Some problems in the determination of the false positive and false negative rates of tuberculin tests.

The estimation of the rates of false positive and negative reactions to the tuberculin tine test was considered when the true sensitivity of patients to tuberculin is unknown and must be estimated from the Mantoux test. It was demonstrated that correct estimates of the error rates of the tine test depend directly and critically on the opposite error rate of the Mantoux test. The exact relationship was derived, and its practical importance was shown to depend on the true prevalence of tuberculosis in the population investigated.