Both the TST and IGRA exhibit nonspecific variation when they are repeated for various reasons.88–93 If serial testing is to be performed, an understanding of the causes and magnitude of test-to-test variability is important for overall interpretation. There are common and unique causes of variation with both the TST and IGRA (see Appendix 1, Section A.2, Table A1). These factors may result in both quantitative differences in measurement and qualitative differences in interpretation (ie, positive vs negative) for both tests.
Serial testing may occur among populations for two specific reasons: (1) to identify a new TB infection after a recent TB exposure, such as in the context of contact or outbreak investigations; and (2) to identify new TB infections in populations at increased risk of exposure to TB, such as workers in healthcare settings, homeless shelters or correctional facilities.
The TST is the preferred test when conducting serial testing, as conversions have been well described in multiple studies and are well understood.85,89,94,95 If the initial TST is at least 10 mm, there is no clinical utility in repeating it. Among persons with an initial TST <10 mm, a TST conversion occurs when a 6 mm or greater increase occurs between two TSTs or when a second test measures at least 10 mm. TST conversion occurs within 3-to-8 weeks of exposure.89
With IGRAs, there remains no clear consensus on what constitutes a real “conversion” such that the change would indicate new TB infection. In multiple studies of serial testing, conversions and reversions occur frequently.32,96 A systematic review suggests that among health care workers, conversions occur 5 times more frequently with an IGRA than with a TST, with no supporting evidence in the form of observed increases in TB disease incidence.97 Therefore, IGRAs are not recommended in serial testing situations.
Recommendation
-
We conditionally recommend use of tuberculin skin testing when serial testing is planned to assess risk of new infection (ie, conversions). This includes repeat testing in a contact investigation, or serial testing of health care workers or other populations (eg, corrections staff or prison inmates) with potential for ongoing exposure. In these situations, interferon-gamma release assays are not acceptable (poor evidence).
Switch To: Français