The average drug costs around a billion dollars to bring to market and decades in development and approval. 23andme might not be a small fish by internet standards, but they are by medical industry standards.
But 23andme do not make drugs. They do diagnostic tests. The cost of development for those are much lower. 23andme aren't giants in that field for sure, but they're pretty big. It's absolutely nonsensical to compare them to a drug company.
More specifically they are busted for interpreting diagnostic tests not doing the tests. If they were two cooperating companies, and one did tests and squirted out results and the other sold diagnostic and interpretive services, the diagnostic services company would be busted but the test company would not be in trouble, as I interpret it at this time.
Skip all the exotic genetic issues and think about personal experiences with something boring, like blood pressure. No legal problem with testing someones blood pressure and giving them numerical results and maybe a medically sane set of typical ranges. My local pharmacy has a machine that does it for free. However, it would be a big issue for a non-MD using that data to diagnose high blood pressure, or a high likelihood of it anyway, then providing some treatment plans.
The issue seems to be selling services. If they just did testing, and provided a free downloadable immense data set and some statistical analysis tools and let the user come up with their own interpretations, they'd probably be in the clear (at least with respect to this individual problem). Now having a cloud service to hold the data and do the math is getting gray area ish but probably still OK as long as its the end user drawing the conclusions and not "the company" drawing the conclusions.
> Skip all the exotic genetic issues and think about personal experiences with something boring, like blood pressure. No legal problem with testing someones blood pressure and giving them numerical results and maybe a medically sane set of typical ranges.
Actually, the device used to take the blood pressure reading, if I understand things right, did have to be approved by the FDA.
The FDA's letter keeps referring to 23andme as a 'device', so it seems like the FDA's authorization/responsibility is specifically to regulate devices rather than services, actually, but they've decided 23andme can be technically classified as a 'device' under the relevant laws/regulations.
If it was instead considered a 'service'... I wonder if there'd instead be a problem with practicing medicine without a license, if it's not MD's who are providing the 'service'!
"the device used to take the blood pressure reading, if I understand things right, did have to be approved by the FDA"
that is true, the FDA would bust a mfgr of a device that claimed to report blood pressure but instead did astrology. This specific example of 23andme is, as near as I can tell, getting busted for interpreting the results, not for using 'cat /dev/urandom' to generate the results and then interpreting them.
Kind of like the same FCC that gets excited about out of band RF interference also got sorta excited by janet jackson's halftime show.
I am not sure there is as big a black-and-white distinction between 'providng' and 'interpreting' as you suggest.
23andme is a 'device' that takes a bodily fluid sample as input, and provides genetic information as output.
A an automated blood pressure cuff is a device that, well, I don't know how it works, but wikipedia says it "They do not measure systolic and diastolic pressures directly, per se, but calculate them from the mean pressure and empirical statistical oscillometric parameters", whatever that means, and produces some blood pressure numbers as output.
23andme is obviously a much more complex 'device'. Both both take some sort of measurement of your person, and then provide medical information.
Even if you could somehow draw a line at some point in this process of translation from physical measurements to medical information, some point at which the 'interpretation' begins, I'm not sure why one would suggest that either side of that line would be exempt from FDA regulation!
>23andme is a 'device' that takes a bodily fluid sample as input, and provides genetic information as output.
I don't think that's true, which is the problem. They don't just provide genetic information. They provide information detailing the impact of those genes, as medical science understands them.
>Even if you could somehow draw a line at some point in this process of translation from physical measurements to medical information, some point at which the 'interpretation' begins, I'm not sure why one would suggest that either side of that line would be exempt from FDA regulation!
Neither side would be exempt, but the physical side would be trivially approved as "yes, this does in fact measure those genes, good work everyone, pop the champagne".
He didn't fail to notice they're different agencies.
He is asserting that they both (being both government agencies) have each been inconsistent in their regulation.
