BACKGROUND The rapid and accurate diagnosis of latent tuberculosis infection (LTBI) is crucial in military settings because military personnel live in crowded circumstances and are of an age group with a high incidence of tuberculosis (TB). We tried to elucidate whether the tuberculin skin test (TST) accurately reflects the risk of TB infection among military personnel, in a setting of intermediate TB prevalence and where bacille Calmette-Guérin (BCG) vaccination is mandatory.

METHODS A multi-stage cluster survey was conducted among military personnel in South Korea. Participants were grouped according to their risk of TB infection: Group 1, no identifiable risk of TB; Groups 2 and 3, recent casual (Group 2) or close (Group 3) contact with smear-positive TB patients.

RESULTS Of 1045 participants, 857 (82.0%) had been BCG-vaccinated. The odds ratio (OR) of a positive TST (10 mm cut-off) for Group 2 (n = 184) and Group 3 (n = 83) compared with Group 1 (n = 778) were 0.95 (95%CI 0.67-1.38) and 1.7 (95%CI 1.06-2.70), respectively (P value for trend 0.16).

CONCLUSIONS The TST does not accurately reflect the risk of LTBI among young military personnel in a setting where there is intermediate TB prevalence and extensive BCG coverage.