It is pretty damn hard for us old guys to do due diligence on a tech company. I was recently trying help out on due diligence on a startup using AI to improve Facebook and Google ad effectiveness. My only AI course was 38 years ago :-), I’m somewhat familiar with the digital ad space since I was around for the beginning, but a shit ton has change. You can ask others with more knowledge to help but they are busy, so at the end of the day, a lot is simply based on trust
Holmes was pretty convincing. Was a she a true believer who was fooling herself, or a con artist or maybe a bit of both?
It certainly goes to show that board made of foreign policy experts is a bad fit for medtech startup.
Two-step due diligence for Theranos:
- Let’s see that device do the tests while we watch.
- By the way, what’s with these bills for third party labs? What do we use them for?
magnet
5771
I don’t know if the board got to see the devices in action, but Joe Biden did. The whole thing was an act.
But hey, surely you don’t need medical training to understand whether a medical testing device is intentionally deceiving you. Biden, like Mattis, must have known Theranos was a fraud. And since we know everyone tells everyone everything, we know that Obama and Pelosi were in on Theranos fraud.
Did they send samples to outside labs? I thought they did their conventional blood tests on their in-house third party machines. And before you ask, a legitimate enterprise would still need those machines for validation of new tech.
No, you just need to audit the results.
Could be.
magnet
5773
How would that work, exactly? You have a pinprick blood sample, and a set of results. You can’t send that pinprick out to a third party lab, because they need a vial of blood.
Now, maybe you meant that you need a prospective clinical trial. Recruit brand new subjects and pay them for two blood draws. Theranos gets one, an independent lab gets the other, and the results are compared.
As far as I know there was no such clinical trial, which should be a huge red flag to anyone trained in biomedical research. And eventually the red flags were raised. But not from the board, because they lacked the necessary training.
What they might have heard is that the machines are routinely tested using known samples. For instance they have a large amount of blood from a known diabetic, and it’s tested every morning and the results show diabetes 100% of the time. Which sounds reassuring, but all it tells you is that the machine can reliably diagnose one particular person.
Absolutely one of the first steps in FDA approval for any medical device is to be able to replicate the results from the predicate medical device (AKA something that is an FDA approved and generally the most common device in a field.) So sending out samples to 3rd parties as well having existing blood analyzers in your labs is completely consistent with getting FDA approval.
One of my investment is startup that is developing a device that helps measure blood coagulation (basically iPhone with cartridge attached that you stick a blood sample into.) Like Thernos this started off a DOD project, for obvious reasons. I’ve been meeting with CEO roughly quarterly for the last 3 years,and I’m only finally now beginning to understand what the process is involved and we are still getting everything together for Phase 1 trials.
Good grief. Hire a third party to conduct the test. Draw multiple vials of blood from multiple test subjects. For each vial produce the pinprick you need for Theranos’ device to run the test. Send the rest of the vial to a third party to run conventional tests. Compare the results.
I don’t have an ounce of training, but I just arrived at the solution one needs the necessary training to arrive at. Go figure.
magnet
5776
What you are saying is precisely “Run a prospective clinical trial”. It’s more complicated and expensive than you suggest, since you can’t just hire some random person to draw blood on random patients and then do whatever you want with it.
You aren’t the first to suggest Theranos should have run a prospective clinical trial. On the other hand, they are almost always analyzed by the group that developed the new drug or technique. So if Holmes were doctoring the results, then she would still have an opportunity to do so. Yes, I recognize that this is a huge potential conflict of interest in all pharma research.
Now, you might suggest that Theranos should have run a prospective clinical trial, but Holmes should not be allowed to participate in the analysis. Which would basically be like the directors of Apple telling Tim Cook that they don’t want him to be involved in the iPhone in any way, and he should not be present at any Apple events. If that’s really the plan, then you might as well just fire the CEO.
They claimed to be able to test a number of conditions, so under your “1 sample is positive for [X]”, they have a number of samples [N], each positive for some readout. You test all of the samples as an array, and each sample that is positive for [X] isn’t positive for [X+1], so you get 1 positive control and [N-1] negative controls for each condition. That sounds OK. Better, you have calibrator lines - (low/medium/high blood sugar) and can test both quantitative as well as qualitative readouts.
From what I remember of reading about theranos, they would take a sample of blood (more significant than a drop), then test it in house on 3rd party equipment and report that as the 1st party results. It makes sense to have all those other test equipment around - they are your validation/gold standard when you get new samples in. What’s obviously less good is the uh…lying part.
This is so true and on point that it’s painful.
Timex
5780
Wasn’t that the guy who sued John Oliver?
And one of my employer’s main competitors, setting some pretty important case law in the process.
Wait, you work in the coal industry, or the wider energy industry?
Media. Financial journalism specifically. The case was very important in delineating the protections of the press when it comes to subscription-based publications. A good chunk of our compliance training is devoted to it.
Ah, I think I understand now. Thanks
The next thing that will come out is that DJT didn’t know about it until it was underway.
