The issue isn't really that they're expensive - they are, they shouldn't be. Proactive access to care is good.
The article kind of hints at this, but it comes down to Bayes theorem [1]. The general population has a relatively low incidence of disease, so even tests that are fairly reliable in a population with a high incidence of disease become super inaccurate when applied to the general population.
Worse, false positives generally - especially in the US, due to high financial liability risk - end up involving expensive, invasive medical follow-up tests, and the probability of a false positive multiplied by the risk of the exploratory procedures outweighs a later diagnosis.
No wonder annual physicals don't do much good.
High-risk people should get periodic tests for the things they're at risk of, and people should be in a position to report issues they discover in the course of their lives. However testing otherwise-healthy people for things they probably don't have isn't likely to yield good results no matter how good the test. It's just math.
[1] https://en.wikipedia.org/wiki/Bayes%27_theorem
The article kind of hints at this, but it comes down to Bayes theorem [1]. The general population has a relatively low incidence of disease, so even tests that are fairly reliable in a population with a high incidence of disease become super inaccurate when applied to the general population.
Worse, false positives generally - especially in the US, due to high financial liability risk - end up involving expensive, invasive medical follow-up tests, and the probability of a false positive multiplied by the risk of the exploratory procedures outweighs a later diagnosis.
No wonder annual physicals don't do much good.
High-risk people should get periodic tests for the things they're at risk of, and people should be in a position to report issues they discover in the course of their lives. However testing otherwise-healthy people for things they probably don't have isn't likely to yield good results no matter how good the test. It's just math.
[1] https://en.wikipedia.org/wiki/Bayes%27_theorem