A lot of the comments here reveal not having read the fine article (which is not very long). If you look at the graphic labeled "How Accurate Tests Can Be Mostly Wrong," about halfway down the displayed Web article, you will see a very carefully worked out (and realistic) example about how even a very accurate medical test can result in mostly false positive indications of a disease--all that is necessary for that, mathematically, is that there is a low base rate of the disease. Examples like this have been commonplace in books about statistical reasoning for making medical decisions for more than a decade, and I have shared links to Hacker News before that make this same point. This is something everyone needs to know (but the investors in Theranos didn't know) to make sound decisions about how much testing to do and what to do with test results.

Other authors who write about this issue are cited in the article kindly submitted for our discussion. I urge everyone here to read a lot of the writings of Dr. John P.A. Ioannidis,[1] who is quoted in the article.

[1] https://med.stanford.edu/profiles/john-ioannidis?tab=publica...