Sensitivity and specificity of a diagnostic test determined by repeated observations in the absence of an external standard

Traditionally, the sensitivity and specificity of a new diagnostic test are derived from its application to two groups of individuals known to have or not to have met the criterion to be detected as judged by means of a "gold standard", some external time-honored test. In this study, a test with unknown accuracy parameters was used to detect an end-point criterion in an ongoing ophthalmological clinical trial. Since no external test method was available to assess the accuracy of this test, equations were derived relating the unknown sensitivity and specificity of the test to data frequencies based on replicate measurements. The solutions to these equations also provided estimates of the incidence rate of the criterion under investigation in the group tested, and of the predictive values of the test. The validity of this method of estimation is discussed, and applications to other situations are suggested.