I read Bad Blood, John Carreyou’s account of the Theranos fiasco over Christmas and must say it changed my conception of what went down there.
I’d written a fair amount about Theranos over at Quora (see here), and was clearly much too sanguine as to what was going on. I thought they were just at the outer range of typical biotech startup behavior – the hype, the overpromising, the willful self-delusion by upper management and the ostracism of those in the lab and middle management unwilling to drink the corporate Kool-Aid. None of that stuff is okay, and it certainly isn’t pretty, but it’s not that unusual and doesn’t mean that a company has developed nothing of value.
Boy was I wrong.
Theranos was more a gang than a biotech. The greed, the demands for absolute loyalty, the cultivation of powerful protectors, the punishment of defectors–the culture there was far more Mafia than startup. No one got sunk in the Bay or strung up in Big Basin, but financial ruin through legal action was used to compel obedience and ensure silence.
I also underestimated just how many people used Theranos’ test results, and thus underestimated the likely medical harm they caused. This paper in JCI showed that Theranos’ results were sloppy but not wildly so; Carreyou reports instances where Theranos put patients at serious risk.
The only criticism I have of the book is that it is weak on the science and on the business. The basic premise of Theranos clinical value was that (1) it could use fingerstick blood to run most blood tests and (2) lowering the barrier to getting blood test information would improve healthcare outcomes.
All reference values for blood analytes are based on venous draws. Fingerstick blood tends to be diluted with lymph (water, really) – just think how your hands swell during a walk on a hot summer day. That means that analytes are both more dilute and more variable than venous blood.
This is the principal barrier to using fingersticks–not miniaturization–and it has been known for decades. The first question any reasonably diligent investor (or reporter) should have asked Theranos is how they solved this problem when no one else has. Carreyou does get around to this point, but not until well into the second half of the book.
The second question is whether Theranos would have done more harm than good if they were totally legit and delivered what they promised. Few blood test results are meaningful absent other clinical information. And few patients are capable of interpreting a result in the context of reference ranges.
The foreseeable result is that people will be frightened for no good reason and seek (and demand) unnecessary medical care. This is already happening with the Apple Watch EKG function.
But these really are minor quibbles. Carreyou did an incredible job of stitching together what are in truth disparate accounts and incidents into a compelling narrative. That is no small feat for any writer.
And Carreyou has laid bare just how shallow and meretricious most tech journalism is. Journalists are so hungry for facile narratives that they are easily manipulated by charismatic personalities with a good story. This went on for years with Theranos.
It’s going on today in other fields, particularly the breakthrough/hero narratives that form around new cancer therapies. We’ve seen one of those heroes exposed for the self-dealer that he is; others are sure to follow. The heroic narrative justifies and enables the enormous sums of money spent on cancer therapeutics (as opposed to cancer prevention). That much money is bound to result in intellectual, clinical, and financial, corruption. Don’t be surprised when these stories start coming out.