A.4.1. Pretest probability of TB infection and predictive value
The pretest probability of TB infection refers to the probability a person truly has infection. The positive predictive value therefore reflects the likelihood that a positive result truly represents infection and the negative predictive value the likelihood a negative result truly represents absence of infection. The positive predictive value of the TST may be significantly reduced by BCG vaccination in populations with low pretest probability of infection; similarly, in populations with significant immune impairment, the negative predictive value of both TST and IGRA may be significantly impacted in populations with high pretest probability of infection.
A.4.2. Discriminatory ability of the test to identify individuals at increased risk of TB disease
The discriminatory ability of the test to identify individuals who will develop TB disease refers to the IRR, that is, the likelihood of developing disease among persons testing positive vs. negative. In populations where false negatives with a TST or IGRA are expected to be common (eg, due to immune impairment or other biologic reasons), the IRR of a test may fall substantially, even with very high specificity. Similarly, in populations where false positives are expected to be common, the IRR of a test may be substantially impacted, even if sensitivity is very high. Therefore, the discriminatory ability of a test is likely to be severely impacted in populations where sensitivity and/or specificity are expected to be reduced.
A.4.3. Risk of TB disease
The risk of TB disease is elevated among persons with medical conditions that affect immunity, recent infection or certain habits, and is described for various populations among persons with a positive test in Table 2 of the main text. Using estimates of IRR, risk of TB disease can be estimated among persons with a negative test. Risk of disease is the most paramount consideration when faced with a test result, as TB disease may result in long-term patient morbidity or even death. Risk of disease at the time the test is done should not be the only consideration; it is important to also consider future risk of disease. This is especially important among certain persons, such as those tested prior to initiating immunosuppressants, prior to transplantation, or early in chronic kidney disease, as risk of disease is likely to increase in the future.
Switch To: Français