TABLE 2-4. Importance of Disease or Condition Prevalence, Bayesian Probability Analysis

Test conditions:

sensitivity of test, 95 percent

specificity of test, 99 percent

Patient Status

Prior Probability

(%)

Probability of a Positive Skin Test

(%)

Product of Probabilities (col 1 * col 2)

(%)

Revised Probability (col 3I * 100/ sum col 3)

(%)

If prevalence is 1%

TB infection

1

95

95

49.0

No TB infection

99

1

99

Sum = 194

51.0

If prevalence is 0.1%

TB infection

0.1

95

9.5

8.7

No TB infection

99.9

1

99.9

Sum = 109.4

91.3

NOTE: TB = tuberculosis.

For individuals, false-negative results are a concern because of the lost opportunity for treatment to reduce their likelihood of developing active tuberculosis. False-positive results are a concern because they may lead to unneeded and potentially harmful treatment for someone who does not actually have latent tuberculosis infection. They can also provoke considerable anxiety. In some situations, a positive test result may be socially stigmatizing. For public and occupational health personnel, false-negative test results may lead to missed opportunities to control the transmission of tuberculosis and identify weaknesses in community or workplace control measures. False-positive results may lead to unnecessary use of limited resources for investigations and medical follow-up.

Other Limitations of the Tuberculin Skin Test

A number of factors can lead to false-positive or false-negative readings for a tuberculin skin test. They include

  • The inadequate conduct or management of the skin test or the skin testing program. With good training and careful practice, the tuberculin skin test is not difficult to administer correctly and to interpret accurately and consistently. A lack of resources and inadequate training and oversight can, however, jeopardize the proper storage of the test material, the quality of testing procedures, and the appropriate recording and use of test results (Chaparas et al., 1985; Ozuah, 1999; see also Perez-Stable and Slutkin [1985]). In addition, differences in the products (reagents) used in



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement