Background: The main parameters to compare binary tests are the sensitivity and the specificity. Case-control sampling to compare two binary tests is frequent in clinical practice. This design consists of applying the two binary tests to all of the individuals in a sample of those who have the disease and in another sample of those who do not have the disease. In this design, the sensitivities (specificities) are compared from the case (control) sample applying the McNemar’s test. Other parameters of a binary test are the predictive values. The predictive values of a binary test represent the clinical accuracy of a binary test when it is applied to the individuals in a population with a determined disease prevalence.
Methods: This article studies the comparison of the predictive values of two diagnostic tests subject to a case-control sampling. A global hypothesis test, based on the chi-square distribution, is proposed to compare the predictive values simultaneously. The comparison of the predictive values is also studied individually. The hypothesis tests studied require knowledge of an estimation of the disease prevalence.
Results: Simulation experiments were carried out to study the type I errors and the powers of the hypothesis tests, as well as to study the effect of a misspecification of the prevalence on the behaviour of the hypothesis tests and on the estimators of the predictive values. The results obtained were applied to an example on the diagnosis of the Human African Trypanosomiasis.
Conclusions: A method has been proposed to compare the predictive values of two diagnostic tests subject to a case-control sampling. This method consists in: 1) Simultaneously comparing the predictive values applying the global hypothesis test based on the chi-square distribution to an error alpha; 2) If the global test is not significant, then the equality of the predictive values is not rejected. If the global test is significant to an error alpha , then the causes of the significance are studied solving the individual hypothesis tests and applying Bonferroni’s method (or Holm’s method) to an error alpha.
Keywords: Binary diagnostic test, Chi-square distribution, Positive and negative predictive values.