An assessment of nucleic acid amplification testing for active mycobacterial infection

Comparison of AFB microscopy, NAAT and AFB plus NAAT, using culture as a reference standard

Using both AFB microscopy and NAAT to diagnose MTB infections was more sensitive than using either test alone for both sputum and non-sputum specimens (Figure ); however, there was a corresponding decrease in specificity when the two tests were combined.

Figure Forest plot showing the pooled sensitivity and specificity values for AFB, NAAT and AFB plus NAAT compared with culture according to specimen type

AFB = acid-fast bacilli; K = the number of studies; NAAT = nucleic acid amplification testing; TB = tuberculosis.

Overall, 38% of all patients (29% of those providing sputum specimens and 54% of those providing non-sputum specimens) had a false-negative AFB microscopy result but only 2–3% were falsely positive when compared with culture. For NAAT only 11% of patients (11% with sputum specimens and 9% with non-sputum specimens) had false-negative results and 6% (5% with sputum specimens and 8% with non-sputum specimens) false-positive results. When the two tests were combined, 5–6% of patients had a false-negative result and 12% (8% with sputum specimens and 17% with non-sputum specimens) were falsely positive. However, as not all patients with a clinical diagnosis of TB will be culture-positive, it is uncertain what proportion of these false-positive patients have been truly misdiagnosed.

The clinical impact of a higher false-positive rate will result in some patients receiving treatment for a disease they do not have, until clinical unresponsiveness is noted or culture results are available. The consequences of a false-negative result are much more severe, as the patient may remain untreated for a longer time period and could potentially spread the disease to more individuals in the community.

When the summary LR+ and LR– values were compared, some differences between the tests were observed. The LR scattergram in Figure shows that the summary LR+ and LR– values for either AFB microscopy or NAAT are in the upper left quadrant. While AFB microscopy is only useful to confirm the presence of TB, NAAT also has some diagnostic value in identifying those without disease as it lies in the shaded area of this quadrant.

In contrast, when AFB microscopy and NAAT are combined the summary values are in the lower right quadrant, indicating that a negative result from both tests is a good indication that the patient will also be culture-negative for MTB. However, decreased certainty in a positive AFB microscopy or NAAT result correlating with a positive culture is due to the 22% false-positive NAAT rate for the AFB-positive population. As discussed above, culture is an imperfect reference standard; hence, many of these patients would receive a clinical diagnosis of TB.

Figure LR scattergram for diagnosis of MTB infection by AFB (A), NAAT (B) and AFB plus NAAT (C) compared with culture in studies using either in-house NAAT or the commercial Xpert NAAT

AFB = acid-fast bacilli; LR = likelihood ratio; NAAT = nucleic acid amplification testing