Re:A Five-Hour, $5k “Super Physical” Exam
by learnstats2 on Thursday March 19, 2020 @01:05PM
(#22629404)
Yes - you should always think about the risk of a false positive when considering a treatment (or, equivalently, the PPV). That risk hasn't vanished at the point of testing.
If you know the risk of a false positive and the benefit and risk of treatment, then you have the option of reducing risk: choosing an option with increased risk is a mistake, but a mistake in judgement, not a mistake due to the test.
You need to think about a tests Postive Predictive Value and Negative Predictive Value. The PPV is the number of people who get a positive test who have the condition divided by the number of people who get a positive test who do not have the condition. When the PPV is low the test can provide functionally no information. Now the important thing to remember is that the PPV changes based on the prevalence of the disease in the population, so a given test could be good for some people and bad for others, all depending on their base risk.