Background
Tackling tuberculosis (TB) requires testing and treatment latent
tuberculosis in high-risk groups. The aim of this study was to estimate
the predictive values of the tuberculin skin test (TST) and interferon
gamma release assays (IGRAs) for development of active TB .
Method
A cohort of migrants and contacts of active TB patients were
prospectively recruited in clinics, the community and primary care. Each
participant received three tests (Quantiferon Gold In-Tube [QFT-GIT], TSPOT.TB
and TST). A positive TST was reported using three thresholds: 5mm
(TST5), 10mm (TST10), and 5mm in BCG-naïve or 15mm in vaccinated (TST15).
Participants were followed for a median of 2.9 years. Incident TB cases
were identified by national TB databases, telephone interview, and
medical note review. Outcomes were ratio of incidence rate ratios and
predictive values for TB development.
Findings
Ninety-seven (1.0%) of 9,610 participants developed active TB (77 of
6,380 with results for all 3 tests). In all tests, TB incidence was very
low in test-negatives (1.2-1.6 per 1000 per year). Incidence rates in
test-positives were highest for T-SPOT.TB (13.2, 95%CI: (9.9,17.4)),
TST15 (11.1 (8.3,14.6)) and QFT-GIT (10.1 (7.4,13.4)). Positive results
for these tests were significantly more predictive of progression than
TST10 and TST5. However, TST5 identified a higher proportion of
progressors than TST10, TST15, T-SPOT.TB and QFT-GIT.
Interpretation IGRA-based or TST15 strategies appear most suited for
screening. Although TST5 and TST10 detect more TB cases, they also
classify more individuals who are unlikely to develop TB as testpositive.
Funding source: National Institute for Health Research Health Technology
Assessment Programme 08-68-01.