Interpretation of the tuberculin skin test in Mycobacterium bovis BCG-vaccinated Singaporean schoolchildren

Tuberculin skin test (TST) interpretation in Singapore is confounded by universal BCG vaccination at birth and by a revaccination policy for schoolchildren who are tuberculin nonreactors (TST < 10 mm, using 1 TU PPD RT 23) at 12 or 16 yr old, with not more than two BCG vaccinations given to any child. School health records for birth cohorts 1978 to 1984 indicate that 82.8% were revaccinated at age 12 yr. By collation with the national database of tuberculosis (TB) notifications, we examined the risk of TB disease in these cohorts for the 4 yr subsequent to TST reading at ages 12 and 16 yr respectively, in intervals of 0-4, 5-9, 10-14, 15-17, and <greater than or equal to>18 mm. Receiver-operating curves (ROCs) were constructed to ascertain the optimum TST cutoffs in screening for the likelihood of development of TB disease. A cutoff of 10 mm performed best for the group tested at age 12 yr, whereas 16 mm was found to be optimum in those tested at age 16 yr (>80% of whom were BCG revaccinated 4 yr earlier). We have extrapolated these findings to the screening of contacts, and utilized these cutoff points to guide the treatment of latent TB infection in this high-risk group.