Can Theranos CEO Elizabeth Holmes Fend Off Her Critics? | Hacker News Reader

Can Theranos CEO Elizabeth Holmes Fend Off Her Critics?

bloomberg.com

78 pointsadventured59 comments

This quote really jumped out at me as almost a caricature of the disconnect between Silicon Valley internet/communication disruption and "disruption" in something like medical services.

"While Draper isn’t a board member or official adviser, he says he sees Holmes frequently. He’s puzzled by Theranos’s troubles and wasn’t aware the company isn’t currently using its famed nanotainer for anything but a herpes test. The idea of pausing the relentless push forward to wait for fusty government approvals doesn’t seem to make sense to one of the investors in Twitter and Skype. “So they’re going to have to go through each one of those tests before they can even use the nanotainer?” he says, sounding horrified. “You can run tests; you just say they are not FDA-approved,” Draper continues, referring to the nanotainer. “If that’s not the way it’s happening, I’m definitely going to give Elizabeth a call about this.” He pauses the interview to pull out his phone. “I would think they can still use the nanotainer and just have some wording,” he says, typing out a text message to Holmes. “I would think you just put a little warning label on the thing and say the FDA has not agreed to the accuracy of these things.”"

When it comes to medical devices, Draper's dumb money. The only reason he invested is because of good ol' nepotism:

> Draper, the venture capitalist and early Theranos investor, says he’s known Holmes since she was 6 years old and a friend of his daughter. When she decided to start her company, she approached him. He called her parents to make sure they were OK with her plan to drop out of school. Then he gave her $1 million.

> He called her parents to make sure they were OK with her plan to drop out of school. Then he gave her $1 million.

Wow, that’s a heck of a narrative. I would never criticize Ms. Holmes for using her connections like this, but when stated as above, this investment sounds like a total freebie. And we all know that, thanks to investor herd mentality, one investment is a springboard to less skeptical follow-on investment. In this case, the first investors don't sound skeptical at all, because they viewed this as a family-friend favor more than an actual business decision. If Theranos were perceived a hell of an opportunity, do you think he’d give a rat’s ass if her parents would rather she finish her degree? Can you imagine an investor calling Zuck’s parents — hey is it OK if your son continues this facebook thing, or should I encourage him to go back to school?

Don't investors generally care if a very key, single-point-of-failure person is in a position where they will receive a lot of resistance from family members to doing necessary tasks?

Edit: For example, if some really competent, unfunded entrepreneur had some brilliant idea to revolutionize liquor distribution, but their family were conservative Baptists (or Muslims, or Mormons, etc), wouldn't you want some kind of reassurance that won't be a problem before you invest?

Actually know an entrepreneur who was friends with Matt Mullenweeg's parents. They wanted him to talk with their son to convince him to not drop out of college. He told me he met Matt, was very impressed with WordPress and told his parents that what he was doing was OK.

> If Theranos were perceived a hell of an opportunity, do you think he’d give a rat’s ass if her parents would rather she finish her degree? Can you imagine an investor calling Zuck’s parents — hey is it OK if your son continues this facebook thing, or should I encourage him to go back to school?

If the investor was a family friend, uh, absolutely.

It's a person with an idea, a prototype on a tiny scale, and no funding. The people that fit that description are a dime a dozen, and the vast, vast majority will get nowhere or fail, no matter how good the idea. You're friends with their parents. Checking with them before giving the kid the means to drop out of college is not remotely ridiculous.

Whoa, such a meritocracy, where all it takes to make it big is to have rich parents and rich friends, and an big idea (good/correct idea optional)

Or was the whole point of his puzzled naive persona to make the point "we could be selling these, but because we take this safety thing so seriously we are not".

Most of HN, and all of Wall Street, knows they are not using their own tests. I doubt this investor is unaware.

“I would think you just put a little warning label on the thing and say the FDA has not agreed to the accuracy of these things.”

Unfortunately that's the key part of a medical test - the accuracy. The reason people can jump so hard on Theranos' back is because at the end of it all the science either works or it does not. And so far nothing seems to point to the technology working just yet.

I worked for a medical diagnostics company for 4 years and this is an incorrect view. Though I'm not fluent in the regulatory side I do know that most of our tests were not FDA approved. Most tests/assays were developed in-house, however, a few tests were licensed from other companies and all of those WERE FDA approved.

Another interesting caveat, when the FDA approved tests failed we could not dissect the process and try different variations/dilutions and were only able to give general reasons for test failure (it was sometimes referred to as "black box" testing). However, this was not the case for in-house developed non-FDA approved tests. For those, we had much more control and could give pretty definitive reasons for testing failure and sometimes we could re-test failed specimens with slight variations and obtain a valid result.

The term you're looking for in FDA speak is 'Laboratory Designed Test', I've included a link to the FDA's guidance document on LDTs below.

The reason why FDA approved tests frequently work as 'black boxes' is because the FDA requires that approved tests reach specific performance characteristics. Typically in order to prove these performance characteristics, companies must constrain the parameters of the testing. Once they do that that test is only considered 'valid' over those parameters. The level of error reporting is dependent on the company who develops the test.

The FDA is working to subjecting certain LDTs (basically stuff that guides treatment choices) to pre-market review, so a lot of these gaps will likely begin to be closed. One of the things that Theranos kinda trumpeted earlier is that their nanotainer HPV test was technically a LDT and therefore did not require a 510k, but they got it anyways.

http://www.fda.gov/MedicalDevices/ProductsandMedicalProcedur...

What would be an example of a test that fails, and how does this happen?

Examples: Not enough DNA in the sample (this normally comes from the White Blood Cells), the primers used in a PCR test (amplifying the DNA) do not bind properly, other markers for various tests don't bind correctly, etc. Not-binding correctly could be a problem with the whole sample (ie something in the blood is throwing off the test, outside contamination by a bad draw, not documenting draw date properly and sample is expired).

With solid tissue samples (often embedded in paraffin wax), there can be extensive problems with the sample being prepared incorrectly. These are normally prepared on-site at your doctor's office and there is a surprising amount of variation.

I'm not very familiar with blood lab tests, but I believe many of them are some form of chromatography, e.g. HPLC, which is "just" a high-tech version of pee-on-a-stick tests. For those, there is a "control" indicator that should always show positive, and failure is when it doesn't show positive.

It's incredible that he thinks that sort of labeling would be ok. "I'm selling you this drug, but the FDA hasn't agreed that it is effective." This is just not the way things are done.

> I'm selling you this drug, but the FDA hasn't agreed that it is effective.

This sort of labeling is already ok. Have you been in a drug store lately? A quarter of the cold medicine is basically placebo garbage. They literally have no active ingredients that are recognized by the FDA, but their cleverly worded boxes sit close to real cold medicine with active ingredients tested by science.

While you're on your drug store field trip, look for the homeopathic cold medicine and look for the active ingredients. Note how it's not measured in milligrams.

With supplements, it's different, thanks to Orrin Hatch and his "Dietary Supplement Health and Education Act". In that field you can sell next to everything, provided you label it correctly. Buyer beware! If your supplement contains any elixir of sulfanilamide it's your fault, you should have educated yourself on what this elixir of sulfanilamide actually is. After all, it contains what it says on the tin.

Diagnostics are thankfully more tightly regulated.

(Shrug) I don't see the problem with this. The opposite scenario, in which everthing that's not explicitly permitted is forbidden, is worse. That's how you get a War on Drugs.

Regulation of certain medical lab tests may be more reasonable. That seems like a case where informed consent on the part of each individual consumer is impractical.

The opposite scenario, in which everthing that's not explicitly permitted is forbidden, is worse

With the degree of information asymmetry in medicine it makes sense to regulate medicinal drugs tightly. Before anyone screams: "But FDA stifles innovation" let's say that the FDA exists since the 1960s and the pharma troubles started in the early 2000s.

An alternative theory is overfinancialization. Just the other day the merger between Dow and DuPont was announced. Why compete in the market when you can just buy up the competition? No expensive R & D needed!

One big difference is that placebo effect doesn't make any sense for blood tests.

You want to cure your cold with placebo? Sure, go ahead. But measuring cholesterol levels or detecting HIV with an inaccurate test is a real health hazard, and a liability for the company.

You are pointing a way forward for Theranos' investors to save their bacon: indemnify the lab testing service.

Within short order the market will offer two kinds of tests: Theranos for the plebs, Labquest for the patricians.

Zinc throat lozenges are a weird case: there's evidence that they're effective but they haven't passed FDA muster (and arguably don't have to, being zinc supplements under US law). Note that they are marked as homeopathic, but the active ingredients are not listed as having been diluted to "basically just water" levels.

The term "homeopathic" has been diluted. It used to mean "a lie dissolved in water", but now, because (ugh) it got popular in some circles, or to try to steal some of the relative legitimacy of natural non-water supplements, "naturopathic" remedies are labeled homeopathic.

I keep almost accidentally buying homeopathic teething gel.

First, that's not a great comparison since drugs are more on the active side of the spectrum. Second, the vast majority of substances do not need FDA approval before people can choose to consume.

A test result will result in active - and possibly bad - decisions. Like not seeing a doctor despite having that weird sensation.

This is why 23andme's health test results were prohibited by the FDA until recently. People were taking action based on it assuming it was a diagnostic test.

They really don't see a difference between disruption like AirBnB/Uber (the 'law breaking and getting away' part I don't agree with; the 'eff rude taxi drivers' agree wholeheartedly) and similar law-breaking 'disruption' in healthcare. When you go in with the mindset that 'wrong' laws _deserve_ to be broken, it can be hard to know when to stop. My money's on them using the money to lobby the FDA to change the laws, just like other 'disrupters' are doing.

Fortunately the FDA is quite used (and resistant) to lobbying from companies that would eat Theranos for breakfast if they wanted to. E.g. Pfizer had "Other operating expenses" in 2010 bigger than the entire much-hyped $9 billion valuation of Theranos.

Edit: if I was trying to shore up support for a company, I'd at least make sure the CEO doesn't look unhinged on business magazine covers. The one ~15% down in the article is just scary.

Edit2: From TFA: "She determined that a vegan, macrobiotic lifestyle would allow her to “train” her body to work all the time and to function on very little sleep, like an android." I'm quite sure "very little sleep" for extended periods of time is bad for your physical and mental health regardless of how macrobiotic your diet is.

SIR! I SAY SIR! WOULD YOU LIKE TO BUY THIS GENU-INE BOTTLE OF CALIFORNIA RATTLESNAKE OIL?! IT IS GUARANTEED* TO CURE WHAT AILS YOU!

*guarantee not FDA approved. Ingestion or skin contact with Rattlesnake oil may result in loss of life, loss of limb, cotton mouth, and spontaneous human combustion.

"My money's on them using the money to lobby the FDA to change the laws, just like other 'disrupters' are doing." - - Exactly. Why else would Theranos' board of directors be filled up with Senators and Washington power brokers instead of scientists and doctors... I mean, for cripes sake- Henry Kissinger was on their board at one point!! Wtf is he doing on the board of a medical device company.

Also on the Theranos board: three former U.S. cabinet secretaries, two former U.S. senators, a retired Navy admiral and a retired Marine Corps general. Ridiculous.

I think it would be good practice in this case to apply MLA rules to this block quote as it is rather confusing to read in it's current state. MLA says that within a set of quotations, any subsequent quotations will then be indicated by '' rather than "".

> leg crossed ankle over knee in a half-lotus manspread power pose. So, a woman with crossed legs is now engaging in manspreading? got it.

"manspreading" is a word?

http://bfy.tw/1hxu

One herpes test and 9 Billion valuation. According to Elizabeth Holmes its none of our business to worry about valuation, since the initial investors are sophisticated. Now, we know the initial investor is a family friend. She had 10+ years to put something out, and I am sure it is hard, but its about time, this farce be dismantled once and for all.

I work at a hedge fund and spend most of my time shorting stocks, especially healthcare/biotech/pharma stocks given the plethora of fraudulent companies in that space. I remember reading about Theranos several years even though it's private, and thinking there is no way this thing is not a fraud. I predict this will eventually be revealed as the biggest fraud ever in silicon valley. It just checks off way too many fraud boxes, based on my personal pattern recognition.

Relevant: On 27 October 2015, Holmes announced Theranos would publish data verifying the accuracy of its Edison testing device. They still haven't done so.

Given the intense scrutiny they're under, if they had the data, you'd think they would be eager to release it to prove their critics wrong.

Maybe it's the graphs on the bottom of https://www.theranos.com/our-lab ?

It's embarrassing that this controversy has been going on for a while and there's still no published hard data on whether this micro-sample thing works. Theranos needs to try some tests given widely, such as blood glucose or cholesterol, and compare results with the standard tests. Those can be tested on a general population.

If this works, it needs to be combined with Google's micro-gun and vacuum needle-less blood sampler. Or maybe Google/Alphabet will come up with more micro-sample techniques that work better than Theranos'.

I've been following this case, so I have to ask a stupid question: can't this be resolved in an afternoon? And one that could have happened years ago?

1) Take blood for the Theranos test.

2) Take blood for normal test.

3) Do they get the same results? If yes, Theranos has viability; if not, they've been scamming.

Confirming the accuracy and precision of a medical device isn't as simple as running a test in an afternoon. Interestingly, Jean-Louis Gassée did almost the same experiment[1] which was discussed earlier this fall[2].

1. http://www.mondaynote.com/2015/10/18/theranos-trouble-a-firs...

2. https://news.ycombinator.com/item?id=10408811

Not even close. Suppose you're testing for something that only occurs in 1% of the population or less. What if the Theranos test just never returns a positive? Then that test is "99% accurate" compared to the "normal" test.

Of course this is just a hypothetical, in real life you have to do a carefully controlled study - which is exactly what the FDA requires.

Perhaps if it's a binary test, but blood tests typically return a cardinal/continuous value; it's enough to verify that e.g. Theranos returns something close to 44 when the reference test returns 44.

No one's disputing that bad statistics can invalidate some comparisons.

> Suppose you're testing for something that only occurs in 1% of the population or less. What if the Theranos test just never returns a positive? Then that test is "99% accurate" compared to the "normal" test.

Understanding statistics is very important here: http://www.kevinmd.com/blog/2015/11/the-controversy-over-the...

Yes, of course: deliver.

My theory is that the company was originally developed for a military procurement play, for battlefield diagnostics or similar. Hence their board.

The Washington Post published a very interesting article earlier this month, in which they used email correspondence to reveal that Theranos pitched their system to the military, but there were concerns from a DoD official about it.[1]

>An official evaluating Theranos’s signature blood-testing technology for the Department of Defense sounded the alarm in 2012 and launched a formal inquiry with the Food and Drug Administration about the company’s intent to distribute its tests without FDA clearance

Holmes contacted General Mattis, who had been promoting Theranos within the military, and he tried to bypass the concerns. In the end, Theranos never demonstrated the system to the military. Mattis joined their board a year later.

>According to a Theranos spokeswoman, Mattis and Holmes met in 2011 at a Marine Memorial event. Mattis, busy overseeing the war in Afghanistan as commander of the U.S. Central Command, expressed interest in testing Theranos’s technology in combat areas, according to the e-mails.

>“I’ve met with my various folks and we’re kicking this into overdrive,” Mattis wrote to Holmes in June 2012. “I’m convinced that your invention will be a game-changer for us and I want it to be given the opportunity for a demonstration in-theater soonest.”

>He urged Holmes to call or e-mail him if she felt they needed to talk.

>A month later, frustrated by the military regulatory expert who had contacted the FDA, Holmes took him up on that offer.

>“I know how incredibly busy you are but thought it was the right thing to let you know the following,” Holmes wrote.

>She noted that a deputy director working for a division of the military that oversees regulatory issues and compliance had launched a formal inquiry to regulators without warning Theranos.

>“I would very much appreciate your help in getting this information corrected with the regulatory agencies,” Holmes wrote in the e-mail to Mattis. “Since this misinformation came from within DoD, it will be invaluable if this information is formally corrected by the right people in DoD.”

>Within hours, Mattis forwarded the exchange to military officials, asking “how do we overcome this new obstacle.”

>“I have tried to get this device tested in theater asap, legally and ethically,” Mattis wrote. “This appears to be relatively straight-forward yet we’re a year into this and not yet deployed.”

>The field demonstration Mattis was seeking never took place.

[1] https://www.washingtonpost.com/news/wonk/wp/2015/12/02/inter...

Strongly agreed on the company's likely intent to integrate monitoring into US military training and deployment kits... imagine the military benefit in knowing every soldier's blood panel and derived data such as stress level on a near-continuous basis--that's a real force multiplier. Worth $10K/yr/man easily.

MD here. The vast majority of soldiers are fit and healthy and will have completely normal labs 99% of the time. The times when they are not normal will be insignificant 99% of the time (had a few drinks last night? your GGT level will be slightly above normal briefly). This information is worth next to nothing to anyone.

Do soldiers remain so healthy-by-the-numbers in the field? I'm wondering if many data points would help to identify individuals on the edge of becoming unfit.

Oh .. and another thing.

I do point of care needlestick blood testing on people every day. I check haemoglobin and blood sugar levels. When doing a needlestick test for Hb I always do it twice because these results are often spurious (especially when you "milk" blood out of the finger). The Hb/liter figure can vary by as much as 20-30% routinely. If this is the case I then do a proper venous draw to find the true result. These tests are marginally useful in screening for really low or really high values, but they are pretty much inadequate compared to a proper venous draw for clinical decision making.

It's not a bad theory, but even the military medical establishment would demand that the tests are accurate.

The military invested in diving rods to detect mines in Iraq. The military is a great place to sell worthless junk at high prices, if you have connections.

Vaporware.

The coverage of Theranos just makes me think of the media as vultures trying to keep a story stirred up until somethingm anything, happens.

They do the same thing with politicians all the time.

Doesnt that serve a good purpose? Is it a good thing to keep investors tied for years in a fruitless endeavor instead of investing in something better?

And the headline should read, "Can Theranos CEO Elizabeth Holmes fend off the critical media critics (us) long enough to achieve something"

They've had a decade to achieve... Something.

They are currently selling their tests. The same tests that, according to third-party verifiers don't work.

I don't think media sharks are the problem, here.

They've been working on it for 10 years and have achieved nothing- then she went on a media PR blitz storm starting in 2013. Suddenly they're upset when not all their coverage is glowing.

Reply on news.ycombinator.com