An assessment of nucleic acid amplification testing for active mycobacterial infection

Population and setting for the economic evaluation

The PASC protocol listed three populations with suspected active mycobacterial infections that would be considered potentially eligible for MBS-funded NAAT. These are patients:

• 
  with clinical signs and symptoms of active TB whose specimen is able to have AFB microscopy and C&S testing
  
• 
  with clinical signs and symptoms of active TB whose specimen is not able to have AFB microscopy but who have C&S testing
  
• 
  suspected of having an NTM infection who are able to have C&S testing.
  

NAAT is proposed to be undertaken as an additional test to existing test procedures in all these populations.

For patients suspected of TB the protocol indicates that initial treatment decisions are based on the clinical suspicion (pre-test probability) of TB, based on clinical judgement of the background epidemiology of the patient, presenting symptoms and imaging features. If TB is clinically suspected, patients are currently initiated on treatment irrespective of the AFB result. However, if the clinical suspicion of TB is deemed low, the decision to initiate treatment is based on AFB, if able to be performed. This is consistent with the current clinical management algorithm presented in Figure .

The introduction of NAAT is not expected to alter treatment initiation decisions in patients with a strong clinical suspicion of TB. However, as NAAT has the ability to identify mutations associated with rifampicin resistance, an appropriate MDR-TB treatment regimen may be initiated sooner in those identified with rifampicin resistance. In patients with a low clinical suspicion of TB, the PASC protocol considers that treatment decisions would be based on the NAAT result if AFB and NAAT are discordant. This is consistent with the proposed clinical management algorithm presented in Figure .

There is inadequate evidence available to demonstrate the effectiveness of NAAT in the second and third populations, and so economic evaluations for these populations would be inappropriate. Any health outcome difference incorporated into the model would not be evidence-based and therefore could only be speculative. Subsequently, a calculation of cost-effectiveness would be inappropriate as it would generate results that do not have an evidentiary basis. Any ICER would be subject to an unacceptable level of uncertainty and could be potentially misleading. However, a costing assessment has been undertaken of the financial implications for the MBS and Australian governments should the proposed listings for the second and third populations be accepted (see ‘Financial implications’)

The PASC protocol indicated that, due to possible differences in the accuracy of NAAT in patients with and without HIV, separate analyses should be presented (with the proposed structure of the decision analytic also considering HIV subgroups). However, as the clinical assessment found little difference in the accuracy between these populations (see ‘Comparison of AFB microscopy and NAAT, using culture as a reference standard in HIV-positive and HIV-negative patients’), and no evidence for change in management was identified, this subgroup has not been modelled separately. As all patients suspected of TB who are known to have an HIV infection would be considered to have a high clinical suspicion of TB, treatment is likely to be initiated on the basis of this suspicion. In this respect, the modelled scenario that best represents this population is that in which all patients are considered to have a high clinical suspicion of TB (see ‘Modelled economic evaluation’).