Abstract
Diagnostic tests are important clinical tools. To assess the sensitivity and specificity of a new test, its results should be compared against a gold standard. However, the gold-standard test is not always available. Herein, I show that we can compare the new test against a well-established diagnostic test (not a gold-standard test, but with known sensitivity and specificity) and compute the sensitivity and specificity of the new test if we would have compared it against the gold-standard test. The technique presented is useful for situations where the gold standard is not readily available.