Main Article Content
Evaluation of diagnostic tests and validation of an instrument have focused on estimation of sensitivity and specificity. In practice, more than two diagnostic tests may be performed either sequentially or simultaneously in the same subject and it may be of interest to evaluate their performance in relation to a gold standard procedure. Another peculiar situation occurs when a diagnostic test is repeated multiple times to the same individual. In recent years there has been an increase of research related to the evaluation of diagnostic tests involving grouped data or other type of data dependency. Several statistical methods have been suggested and evaluated in order to correct the bias and the variance estimates of sensitivity and specificity in different dependency structures. In this work we summarized and compared some of the methods available in the literature for point and interval estimation for sensitivity and specificity when there are more than two diagnostic tests performed in the same individual, or when more than a measure of the same test is available for each individual. Simulation studies are conducted and a data analysis related to Ascaris infection is shown. In any data dependency scenario, the estimation methods need to take into account the correlation between multiple measurements of the same individual. Definition of new methods is needed to cope with the increased complexity of the data structure of studies considering, for example, multiple measurements across multiple diagnostic tests.
Authors who publish with this journal agree to the following terms:
- Authors retain copyright and grant the journal right of first publication with the work simultaneously licensed under a Creative Commons Attribution License that allows others to share the work with an acknowledgement of the work's authorship and initial publication in this journal.
- Authors are able to enter into separate, additional contractual arrangements for the non-exclusive distribution of the journal's published version of the work (e.g., post it to an institutional repository or publish it in a book), with an acknowledgement of its initial publication in this journal.
- Authors are permitted and encouraged to post their work online (e.g., in institutional repositories or on their website) prior to and during the submission process, as it can lead to productive exchanges, as well as earlier and greater citation of published work (See The Effect of Open Access).