I’m not questioning the magic box part. Just whether the drop of blood is first put on a $10 disposable assay chip, and then the chip is put in the box, or if they were really claiming you put the blood directly in the box. That would certainly be a “wait, what?” moment for me as a prospective investor, even given my own very limited knowledge.
Several Stanford faculty members told her very early on that what she was proposing wasn’t possible. In particular, that you really can’t conduct all of those tests using just a pinprick of blood, which is why the phlebotomist draws a couple of vials for them.
That doesn’t seem like a great argument. The phlebotomist draws a few vials because that’s what the existing machines required.
I don’t think it’s a stretch to say that you probably don’t need a full vial for many of the tests, given that they all must have different requirements. But the vials are a standard size and easy to handle, and once you have a needle in someone you may as well collect a few. The problem appears to be:
- A single drop really is too little for almost anything, let alone several tests
- Collecting that drop with a finger prick contaminates it too much anyway
Not to mention that Holmes herself seemed to be incompetent.
As I recall, a finger prick damages a large percentage of the blood cells in the sample (resulting in bad results) , while a phlebotomists sample damages only a very small percentage of the larger sample.
At least some types of tests shouldn’t be affected, though. After all, blood sugar tests work with a finger prick. Presumably, damaged cells and other contamination don’t affect the glucose levels enough to be a factor.
There was probably the possibility of good science here, just not what they claimed. Among the harm that Holmes did was damaging the credibility of all research in the area. No one is going to want to invest in a blood test machine like this, even if it was much more limited in scope and was only claimed to measure things for which the limitations of a finger prick aren’t relevant.
Agreed.
Having given a lot of vials of blood for research recently, no they aren’t. There are several different sizes in use, and my last donation was collected into at least 4 sizes of vials.
Fair enough. My experience is more limited. Still–clearly there isn’t much additional cost to collecting more blood than needed. The theoretical benefit to the Theranos machine wasn’t exactly that it operated on so little blood, but rather that it could be done with a pinprick. That’s what would have gotten it into retail pharmacies. Even if it actually could have operated on a drop, it wouldn’t have been very useful if it still needed to be collected with a needle.
Genotyping works by extracting and comparing sequence of DNA. It actually only requires a very small amount of genetic material compared to a broad blood panel that needs a sufficiently large sample both for running individual tests and to get an adequate statistical measurement of the concentration of different minerals, ions, pathogens, et cetera. Blood glucose testing can be done with a single drop because glucose levels are very well distributed through blood quickly and reacts readily to glucose oxidase to produce a reaction rate that is easily measurable by concentration. The vast majority of blood tests are not this easy and require substantially more volume. This isn’t just a “this is how it has always been done” issue; it is a fundamental necessity for the testing.
There are plenty of automated diagnostics for blood tests that are available from credible biomedical manufacturers working on established principles that are validated against accepted baselines. Nobody should “want to invest in a blood test machine like this” because it was fundamentally unsound in how it was purported to operate, and in general Silicon Valley entrepreneurs should stay out of the field of biomedical testing and genetics unless they have an already well-established concept because the difficulties and potential harms of promoting and developing medical ‘vaporware’ are significantly more impactful than social media or consumer products. “Move fast and break things” is an okay philosophy—I guess—when it comes to something that doesn’t put peoples lives at risk and can be fixed with a firmware update, but it’s not a great way to develop anything healthcare related.
Stranger
I don’t see any sterilization component in the exploded view. It appears that the drop enters via their sealed “nanotainer” and stays there, with the measurements happening via non-contact devices like a camera and fluorometer (I’ve no idea how that’s supposed to actually work, but that’s what it shows). Where is the part where something would need to be sterilized?
Hence why I mentioned “much more limited in scope”. Had someone competent set more modest goals, especially with the idea that it was a process of discovery as much as anything, we might have seen some actual innovation.
Looking around, it appears that a few diseases at least (COVID-19 and HIV among them) can be detected via pinprick. A retail machine that can perform these seems useful.
Ok, found a video of Holmes speaking about it, and while I’ve just been skimming it, this came up:
So it looks like all the material handling, etc. really does reside on a single-use cartridge, and the idea that it would have to sterilize some shared microfluidics isn’t correct.
Not that any of this excuses the fact that it was fictional from beginning to end, but I can see why some aspects might be compelling. And also why professionals that worked for the company might not have a clear picture of the fraud, particularly if they had been compartmentalized from each other.
A sheaf of cheap pinprick tests seems like a fine idea. It’s the universal box that does it all that seems really dicey.
What they seem to be claiming is that while the box is “universal”, it just has the common elements like cameras, material handling robots, etc. The actual reagents, imaging slides, etc. are contained in the disposable cartridge. And the cartridges are not universal; they’re customized for the subset of tests you care about.
No, that isn’t right. If you want to watch a long explanation of how the system purportedly worked you can start at about the ten minute mark on this presentation (though I would frankly rather stick bamboo skewers through my ears than listen to Holmes’ fake baritone narration for an hour) but you can see that there is a centrifuge, cytometer, spectrometer, thermocycler, luminometer/fluorometer, et cetera, all of which require individual samples of the blood sample in the cartrige (and the reagents that are stored there), which are presumably delivered by the “material handling robot”. Holmes doesn’t go into how these instruments would be cleaned and sanitized between assays because there isn’t any plan for it and no practical way to do so. The cartridge is literally just a container for the blood sample and reagents and there is no onboard diagnostics on this.
SARS-CoV-2 and HIV can be detected by rtPCR testing, which requires a sample of some fluid (in the case of HIV it has to be blood, for SARS-CoV-2 and other respiratory pathogen it can be an oropharyngeal sample), or the presence of an immune response to many pathogens can be detected via ‘lateral flow’ (enzymatic response) test, albeit with significant variability in sensitivity and specificity. There are, again, many machines that can do mostly automated testing for some small subset of tests in samples correctly prepared by a laboratory technician. The point of the Theranos ‘Edison’ machine was to recreate the blood panel assay that is typically performed as part of an extensive blood diagnostics but without a phlebotomist taking multiple vials of venous blood, which is just not realistic short of having some kind of “test on a chip” that is essentially science fiction. This is why medical diagnostic and device development should be done by biomedical engineers as overseen by physicians and biomedical professionals in a qualified laboratory environment instead of a secretive startup which segregates its development and diagnostics and signs everyone to restrictive NDAs that make it essentially illegal to report even gross malfeasance and falsification of results.
Stranger
I watched a movie that weirdly came out around the time of the trial, and it included a lot of clips of Holmes taking up the system. And it really was intended to be a magic box. And her employees struggled to pretend to use the thing when investors and such visited the plant.
Inventor: Out For Blood In Silicon Valley, The The Inventor: Out For Blood In Silicon Valley - Google Search
Yeah, that’s the same video I linked to above. It appears to me that all the microfluidics stuff is part of the disposable cartridge; in particular the “cytometry slide.” The cartridge also contains a collection of disposable tips, vessels, optical cuvettes, etc. so there is no direct contact between any element of the machine and the samples.
I mean, the whole machine is fake so it doesn’t matter, but if it could be built, it would seem that is enough to not require any extra sterilization step.
Bail, formally “judicial interim release”, is a release from custody ordered by a court. It is available at both the trial level, and the appellate level. Prior to trial, a trial court can grant bail to the accused, but the trial court’s jurisdiction ceases once there’s been a conviction and sentence. It is functus officio. The convict is taken into custody immediately.
If the convict lodges an appeal with the appellate court, that court can entertain a bail application, pending the appeal. The appellate court takes into account factors such as flight risk, threat to the community, strength of the grounds of appeal, and so on. The appellate court can grant bail pending the appeal, on defined conditions, such as surrendering passport, regular check-in with probation officers, and so on.
For detailed review of the legal principles, see the Supreme Court of Canada’s decision in R v Oland:
How would a centrifuge possibly work without transferring fluid into it to be subjected to centrifugal differentiation? And how do you imagine that the other tools (cytometer, spectrometer, thermocycler, luminometer/fluorometer) could possibly work without having individual samples, notwithstanding the issues of using capillary blood contaminated with whatever might be collected from the dermis?
You seem to be trying to argue that while the Theranos execution was flawed, the basic concept is sound, which is utterly false. The entire basis for the technology was unworkable in fundamental ways which anyone doing the due diligence to consult with actual experts in the fields of microfluidics and hematological diagnostics would have immediately recognized. Holmes preyed on people with enthusiasm in her narrative but who lacked the basic understanding of why it wasn’t workable (and incidentally, why other biomedical companies weren’t pursuing such a device despite the obvious benefits if it could be made to work).
Holmes was a Harold Hill, selling the dream of a marching band to people who couldn’t hold a tune and decrying anyone who questioned her as naysayers. And this is what a large majority of Silicon Valley startups do; they push a technology that is nascent at best as achievable in some minimal amount of development time, create fear of missing out to ratchet up investment, and hope for a miracle to pull their hypothetical irons out of the fire even though they have dead coals. SiVal was instrumental in developing silicon transistors, turning academic research in operating systems into workable business and recreational computing, and fomenting the personal computing revolution, but that doesn’t translate into the right path for innovation in all fields. Medical technology and longevity in particular are areas of research that Silicon Valley startups have tried and massively failed to adequately develop because they just don’t follow the model of computing and software development.
Stranger
The sample is in a capsule. Normal medical centrifuges don’t have to be sterilized after each use, at least not as long as the capsule stays intact. The capsule is disposable and comes with the cartridge, as are the tips that transfer fluid. All the cartridge components are presumably sterile already as the whole thing is sealed from the manufacturer.
Based on their images, I would be concerned that there appears to be no barrier between the fluid and the tube that provides the suction for their transfer tips. They claimed a high precision, but if there was a failure it would be difficult to detect and then you really would have to sterilize everything. I’d put a piston or maybe a flexible seal in there to really eliminate any chance at contamination.
Not exactly. It’s clear that for many or most of their claimed tests, a drop of blood was not enough for accurate results (which they then lied about), plus the contamination/damage from the finger prick sample largely ruined any remaining utility. And the idea of taking multiple such tests with one drop is even more ridiculous.
It’s clear that there was also an execution problem, in that the stuff they were building were basically fakes.
However, based on the fact that some finger prick tests already give useful results in limited other contexts, and the likelihood that with further research more possibilities could be found, it seems that there is room in the market for a machine of this general nature. Not a magic box that can perform any blood test imaginable, but one that gives a useful subset of tests and is both automated and cheap enough that they could be put in retail pharmacies. Whether there is actually an intersection between the cost and utility of such a machine to be economic I couldn’t say.
Alternatively, perhaps there is room in the market for a box that works on “large” samples, but does so in fully automated fashion. The times I’ve needed blood tests, I had to find a Quest clinic, get the blood taken, and then wait a couple of days for the results to arrive. If they could process the blood there, at least for some common tests like liver function, it would be an increase in convience.
I’m seeing a new business opportunity – Cigars saying “Congratulations it’s a boy-voiced girl!”