Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

We Shouldn’t Be Surprised at the Theranos Fraud

nautil.us

Post reply on HN
Print view View original

21–30 of 257 posts

Re: We Shouldn’t Be Surprised at the Theranos Fraud

#21

Post by tptacek »

xevb3k wrote: post #10
rotskoff wrote: post #6

While Theranos relied heavily on excessive optimism about technology's potential to solve a major problem, it also relied heavily on outright fraud. It's true that they had assembled a board of non-experts that were duped by unrealistic promises, but when the investors were lied to (as is being alleged in court) it seems difficult to say this is a symptom of some sort of cultural problem among investors.

I just finished reading the new book on Theranos. I’ve seen very similar things at other science based/biotech startups. In particular, over-promising and misrepresenting the state of development. Quite often these companies are funded by traditional tech investors who don’t know how to do DD on science-based/biotech startups. The difference with Theranos is that they messed with the FDA. They did consumer testing, a…

I think most of what went wrong with the FDA happened after Carreyrou's first WSJ article came out, didn't it? That WSJ article basically killed the company.

Re: We Shouldn’t Be Surprised at the Theranos Fraud

#22

Post by cepth »

ezzaf wrote: post #18

This snippet is amazing: "Theranos kept up the illusion that they were running a lot of the tests from finger-stick blood draws by hacking the Siemens machines to work with small samples" No wonder they had so many inaccurate test results.

The full process as described in the book is even more shocking.

Theranos would dilute the finger stick draws, and then run them on the Siemens machines 5 times. In theory, you could then calculate the concentration, and adjust for the dilution. In reality, the diluted samples would have concentrations of the relevant cells and elements that were too low to pick up.

To “fix” this, they would do those 5 runs, then calculate the median. And, if you had 5 outliers in all different directions, they’d discard the run and do 5 again.

The fact that PhDs and doctors who swore the Hippocratic oath allowed this to go on for over a year is shocking.

Re: We Shouldn’t Be Surprised at the Theranos Fraud

#23

Post by tptacek »

Animats wrote: post #9

Please stop linking to this paywalled site. Thanks.

You've been here long enough to know --- apart from whether or not Nautilus counts as "paywalled" --- that paywalled sites are fine on HN as long as there is a straightforward way to get around the paywall. Complaints about paywalls are, according to the site guidelines, off topic.

Re: We Shouldn’t Be Surprised at the Theranos Fraud

#24

Post by malshe »

tptacek wrote: post #20

The book is really good, well-written and paced like a thriller. I never paid any attention to Theranos during its heyday, and it was especially interesting after reading it to go back to the old HN threads about it and see what the community here predicted about it.

I agree. I bought the audiobook and I was looking for every opportunity to listen to it and get to the end. It made me actually stay on the elliptical a little longer in the gym :)

Re: We Shouldn’t Be Surprised at the Theranos Fraud

#25

Post by cepth »

tptacek wrote: post #21
xevb3k wrote: post #10

Earlier quoted context omitted.

I just finished reading the new book on Theranos. I’ve seen very similar things at other science based/biotech startups. In particular, over-promising and misrepresenting the state of development. Quite often these companies are funded by traditional tech investors who don’t know how to do DD on science-based/biotech startups. The difference with Theranos is that they messed with the FDA. They did consumer testing, a…

I think most of what went wrong with the FDA happened after Carreyrou's first WSJ article came out, didn't it? That WSJ article basically killed the company.

The timeline/impact is a bit murkier.

Carreyou had two big revelations. One was that Theranos was not running blood tests on their proprietary machines, which were supposed to do finger prick draws. The other was that Theranos had submitted fraudulent validation data to the Center for Medicare and Medicaid Services (CMS) and state regulatory bodies.

They did a couple brazenly unethical and illegal things.

1) CMS subcontracts lab inspections to state regulatory bodies. When California lab inspectors visited Theranos, they physically blocked off the door leading to a floor that had the Siemens equipment, presumably to hide the fact that they were not conducting tests on their own proprietary machines. So, the inspectors never actually inspected the relevant labs.

2) Theranos violated "proficiency testing" guidelines. Proficiency testing is the practice of running tests on a sample with a known concentration/result, and comparing results. The spirit, if not the letter, of the guidelines is that you run the tests on the same equipment as would be commercially used. Thus, Theranos sent in fraudulent tests by testing using Siemens machines. This proficiency testing duping was what caused Tyler Schultz to send in an anonymous tip to the New York laboratory regulatory body, and led to his final decision to leave the company.

3) Theranos tried to claim with the FDA that its tests were "laboratory developed tests", i.e. they were unique tests that could not be compared to existing market tests. Carreyrou has a scoop in the book that Theranos benefited greatly from a bureaucratic power grab between the FDA and CMS. Both agencies claim some sort of regulatory oversight for the blood testing market.

At least initially, there were many in the FDA who were suspicious, and wanted to send inspectors. But, Theranos's seeming passing of CMS inspections allayed fears in the short term. It was after Carreyrou's first two articles (IIRC), that the FDA sent a surprise inspection.

EDIT: I also think that in the book Carreyrou mentions that a number of doctors in the Phoenix area, where Theranos was doing a test pilot of their finger prick tests, had filed complaints with the state or federal regulatory bodies. So, there were definitely people at CMS or FDA who were beginning to pay attention.

Re: We Shouldn’t Be Surprised at the Theranos Fraud

#26

Post by kumarvvr »

The most amazing thing about this whole situation is that there were experts in the field who were shouting out about the impossibility of Holmes' claims right from the beginning.

The media, though, never paid enough attention (or was possibly subverted by Theranos itself) to the naysayers.

These incidents ought to remind us the exponentially important role media plays in our lives and how important it is that we come up with a framework to penalize misleading journalism, while at the same time, preserving the independence of media.

Maybe that is the next social challenge for entrepreneurs in the coming decade.

Re: We Shouldn’t Be Surprised at the Theranos Fraud

#28

Post by mikekchar »

cepth wrote: post #22
ezzaf wrote: post #18

This snippet is amazing: "Theranos kept up the illusion that they were running a lot of the tests from finger-stick blood draws by hacking the Siemens machines to work with small samples" No wonder they had so many inaccurate test results.

The full process as described in the book is even more shocking. Theranos would dilute the finger stick draws, and then run them on the Siemens machines 5 times. In theory, you could then calculate the concentration, and adjust for the dilution. In reality, the diluted samples would have concentrations of the relevant cells and elements that were too low to pick up. To “fix” this, they would do those 5 runs, then cal…

I don't know much about Theranos, but I remember writing a proof in university. The TA in the class was a PhD student doing research in group theory. My proof relied on doing some manipulation of exponents and there was a point where I was doing "1 + (-1) = 0". The TA just could not understand that bit. He was absolutely sure that in some circumstances this wouldn't hold. No amount of discussion about how integers form a field under addition and multiplication would convince him. He eventually got his PhD and I can only assume he's working for a company like Theranos.

Cue someone replying and telling me why I was actually the one that was wrong ;-) But the point is that even well educated people make really, really stupid mistakes.

Re: We Shouldn’t Be Surprised at the Theranos Fraud

#30

Post by cepth »

mikekchar wrote: post #28
cepth wrote: post #22

Earlier quoted context omitted.

The full process as described in the book is even more shocking. Theranos would dilute the finger stick draws, and then run them on the Siemens machines 5 times. In theory, you could then calculate the concentration, and adjust for the dilution. In reality, the diluted samples would have concentrations of the relevant cells and elements that were too low to pick up. To “fix” this, they would do those 5 runs, then cal…

I don't know much about Theranos, but I remember writing a proof in university. The TA in the class was a PhD student doing research in group theory. My proof relied on doing some manipulation of exponents and there was a point where I was doing "1 + (-1) = 0". The TA just could not understand that bit. He was absolutely sure that in some circumstances this wouldn't hold. No amount of discussion about how integers fo…

Fair enough! The book is worth the read.

I think I may be being a bit unfair to the scientists. Basically, Theranos was a hot enough startup that Elizabeth Holmes and Sunny Balwani would fire, demote, or marginalize anyone who made a fuss. Lifelong academics may have never had to deal with that kind of bullying/pressure before.

On the other hand, they had some pretty incredible scientific talent on staff. You would think that people who are highly cited and have a reputation to protect would have spoke up sooner. Instead, the first two people to make any kind of sustained fuss were two fresh out of college hires.

Post reply on HN