Interpretation of the tuberculin skin test

Area of Science:

• Epidemiology
• Infectious Disease Diagnostics
• Public Health

Background:

• The tuberculin skin test (TST), using purified protein derivative, is a key tool for tuberculosis screening.
• Traditional interpretation of TST results lacks nuance regarding population-specific characteristics.
• Modern approaches necessitate re-evaluating TST performance using contemporary statistical methods.

Purpose of the Study:

• To re-evaluate epidemiologic data of the TST using modern test characteristic analysis.
• To determine appropriate cutpoints for positive TST results across diverse populations.
• To calculate TST operating characteristics, including sensitivity, specificity, predictive values, and ROC curve analysis.

Main Methods:

• Utilized standard epidemiologic assumptions to differentiate Mycobacterium tuberculosis-infected from uninfected individuals.
• Employed data from U.S. Navy recruit and World Health Organization tuberculosis surveys.
• Applied Bayesian analysis for robust statistical interpretation.

Main Results:

• TST sensitivity ranged from 0.59 to 1.0, specificity from 0.95 to 1.0, and positive predictive value from 0.44 to 1.0, varying by cutpoint in the general U.S. population.
• The area under the receiver operating characteristic (ROC) curve was 0.997, indicating excellent discriminatory power.
• Optimal cutpoints for positivity varied significantly (2 mm to 16 mm) based on prevalence and testing objectives.

Conclusions:

• The TST exhibits superior operating characteristics compared to most screening and diagnostic tests for tuberculosis.
• Effective interpretation of TST requires considering disease prevalence and the specific goals of testing.
• Using variable cutpoints for different populations and interpreting induration size as a probability of infection enhances diagnostic accuracy.