An assessment of nucleic acid amplification testing for active mycobacterial infection

Component	A Excellent	B Good	C Satisfactory
D Poor
Evidence-base^a		One or two level II studies with a low risk of bias, or an SR or several level III studies with a low risk of bias
Consistency		Most studies consistent and inconsistency may be explained
Clinical impact				Slight or restricted
Generalisability		Population(s) studied in the body of evidence are similar to target population
Applicability			Probably applicable to Australian healthcare context with some caveats

SR = systematic review; several = more than two studies

^a Level of evidence determined from the NHMRC evidence hierarchy (see Table ).

Source: Adapted from NHMRC (2009)

Time to diagnosis and/or treatment

Fourteen studies reported data on time to TB diagnosis or anti-TB treatment after the intervention/comparator. Eight of these studies were conducted in countries with relatively high TB prevalence and 1 with intermediate TB prevalence (Table in Appendix ). Median time to TB diagnosis and median time to therapy are shown in Table and Table , respectively. It was shown that time to diagnosis is shorter with NAAT, compared with liquid and solid culture, and similar to AFB microscopy. Median time to therapy is also decreased with the use of NAAT (in these cases, Xpert) compared with other methods of diagnosis (especially culture), as shown in Table . These results correspond with those from two other studies: (1) a prospective cohort study by Sohn et al. (2014), which stated that for five subjects in their study who had AFB-negative Xpert-positive results, treatment would have started a median of 12 days (IQR 4–23) earlier if results had been shared with the physicians, whereas treatment would have been only around 1 day sooner for AFB-positive cases; and (2) a Spanish retrospective cohort study by Buchelli Ramirez et al. (2014), which reported that in the sputum AFB-negative group, Xpert-positive results allowed for an early treatment start. In this study treatment was brought forward by 26.1 ± 14.5 days, without waiting for culture results.

In addition to the median time to diagnosis and treatment data, a historical control study by Yoon et al. (2012) reported that the proportion of TB patients diagnosed on day 1 using AFB microscopy was 55%, compared with 78% in patients diagnosed by AFB microscopy plus Xpert NAAT (p<0.001). The study by Theron et al. (2014) showed that 44% (67/154) of culture-positive patients in a group that had AFB microscopy started treatment on the day of presentation, compared with 66% (122/170) in a group that had Xpert NAAT (p<0.0001). Furthermore, a medium-quality retrospective cohort study by Kwak et al. (2013) reported that the median turnaround time for Xpert results in Korea was 0 days (IQR 0–1), which was significantly less than AFB microscopy with a turnaround time of 1 day (IQR 0–1), liquid or solid culture with 14 days (IQR 10.25–1.75) and 24 days (IQR 17–30) respectively, and DST with 78 days (IQR 65–96). Time to confirmation of results by a physician was also significantly shorter for Xpert results in this study, with a median of 6 (IQR 3–7) days, compared with 12 (IQR 7–19.25), 21 (IQR 7–19.25), 38.5 (IQR 25.75–50.25) and 90 (IQR 75.75–106) days for AFB microscopy, liquid culture, solid culture and DST, respectively. Median turnaround time (from sampling to reporting) was also reported in the retrospective study by Omrani et al. (2014), which was 1 day for Xpert NAAT, 1 day for AFB microscopy (p>0.999) and 44 days for mycobacterial cultures (p<0.001). Laboratory processing times for AFB microscopy were 2.5 times as long as Xpert NAAT (23.2 hours, IQR 15.3–32.6 versus 9.1 hours, IQR 5.5–15.6, p<0.001), as stated by a cohort study done in the US (Lippincott et al. 2014).

Table Median time to TB diagnosis/detection using NAAT versus comparator