BACKGROUNDThe health-care workers (HCWs) are at high risk of acquiring infection with Mycobacterium tuberculosis. The objectives of this study were to compare the performance of the T-SPOT.TB and tuberculin skin test (TST) for latent tuberculosis infection (LTBI), evaluate diagnostic concordance and risk factors for LTBI, and observe the progression to active tuberculosis (TB) disease among HCWs in a general hospital in Beijing.

METHODSThe prospective cohort study enrolled HCWs in a tertiary general hospital in Beijing, China, to evaluate LTBI with T-SPOT.TB and TST. The subjects were evaluated every 12 months during the 60-month follow-up.

RESULTSOf 101 participating HCWs, 96 and 101 had valid TST and T-SPOT.TB results, respectively. Twenty-nine (28.7%, 95% confidence interval (CI), 19.9% - 37.5%) were defined as positive by T-SPOT.TB and 53 (55.2%, 95%CI, 45.2% - 64.9%) were defined as positive by TST (using a ≥ 10 mm cutoff). An agreement between the two tests was poor (57.3%, κ = 0.18, 95%CI, 0.01% - 0.52%). In multivariate analysis, direct exposure to sputum smear-positive TB patients was a significant risk factor for a positive T-SPOT.TB (OR 5.76; 95%CI 1.38 - 24.00). Pooled frequency of antigen-specific IFN-γ secreting T-cells for subjects who reported direct contact with sputum smear-positive TB patients was significantly higher than that for participants without direct contact (P = 0.045). One of 20 participants with positive result of T-SPOT.TB and TST developed active TB at 24-month follow-up.

CONCLUSIONT-SPOT.TB is a more accurate, targeted method of diagnosing LTBI than TST.