Hacker News Reader

The first part seems probable (sell to the DoD), but intentional fraud is unlikely. Remember Hanlon's Razor - "Never ascribe to malice that which can be explained by incompetence."

It seems more likely they were very close on the science, and decided to run with it while they worked out the last 10%. That works great for most tech companies, but healthcare technology is very different.

Has there been any publicized and verifiable facts that they were close to having something that worked?

No. In fact all the independent data suggests that it is impossible with less than 4 drops of blood (~100µl) due to sampling issues. No matter how great the detection technology is if you are not able to get a representative sample then the end result will be garbage.

If 4 drops is a low bound, couldn't they have aimed for say 10? Can't you draw 10 drops of blood relatively easily from the finger? I feel like I've had that done before when donating blood. It'd still be a much less annoying process than vials after vials like it seems like most blood testing is.

It's hard for me to imagine they wouldn't have said "well, one drop didn't work but we could decently with four but we won't do that."

I haven't followed tremendously the story though.

It is really hard to get more than a single drop out of fingerprick. The reason why is the capillaries close up very quickly - if you are really lucky you can get 2 drops out, but anything more is a struggle.

The second problem with fingerprick blood is you pick up quite a bit of skin tissue in the process, but the amount collected is quite variable. What you end up measuring is blood + some unknown percentage of skin tissue derived material. None of these things are good for reproducibility.

A simpler theory might be that once the board picked up a couple of members in certain areas, it expanded along the social network of the board.

That's my read on it - the board members probably all know each other, or are friends-of-friends. Just like you and I, corporate board members recommend their friends when positions are open.

It probably doesn't hurt to have people like that on your board when you go looking for funding.

Being on a corporate board has got to be the easiest job in the world. When things are going well you collect a six or seven figure paycheck for doing almost nothing.

> intentional fraud is unlikely. Remember Hanlon's Razor

Hanlon's razor isn't proof of anything.

No, it's more like a prior.

Tillett’s Razor - ”Never ascribe to incompetence that which can be explained by corruption”

Specialist's Corollary - "To observers, corruption and incompetence are indistinguishable."

Unless you get to see the offshore bank balance.

Why does Hanlon's Razor exist when it can just as easily be put the other way around?

P(malice) <= P(competence & intent); P(stupidity) >> P(competence & intent)

That said, Hanlon's razor can't explain away evidence.

   if (benefit > 0) {
      P(malice) >> P(stupidity)
   }

Yup, when I make a mistake in my income tax filing and over pay by $2k that's incompetence. I wouldn't be so sure about attributing an unreported offshore shell company holding millions of dollars as incompetence.

More like they had a taste of the science and decided to work out the last 90%.. But that's actually the right way to do science generally.... Just not in this case.

When billions are at stake, sophisticated fraud becomes more likely though. I wouldn't dismiss the theory out of hand.

Hanlon's razor is used by the malicious to fool the incompetent.

I agree with your first section, but not the second. A fart can look close to getting to the moon given enough layers of incompetence.

I'm a little confused, does the military acquire companies wholesale? I've never heard of that happening, that's not to say it can't but what are some examples of this kind of scheme?

Instead was the plan to milk the military for a long time on R&D expenses and just pocket the cash as one goes along? If so how does that work for the VC-backed structure of the company?

I have no expertise in any of this, but I was assuming it would be something along the lines of an IDIQ (indefinite delivery, indefinite quantity) contract vehicle: https://en.wikipedia.org/wiki/IDIQ

That makes some sense, though it seems they'd have to deliver something in that case. I'm sure one could keep the charade going for a little bit by delaying/delivering stuff that didn't work, but eventually the docs in the army medical corp would start to raise red flags and the jig would be up. One hopes anyway?

The military pays as much, if not more, for contracts and products as the private sector would pay to acquire the companies involved.

I can answer that for you. No, the US military does not acquire private companies, except in extraordinary circumstances of national security (think: WW2 type events). They'd only do it if the company possesses something of extreme strategic importance and could no longer operate as an on-going concern, and the first thing they'd do is try to arrange a sale to another friendly defense contractor. Buying a private company would be the very last thing they would do in the chain of actions there; further, what they would do, is purchase the technology, not the corporation.

Accordingly, the theory about Theranos being set-up to be sold to the DoD, falls entirely flat.

>The plan appeared to be to use their connection and access to some rather “helpful” high brass to offload the technology for billions and then later sweep the fact that it didn’t work under the rug (this tactic has worked fabulously for the last 20 years). Avoid the FDA altogether.

they actually attempted exactly that back in 2012

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

IMO The Theranos basic concept is viable, but their implementation is flawed.

There have been somewhat recent cases of the SECDEF firing a General or Admiral if a development or acquisition goes awry. The JSF Program Executive Officer, MG David Heinz [0] was one example. I doubt a senior officer would risk a $10Bn acquisition program on unverified technology. Development programs are another story.

The Defense Contract Management Agency (DCMA) [1] are known for being rigid when it comes to the units meeting contracted specifications. They also require technical documentation, so trade secret wouldn't be an excuse. If DCMA are on vacation, then no deliveries take place. This happened during federal shutdowns.

If Theranos sold the DoD a device that was faked and didn't meet specifications, It's entirely likely DCMA would reject the delivery.

Development programs with DoD funding sometimes (often?) go awry, and sometimes specifications get changed (e.g F-35) but the unit must meet contractual specifications.

[0] http://www.washingtonpost.com/wp-dyn/content/article/2010/02...

[1] https://en.wikipedia.org/wiki/Defense_Contract_Management_Ag...

"IMO The Theranos basic concept is viable, but their implementation is flawed."

Can you point to any sources indicating that the basic concept is viable?

There is a market for low cost, minimally invasive lab tests.

Using a smaller needle can be done. A skilled doctor or nurse can use a butterfly needle [0] with a fine tip, and the patient barely notices. As I recall, a 27G needle is about the smallest and not damage the blood cells.

Microfluidic Arrays [1] are one way to carry out some lab tests with smaller samples. What steps are required to get the results from the array would vary between tests.

YC (W16) funded Unima [2] is working on reducing the cost of certain lab tests with paper test strips.

[0] https://en.wikipedia.org/wiki/Winged_infusion_set

[1] https://en.wikipedia.org/wiki/Microfluidics

[2] http://www.unima.com.mx/nsite/index.html

The plan appeared to be to use their connection and access to some rather “helpful” high brass to offload the technology for billions and then later sweep the fact that it didn’t work under the rug (this tactic has worked fabulously for the last 20 years). Avoid the FDA altogether.

While I'm sure this has happened in weapons technology, can you give an example in Health Care? (Where people deliberately avoided the FDA this way) I know there's a ton of waste in the military, I wasn't aware it spread this wide and deep.

The company didn't start as a way to draw blood for civilians but for a quicker/easier way to draw blood on the battle field...hence the original board composition.

That's not strictly true, the board composition was far more traditional prior to 2013, a lot of these government guys came on at that point. From the wikipedia article:

https://en.wikipedia.org/wiki/Theranos

In July 2013, the composition of the Theranos board of directors changed markedly, with departure of Channing Robertson (emeritus professor, chemical engineering, Stanford University),[39] experienced pharma and biotech executive Robert B. Shapiro (former chairman/CEO of the Pharmacia, Monsanto, and G.D. Searle group of companies), and financier Pete Thomas (principal, ATA Ventures).[6][38] Remaining from the original board were Theranos President and COO Ramesh "Sunny" Balwani and former Secretary of State George Shultz;

Let's give the powers that be a hearty welcome to the Age of Transparency. :)

I think you are cheapening the meaning of "fraud" if you think there is hundreds of billions worth of it going on in the government.

Medicare fraud is something like $60 billion or 10% of total spending. It's conceivable that the rest of the government would approach some multiple of $100 billion, even having lower rates of fraud than Medicare.

I think you need to read what the military have to say on this topic [1].

It is estimated that six percent of all procurement dollars are lost to fraud nation-wide annually. If you apply that amount to USAspending.gov’s figure of $700 billion dollars that the Federal Government spent on average each of the past nine years, that would total $42 billion dollars per year or $378 billion dollars lost to procurement fraud during the past nine years.

1. http://www.dodig.mil/iginformation/archives/finalwebsnips.pd...