An assessment of nucleic acid amplification testing for active mycobacterial infection

Impact on TB treatment

Unnecessary treatment and overtreatment for TB were reported in three studies. Forty-seven out of 143 patients without culture-positive TB were initially treated empirically pending culture results in the medium-quality cohort study by Davis et al. (2014). Xpert results were negative in 45/47 (95.7%) of these patients, whereas 1 patient who was not treated had a positive result. Only 8 (18%) of these 45 patients were clinically diagnosed with culture-negative TB. In conclusion, 82% (37/45) of patients correctly classified by Xpert were over-treated for active TB. If Xpert had been used to guide initial treatment decisions, the median duration of overtreatment would have been 1 day (IQR 1–3) compared with 46 days (IQR 45–49), a median difference of 44 days (IQR 43–47). In this scenario 44 fewer patients would have started empirical TB treatment, and during the 13-month study period the total number of overtreatment days would have decreased by 95%, from 2,280 (95%CI 2,081, 2,479) to 111 (95%CI 0, 256) days.

Although these results were hypothetical, they correspond with results from a historical control study of poor quality by Guerra et al. (2007), which reported a median duration of non-indicated TB treatment of 6 days for patients undergoing NAAT (MTD; Gen Probe, San Diego, CA) and 31 days for the non-NAAT group (p=0.002). Furthermore, a retrospective cohort study of medium quality by Marks et al. (2013) reported culture-negative NAAT-negative patients had significantly fewer average days on outpatient medications, compared with those receiving no NAAT, with an average of 3 versus 57 days for AFB-positive culture-negative patients and 58 versus 100 days for AFB-negative culture-negative patients, respectively (p<0.05).

Conversely, a retrospective cohort study of medium quality from Saudi Arabia (mediun TB incidence of 15/100,000 people¹²) reported a lack of change in overtreatment and patient management after NAAT in current clinical practice (Omrani et al. 2014). Anti-TB therapy was not discontinued in any patients with negative Xpert results that started therapy empirically (n=8). Furthermore, Xpert was requested in only 54.3% (76/140) of patients and, overall, an Xpert-positive result was the reason for therapy initiation in just 12.1% (17/140) of patients. The authors concluded that physicians who are highly experienced in the diagnosis and treatment of TB underused the Xpert NAAT and it had only a limited impact on their decisions related to starting or stopping anti-TB therapy.

A cohort study of medium quality by Sohn et al. (2014), conducted in Canada (low TB incidence of 4.6/100,000 people), also reported that the Xpert NAAT had no impact in preventing unnecessary TB treatment; however, in these patients species confirmation was done by existing NAAT in the clinical lab within a day of the positive AFB microscopy, and clinical suspicion in these cases was low.

NAAT had a clinical impact on management in 39% (20/51, 95%CI 27%, 53%) of patients in the retrospective cohort study of poor quality by Taegtmeyer et al. (2008), conducted in the UK (medium TB incidence of 15/100,000 people). In 7 patients for whom there was uncertainty about TB diagnosis, TB was confirmed by NAAT and TB therapy continued. Three patients who had started empiric TB treatment were able to stop because of NAAT results, 2/4 patients with MRD-TB were identified by NAAT, and in 5 patients (previously treated) MDR-TB was excluded. In 3 patients the need for a hospital contact-tracing exercise was confirmed. The study further hypothesised that if NAAT had been used in the other 36 patients for whom it was indicated, it could have had a clinical impact on 8 of them (22%). If NAAT had been used in the 36 patients who were AFB-positive (not indicated by British guidelines), the results could have stopped unnecessary treatment in 14% (5/35, 95%CI 5%, 29%) of patients who did not have TB, provided rapid confirmation in 19 patients and excluded TB in a further 12 patients. The authors concluded that there were significant clinical benefits from the use of NAAT in low-prevalence settings.