• Bryan Kestenbaum

Learning Objectives

  1. 1.
    Important characteristics of diseases that are appropriate for screening include:
    1. (a)

      The disease should be important in the screened population.

    2. (b)

      The disease process should have a preclinical phase.

    3. (c)

      Treating the disease process at an early stage should provide benefit.

  2. 2.

    Reliability refers to the ability of a test to provide repeatable results.

  3. 3.

    Validity refers to the ability of a test to detect true disease, as defined by a gold standard.

  4. 4.
    Two important measures of validity are sensitivity and specificity:
    1. (a)

      Sensitivity is the probability of testing positive given the presence of disease.

    2. (b)

      Specificity is the probability of testing negative given the absence of disease.

  5. 5.
    The predictive values of a test are defined as:
    1. (a)

      Positive predictive value is the probability of true disease given a positive test.

    2. (b)

      Negative predictive value is the probability of no true disease given a negative test.

  6. 6.

    Disease prevalence is required for calculating the predictive values of a test.

  7. 7.

    High test specificity is needed to reduce false positives when screening for a rare disease.

  8. 8.

    ROC curves present sensitivity and specificity characteristics for all possible cutoff values of a continuous screening test.

  9. 9.
    Four potential biases in screening studies that may lead to spurious associations of a screening program with health outcomes are
    1. (a)

      Referral bias

    2. (b)

      Lead time bias

    3. (c)

      Length bias sampling

    4. (d)

      Overdiagnosis bias

  10. 10.

    The association of a factor with disease must be extraordinarily strong to qualify that factor as a potentially useful screening test.



Receiver Operating Characteristic Curve Prostate Specific Antigen Screen Test Prostate Specific Antigen Level Detect Breast Cancer 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.


  1. 43.
    Ware JH. The limitations of risk factors as prognostic tools. N Engl J Med. Dec 21 2006;355(25):2615–2617.CrossRefPubMedGoogle Scholar

Copyright information

© Springer Science+Business Media, LLC 2009

Authors and Affiliations

  1. 1.University of WashingtonSeattleUSA

Personalised recommendations