An assessment of nucleic acid amplification testing for active mycobacterial infection

Point-of-care NAAT for the detection of MTB and rifampicin resistance

Indigenous Australians who live in remote communities face specific challenges in being able to access healthcare initiatives such as TB control programs. Rapid diagnosis and treatment is essential to contain the spread of TB in these communities, especially to children and infants. Thus, point-of-care testing with same-day results offers easier access to diagnosis and more-rapid treatment initiation for people living in these isolated communities.

Xpert is the first fully automated NAAT developed for point-of-care diagnosis of MTB and rifampicin-resistant MTB, and was endorsed by the WHO in December 2010 (WHO 2014). WHO stated that ‘Xpert testing should not be placed solely in centralized reference laboratories since patients gain the greatest benefit from the test when it is placed as close as possible to the point of care’. However, WHO also noted that certain conditions and infrastructure need to be available to ensure its efficient use. These include a stable and continuous electrical supply, an ambient temperature of 15–30 °C in the testing room, trained staff to perform the test, and biosafety precautions similar to those needed for direct AFB microscopy.

Three studies that met the inclusion criteria and looked at the use of Xpert in a point-of-care setting were included in this report. Only 1 of these studies provided any diagnostic accuracy data, reporting on the concordance between Xpert conducted by a nurse and by a trained laboratory technician. All 3 studies reported on differences in time to treatment initiation, with 2 studies also reporting health-related treatment outcomes.

A randomised, parallel-group, multicentre trial conducted by Theron et al. (2014) randomised adults with symptoms suggestive of active TB from five primary healthcare facilities in South Africa, Zimbabwe, Zambia, and Tanzania to nurse-performed Xpert NAAT or sputum AFB microscopy at the clinic. In this study nurse-administered Xpert had substantial agreement with that done by a laboratory technician on a paired sputum specimen (κ=0·69; 95%CI 0·64, 0·74), and had a similar sensitivity and proportion of unusable results. The authors reported that nurse-administered Xpert detected 154 (83%) of 185 culture-positive patients, 112 of whom started treatment on the same day. However, as AFB microscopy was also done on site, the delay in treatment initiation in 91 (50%) of 182 culture-positive patients was only 1 day (IQR 0–4). Thus, it was not surprising that there were no significant differences in morbidity and mortality at either 2 months or 6 months follow-up between the two groups. Nevertheless, nurse-administered Xpert detected a larger proportion of culture-positive cases than AFB microscopy (83% versus 50%). Thus, more patients would start treatment immediately after Xpert than after AFB microscopy.

A cohort study by Van Rie et al. (2013a) reported on the use of Xpert at a large primary care clinic in South Africa between April and October 2010. On presentation, two sputum samples were collected for AFB microscopy (and NAAT if the patient consented to participate in the study), and the patient was given a 5-day course of antibiotics if clinically indicated and asked to return within 5–7 days. Individuals returned to the clinic for their results after a median of 8 days (IQR 6–22). A third sputum sample was then collected and sent to a central laboratory for fluorescent AFB microscopy and liquid culture. The authors reported that patients who were Xpert-positive were started on anti-TB treatment on the same day as collection of the third sputum specimen in 15/16 cases, compared with a median delay of 13 days (IQR 7–27) for 38 patients diagnosed by chest X-ray and 34 days for 1 patient diagnosed by culture. Three patients were identified as AFB-positive and Xpert-negative (two were also culture-negative); however, none of them started treatment due to unsuccessful tracing.

A third study by Hanrahan et al. (2013) was conducted between July and September 2011 at the same South African primary care clinic as the study by Van Rie et al. (2013a). In this study 96% (48/50) of Xpert-positive patients were started on treatment on the same day as presenting with symptoms (IQR 0–0), compared with a treatment delay of 14 days (IQR 7–29) for 18 Xpert-negative patients who were diagnosed by chest X-ray, 144 days (IQR 28–180) for 14 patients diagnosed by culture and 14 days (IQR 5–35) for those diagnosed empirically according to symptoms. However, at 6 months, treatment outcomes did not differ significantly between patients who were initially Xpert-positive or -negative (p=0.46). Among the 48 Xpert-positive cases started on treatment, 48% had a successful treatment outcome (i.e. 6-month treatment completion or cure) and 2% died. Among the 58 Xpert-negative patients started on treatment, 64% completed the treatment or were cured, and 2% died.

Thus, Xpert could be suitable for use in small regional hospitals and clinics in rural areas of Australia if suitable training of personnel was available. This would reduce the time between specimen collection and availability of test results for people living in remote communities, and may result in quicker treatment initiation. In addition, the early knowledge of rifampicin resistance may influence treatment decisions, ensuring that appropriate anti-TB drugs are given immediately. The linked evidence on patient outcomes due to a change in management indicated that there does not appear to be any advantage for patient health-related outcomes (e.g. cure) with early versus delayed treatment. However, early appropriate drug treatment reduced both the spread of TB to close contacts and the likelihood of developing drug resistance. Both of these are important public health outcomes essential for the control of TB in Australia.