensitivity 1
• Two specimens: Increases sensitivity by mean of 11.1% (to approximately 65%) 1
• Three specimens: Achieves approximately 70% sensitivity total, with third specimen adding only 2-5% 1
• First morning specimen: 12% more sensitive than single spot specimen 1
• Concentrated specimens: 18% mean increase in sensitivity compared to non-concentrated specimens 1
• Fluorescence microscopy: 10% more sensitive on average than conventional microscopy 1

Specificity

• Specificity: ≥90% (relatively high) 1
• Positive predictive value: 70-90%, varying with prevalence of TB versus nontuberculous mycobacterial disease 1

Extrapulmonary Tuberculosis Performance

AFB smear microscopy performs significantly worse in extrapulmonary specimens, with sensitivity ranging from 0-42% depending on site, but maintains high specificity of ≥90%. 1

Site-Specific Sensitivities

• Pleural fluid: 0-10% 1
• Pleural tissue: 14-39% 1
• Urine: 10-30% 1
• Cerebrospinal fluid: <5% 1
• Peritoneal fluid: 0-42% 1
• Pericardial fluid: Variable, generally poor 1

Specificity for Extrapulmonary TB

• Specificity: ≥90% across all specimen types 1

Critical Clinical Implications

False-Negative Results

• A negative AFB smear cannot exclude TB because false-negative results are exceedingly common, particularly in extrapulmonary TB where sensitivity is <50% for all specimen types 1, 2
• HIV-infected patients are even less likely to have positive AFB smears, consistent with lower rates of cavitary disease 2
• 40% of culture-confirmed pulmonary TB cases will have negative smears even with optimal technique 2

False-Positive Results

• A positive AFB smear is highly reliable with specificity ≥90%, meaning false-positive results occur only ≤10% of the time 1
• Positive results warrant immediate treatment initiation because infection is highly likely present 1

Optimization Strategies to Maximize Sensitivity

Volume Requirements

• Minimum 3 mL sputum required, but optimal volume is 5-10 mL 1
• Using ≥5.0 mL sputum increases sensitivity to 92% compared to 72.5% when processing specimens of any volume 3

Specimen Processing

• Concentrated specimens are strongly preferred over direct smears, increasing sensitivity from 34% to 58% for all AFB culture-positive specimens, and from 42% to 74% specifically for M. tuberculosis 1, 4
• Fluorescence microscopy preferred over conventional microscopy 1

Number of Specimens

• Three specimens on different days should be collected as standard practice in the United States 1, 2
• Testing of three specimens is normative practice strongly recommended by CDC and National Tuberculosis Controllers Association to improve sensitivity given pervasive issue of poor sample quality 1

Common Pitfalls to Avoid

• Never use negative AFB smear to exclude TB - additional diagnostic testing with culture and molecular methods is mandatory 1, 2
• Never rely on single specimen - sensitivity increases substantially with multiple specimens 1
• Never accept inadequate sputum volume - volumes <5 mL significantly reduce diagnostic yield 1, 3
• Never skip culture even with positive smear - culture is essential for drug susceptibility testing and definitive diagnosis 1, 2
• Never use direct smears when concentration is available - concentrated specimens are significantly more sensitive 1, 4