Browse the glossary:
diagnostic paired design
A paired study design for diagnostic tests makes it possible to directly compare the diagnostic accuracy of the tests within each patient.
In a fully paired direct comparison, all study participants receive all tests, as well as the reference standard.
If the design is not fully paired, participants receive only a subset of the tests. Within the set of not fully paired designs, randomized direct comparisons offer the best opportunity to avoid selection bias.
index test · reference standard test
If you feel that this definition hasn't helped you to understand the term, click on our monkey to let us know.