The pattern of utilisation and accuracy of a commercial nucleic acid amplification test for the rapid diagnosis of Mycobacterium tuberculosis in routine clinical practice.

BACKGROUND

Several nucleic acid amplification (NAA) tests are available for the rapid detection of Mycobacterium tuberculosis (MTB) in clinical specimens.

AIMS

To identify the pattern of utilisation and accuracy of the AMPLICOR test in routine clinical practice in an acute care setting.

DESIGN

A retrospective descriptive study.

METHOD

We studied 159 consecutive specimens in which the AMPLICOR (Roche; Branchburg, NJ) test was requested by attending doctors. The sensitivities and specificities of the AMPLICOR for detection of active tuberculosis (TB) were calculated in relation to types of specimens, smear and culture results.

RESULTS

The number of requests more than doubled from 1999 to 2000. Thirty-eight percent of the specimens were not from the respiratory tract. The majority of the specimens had requests for one or more additional test (mean 1.8). The rate of active TB was 18%. The sensitivities of the AMPLICOR on per specimen, per patient, per smear negative specimen and per smear negative patient basis were found to be 81%, 80%, 66.7% and 71.4% respectively. The specificities for these groups accordingly were 99%, 98.6%, 99% and 98.6% respectively. The sensitivity and specificity for respiratory specimens were 97.5% and 98.5%, while for non-respiratory specimens, they were 60% and 100%. In smear negative specimens, the sensitivity and specificity for respiratory specimens were 60% and 98.5%, while for non-respiratory specimens, they were 75% and 100%. The AMPLICOR assay was negative in all 21 specimens of pleural or spinal fluid.

CONCLUSIONS

There is a growing demand for NAA in the rapid diagnosis of TB with a high proportion of non-respiratory specimens. The number of additional diagnostic tests performed on each specimen should be limited. In routine clinical practice, the AMPLICOR assay is a useful confirmatory test for active pulmonary TB. The utility of the AMPLICOR assay for MTB detection in exudative fluid specimens needs further evaluation.