[theshadowknows]

When I first heard about the Theranos meltdown I was so sure that Holmes was a victim of big pharma trashing her name and her company. But man..as the story unfolds in front of us that is obviously not the case. I’m pretty bummed about it because I was really excited a few years ago when the company was so hot. Hopefully some of their tech or ideas are solid and someone really can build on them to make good healthcare products for people.

[nradov]

None of their tech or ideas were ever solid. It is biologically impossible to measure accurate blood levels of all but a few analytes from a fingertip blood drop.

[theshadowknows]

Yeah that’s what I’m understanding now. Really would have been amazing if it were possible. If you’ll indulge me, is there some theoretical technology that could enable such a breakthrough or is it so far from possible that it’s not even feasible for some reason?

[hn_throwaway_99]

No, it's physically impossible. There is plenty of evidence that capillary blood simply isn't homogeneous enough in small volumes to do quantitative analysis.

But honestly, I actually don't think that's much of a big deal. The difference between taking a venous draw and doing a finger stick as not as big of a deal as Theranos made it out to be.

If instead Theranos had a small, portable analysis machine that could work with small volumes of venous blood, and they had a big deal with Walgreens to use that together with things like low, up-front pricing for tests, I think that still would have been a great advance. Note there are startups like Everlywell basically trying to make that vision (cheap, accessible, up-front pricing for medical tests) possible.

[nradov]

Theoretically what could work is to flip things around. Instead of drawing blood from the patient and analyzing it in a lab, miniaturize the lab and implant it in the patient. That way the sensors have continuous direct access to blood or other fluids.

Obviously there are huge challenges to designing such an implanted medical device, and it could only ever work for a limited set of assays, but I expect something like that will eventually be possible. We already see diabetics wearing continuous glucose monitors, which gives a preview of how such implanted sensors might operate.