184 PART 4 Comparing Groups
best determined by some gold standard test that the medical community accepts as
perfectly accurate in diagnosing the condition. For COVID-19, the polymerase
chain reaction (PCR) test is considered a gold standard test because of its high
level of accuracy. But gold standard diagnostic procedures (like PCR tests) can be
time-consuming and expensive, and in the case of invasive procedures like biop-
sies, they may be very unpleasant for the patient. Therefore, quick, inexpensive,
and relatively noninvasive screening tests are very valuable, even if they are not
perfectly accurate. They just need to be accurate enough to help filter in the best
candidates for a gold standard diagnostic test.
Most screening tests produce some false positive results, which is when the result
of the test is positive, but the patient is actually negative for the condition. Screen-
ing tests also produce some false negative results, where the result is negative in
patients where the condition is present. Because of this, it is important to know
false positive rates, false negative rates, and other features of screening tests to
consider their level of accuracy in your interpretation of their results.
You usually evaluate a new, experimental screening test for a particular medical
condition by administering the new test to a group of participants. These partici-
pants include some who have the condition and some who do not. For all the
participants in the study, their status with respect to the particular medical condi-
tion has been determined by the gold standard method, and you are seeing how
well your new, experimental screening test compares. You can then cross-
tabulate the new screening test results against the gold standard results repre-
senting the true condition in the participants. You would create a fourfold table in
a framework as shown in Figure 13-3.
FIGURE 13-3:
This is how data
are summarized
when evaluating
a proposed new
diagnostic
screening test.
© John Wiley & Sons, Inc.