I post a fair number of stories about regulation of medical treatments here on Hacker News. I see this one has attracted a lot of reader interest. The comments

I post a fair number of stories about regulation of medical treatments here on Hacker News. I see this one has attracted a lot of reader interest. The comments already posted as I post this comment point out that once someone sets up a regulatory program, the incentives become powerful to figure out ways to cheat the regulators. And, as usual, scientific research is hard (another topic I post a lot about here), and researchers all operate under incentives to publish interesting and hopeful results rather than acknowledge that their research program didn't show what the researchers hoped to show. The submitted article identifies a specific problem in how the FDA doesn't publish identifying details of studies that its own inspectors discovered to have problems. Biomedical researchers, then, don't know which previous study findings are flawed, even though the FDA detected flaws, and much time and effort is wasted trying to pursue research results that don't reflect reality. That has to change. Meanwhile, there is a different pattern of regulation for some products that consumers in the United States take regularly as they seek better health, namely "natural supplements." For natural supplements, the regulatory problem is far worse, because those products can go on the market with no research done on safety or effectiveness at all, and without even any guarantee that the product contains the labeled active ingredient,[1] a plain violation of the warranty of merchantability in the Uniform Commercial Code. So the better trade-off for the consumer here is to amend existing law so that as drugs go through clinical trials, disclosure and peer review of FDA inspection findings during the trials is a condition for marketing approval. If manufacturers can't deal with that kind of regulation, tough. But it's only fair that natural supplement products, a multi-billion-dollar industry, should be regulated the same way. (To answer a question asked in another comment, already well answered overnight in my time zone, yes, "science-based medicine" includes "evidence-based medicine," that is clinical trials, as a subset of practices that include also consideration of fundamental principles of science and statistical understanding of prior plausibility before interpreting clinical trial results. If a sample size is small enough, many implausible treatments can look like successes in a clinical trial, especially for subjective, patient-self-reported symptoms, but we shouldn't call such treatments effective until they have been replicated in larger samples and verified with independently gathered "hard endpoint" data about clinical signs.) [1] "Searching for the supplement in your supplement" http://www.sciencebasedmedicine.org/searching-for-the-supplement-in-your-supplement/ http://www.sciencebasedmedicine.org/searching-for-the-supple...