A Theranos whistleblower speaks out

I've refrained from commenting on Theranos on the blog, because I haven't had much to say other than "Gee, it's not clear to me what the heck Theranos was actually selling" over and over and over again. But I do want to recommend that you read John Carreyrou's long piece about one of his sources, Tyler Shultz, former Theranos employee and a grandson of George Shultz (one of Theranos' board of directors): 

...The younger Mr. Shultz and Ms. Holmes met in late 2011 while he was visiting his grandfather’s house next to the Stanford campus. Tyler Shultz was a junior at Stanford majoring in mechanical engineering. Mr. Shultz interned at Theranos that summer and went to work there full-time in September 2013. He had just graduated after changing his major to biology to better prepare for a career at the startup, he says. 

The new employee was assigned to the assay validation team, which was responsible for verifying and documenting the accuracy of blood tests run on Edison machines before they were deployed in the lab for use with patients. 

Mr. Shultz says he found that results varied widely when tests were rerun with the same blood samples. To reduce that variability, Theranos routinely discarded outlying values from validation reports it compiled, he says. 

One validation report about an Edison test to detect a sexually-transmitted infectious disease said the test was sensitive enough to detect the disease 95% of the time. But when Mr. Shultz looked at the two sets of experiments from which the report was compiled, they showed sensitivities of 65% and 80%. That meant that if 100 people infected with the disease were tested only with the Edison device, as many as 35 of them would likely incorrectly conclude they were disease-free. 

A few months later, Mr. Shultz moved to Theranos’s production team, where he quantified by how much patient tests should be allowed to vary during daily quality-control checks. Under federal rules, labs are allowed to set those parameters on their own within the bounds of accepted industry guidelines. 

He says he noticed Edison machines often flunked Theranos’s quality-control standards. He says Mr. Balwani, the No. 2 executive at the company, pressured lab employees to ignore the failures and run blood tests on the machines anyway, contrary to accepted lab practices. 

Mr. Shultz says he took his concerns directly to Ms. Holmes. When they met in early 2014, she encouraged him to talk to Daniel Young, a Theranos vice president in charge of biostatistics. 

According to Mr. Shultz, Mr. Young said the differences with the sexually-transmitted infectious disease test occurred because some results fell inside an “equivocal zone,” meaning they were unclear at first but clarified later through other methods...

There's a lot more to the story, including quite a bit of intrigue with lawyers and his family. Read the whole thing.

One of the funny things about life is how early and often ethical dilemmas come. It seems to me that the younger Mr. Shultz acquitted himself well.