Prevalence and factors associated with tuberculosis infection among new school entrants, New York City, 1991-1993.

SETTING New York City public (or state-run) and private schools-elementary and secondary. OBJECTIVE To describe the prevalence and determine factors associated with positive tuberculin skin tests (TSTs) in school children. DESIGN Mandatory TST surveys among cohorts of new school entrants for the 1991, 1992 and 1993 school years, of whom birthplace was known for 81%. A positive tuberculin skin test defined as > or =10 mm induration. RESULTS Of the 298506 new school entrants, 2.1% (6326) were tuberculin test positive. The proportion that was tuberculin test positive was 0.5% (931/199 728) among US-born and 9.2% (3794/41 346) among foreign-born students. Foreign-born (FB) students with a history of BCG vaccination were much more likely to have a positive tuberculin test than US-born students (13.6% vs. 0.5%, odds ratio [OR] = 33.6, 95% confidence interval [CI] 31.7, 35.6), and were more likely to have a positive tuberculin test than FB students with no history of BCG (13.6% vs. 4.4%, OR = 3.4, 95% CI 2.5, 4.6). Older age was independently associated with tuberculin test positivity, except among foreign-born BCG-vaccinated children, in whom the youngest were more likely to have a positive tuberculin test. CONCLUSIONS Even in the midst of a tuberculosis resurgence such as that experienced by New York City, where tuberculosis cases nearly tripled from 1978 to 1992, the risk of tuberculosis infection among school children remained quite low. Given the reduced predictive value of the tuberculin test among low risk children and the effects of BCG vaccination, many children (especially younger children) with positive tuberculin test results are probably not infected with Mycobacterium tuberculosis. To reduce unnecessary evaluation and treatment, routine tuberculin tests should be administered only to high risk groups such as older children from countries with high rates of tuberculosis.