Nucleic acid amplification testing may be effective in diagnosing TB
Nucleic acid amplification testing may be helpful in determining whether patients who have tested positive for acid-fast bacilli have tuberculosis, according to the findings of a new study.
Researchers reviewed records of patients suspected to have TB during 2000-2004. They analyzed the performance of nucleic acid amplification on specimens that had been collected within seven days of TB treatment initiation to determine the sensitivity, specificity and positive and negative predictive values of the assay.
In 2000, the proportion of patients with confirmed or suspected TB whose respiratory tract specimens were tested using the assay was 12.9%. That proportion had increased to 15.6% by 2004 (P<.001).
Among patients who had respiratory tract specimens that had tested positive for acid-fast bacilli, 43.6% had been tested using nucleic acid amplification in 2000. In 2004, 55.5% of similarly-tested specimens had been analyzed using the assay.
Overall, there were 16,511 patients evaluated for pulmonary TB, 28.1% (n=4,642) of whom had specimens that had tested positive for acid-fast bacilli on a smear. Nucleic acid amplification had been performed on 48.3% (n=2,241) of those 4,642 specimens. Positive test results were observed in 57.1% (n=1,279) of those patients. Confirmed TB diagnosis was made in 98.7% of those 1,279 patients.
The nucleic acid amplification assay had a sensitivity of 96.0%, a specificity of 95.3%, a positive predictive value of 98.0% and a negative predictive value of 90.9% on 1,861 patients who had specimens that were positive for acid-fast bacilli on a smear.
There were 158 patients who had specimens that tested negative for acid-fast bacilli on a smear. The assay had a sensitivity of 79.3%, a specificity of 80.3%, a positive predictive value of 83.1% and a negative predictive value of 76.0% on those specimens.
The assay had a sensitivity of 97.5%, a specificity of 93.6%, a positive predictive value of 95.1% and a negative predictive value of 96.8% among 215 specimens that tested positive for acid-fast bacilli on a smear.
There was an association between a high grade smear and better test performance.
Nucleic acid amplification (NAA) tests are among the most sensitive diagnostic measures in microbiology and infectious disease. Laraque and colleagues summarize extensive experience in the diagnosis of tuberculosis from the Bureau of Tuberculosis Control New York City Department of Health. Using as a basis those specimens that eventually grew Mycobacterium tuberculosis in culture the specificity and sensitivity of a positive test result proved to be excellent: in excess of 90% and approaching 100%. Perhaps more important, for those individuals that were not smear-positive, NAA detected about 75% of cases of pulmonary tuberculosis that were eventually microbiologically-confirmed. Of respiratory tract specimens that are positive for AFB, NAA tests can reliably distinguish M. tuberculosis from non-tuberculous mycobacteria and this has implications for the selection of initial antimicrobial therapy.
Unfortunately, NAA testing of non-respiratory specimens proved to be less sensitive. In other words, for smear-negative samples of spinal fluid, peritoneal fluid and lymph node biopsy materials, a negative smear does not exclude the diagnosis. Thus, NAA tests are not currently FDA licensed for use with non-respiratory specimens.
–Lowell Young, MD
Infectious Disease News Editorial Board member