The impact of the prevalence of tuberculous infection on the predictive value of a positive tuberculin skin test is exemplified in this model applied to findings in the US Navy recruit study.
Navy recruits who have a history of contact with a tuberculosis case are expected to have a higher prevalence of infection than those without such a history.
Among persons without a known history of contact only about 10% with an induration of 10 mm will actually have tuberculous infection. Among recruits with a history of contact, the expected prevalence of tuberculous infection is much larger. With the same induration of 10 mm, the test is now correct in about 80%.
