Context One or both commercial tuberculin skin test reagents
(Aplisol and Tubersol) may have a high rate of false-positive
Objective To compare the reaction size and specificity of skin
testing with Aplisol, Tubersol, and the standard purified protein
Design Double-blind trial, conducted between May 14, 1997,
and October 28, 1997, in which each individual received 4 tuberculin
skin reagents at sites assigned at random.
Setting Health departments and universities in 6 US cities.
Participants A total of 1555 persons at low risk of latent
Intervention Simultaneous skin tests with Aplisol, Tubersol,
PPD-S1, and either a second PPD-S1 or PPD-S2 (a proposed new standard).
Main Outcome Measure Reaction size at each injection site measured
by 2 investigators blinded to type of reagent.
Results Aplisol produced slightly larger reactions than
Tubersol, but this difference did not significantly change skin test
interpretation. The mean ± SD reaction sizes were 3.4 ± 4.2 mm with
Aplisol, 2.1 ± 3.2 mm with Tubersol, and 2.5 ± 3.6 mm with PPD-S1.
Assuming that all participants were uninfected and using a 10-mm
cutoff, the specificities of the tests were high: Aplisol, 98.2%;
Tubersol, 99.2%; and PPD-S1, 98.9%. Significant variability was not
detected in interobserver, host, and lot-to-lot reagent comparisons.
Conclusion Using a cutoff of at least 10 mm, testing with 3
different PPD reagents resulted in similar numbers of uninfected
persons being correctly classified.