Conclusion
Bayes' Theorem is important: with low prevalence of disease, as is often the case in primary care or when screening for disease, even great tests can have significant false positive rates. Thus, getting a test "just to see" may actually cause harm, because of the increased likelihood of a false positive test (Remember Baby Jeff).
On the other hand, over testing can be just as confusing. Obtaining "just one more to test" to really nail down a diagnosis can result in false negative results, just causing confusion.
Clinicians thus have a responsibility to understand how well the tests they use "work" in different situations - this involves knowing the prevalence of disease in their practice and the characteristics of the tests they use.
