Working out sensitivity, specificity and positive predictive value can be difficult enough when using just one test to diagnose an infectious disease (or any disease for that matter). What about if you are using two or more tests?
For example if you are trying to diagnose Disease X and you have performed two tests, A & B. What are the effects of combining the tests with regards to sensitivity and specificity?
It really depends on what you regard as a valid result. If you diagnose the disease based on only one of the two tests being positive then the combined sesnitivity is higher and the specificity lower than that of the two tests individually.
However if you will only make a diagnosis of disease if both tests are positive, then the combined sensitivity is lower but specificity is higher than the two tests performed individually.
When using three tests things become even more complicated!
Whether we regard a valid result as one test in the set being positive or all the tests being positive is very much up to us. There is no right or wrong answer. We can use multiple testing to our advantage when we have individual tests that either lack sensitivity or specificity.
However we often interpret multiple tests for the one disease the way we want to interpret them. For example if we want to diagnose the disease, then we will probably be happy with one result out of the set being positive. If we want the diagnosis to be negative then we will probably believe the negative result from the set. Retaining objectivity becomes very difficult.
Told you it was a minefield!
I have added a quick question and answer tutorial on the basics of Maldi-tof