If you're a digital health entrepreneur and treating what's happening with Theranos as a wakeup call to do rigorous science, you could do worse than to work with the UCSF Center for Digital Health innovation:
I'm a machine learning engineer working with UCSF cardiology. I certainly don't know all the answers, but I'm happy to try to answer questions or direct people to the right place.
I think it's quite dangerous when there's a company that says "we have breakthrough science, but we definitely can't tell you what it is", which was Theranos's M.O. for a long time.
Working in medicine is hard, working with biology is hard, not just because of regulatory hurdles, but because fundamentally the problems in those fields are much more commonly questions of basic science than engineering.
Good science leads to good companies, and if you can't do good science you're probably going to fail in this space. Good to see solid research institutions creating spaces to nurture good science at startups.
"we have breakthrough science, but we definitely can't tell you what it is"
Haven't followed the story but that is stock snake oil marketing boilerplate. If they really were/are saying that, no surprise that they were shown to be frauds.
Yep. I think there's room for scientists, engineers, and designers to contribute, but you have to publish peer-reviewed studies showing hard outcomes. Secrecy doesn't work in medicine.
They claim that it's because they can't patent their inventions and are instead relying on trade secrets. Revealing their trade secrets wouldn't make them trade secrets anymore.
A lot of the things they've claimed to be developing (microfluidic devices, signal amplification strategies, novel assays, etc.) are totally patentable. Maybe they don't want to reveal their secrets to inhibit competition, but in medicine you have to prove the thing works before you actually use the thing.
I don't much care if they want to hide some aspects of how their process works, but under no circumstances should anyone in this space operate without conclusively, publicly and openly proving that the technology works. The standard for that is scientific publication, and the ability of disinterested parties to test the equipment and verify it does what it claims to do.
Plenty of medical/biotech/diagnostics companies have cleared that bar while maintaining trade secrets. Theranos should be submitted to the same standards, and as far as I can tell, whenever they have been tested, they have failed.
http://centerfordigitalhealthinnovation.org/
I'm a machine learning engineer working with UCSF cardiology. I certainly don't know all the answers, but I'm happy to try to answer questions or direct people to the right place.