Thread by @eringriffith

🩸 Back in San Jose for another day of testimony in US v Holmes.

I have LOUD TYPING ANXIETY (theverge.com/platform/amp/2…) so I am sitting in the overflow room this morning.

The lawyers are once again fighting about upcoming testimony.

Fun times, we still have not started because of some kind of technical problem.

Oh my god we are FINALLY starting.

They're projecting the exhibits onto a wall, which is not really coming through in the overflow room. A juror had to run and get glasses.

The judge says he's "very embarrassed" that the courtroom is having these problems and apologizes.

The defense is continuing its cross-exam of Lynette Sawyer, who was a co-director of the Theranos lab for 6 mo.

Sawyer had testified that she was frustrated that Theranos did not give her a list of the lab employees.

Lance Wade, lawyer for Holmes, pushed her to say she only asked for the list once.

We also discuss an email where Sawyer agreed to extend her time at Theranos (meant to undercut her comments that she quit over concerns about the lab).

Strange exchange with former Theranos lab director abt nanotainers.

Wade: What is a nanotainer?
Sawyer: I don't know
W: Did you understand it was the small sample vial that Theranos used at the time?
S: No.
W: Did you ask Mr. Balwani any questions about the nanotainer?
S: No.

talkin bout QA QC LDTs (quality assurance quality control lab developed tests)

Wade asks Sawyer if she signed off on an LDT for finger stick testing. "Had you ever seen such an elaborate and specific procedure?" Sawyer says no.

Wade asks about "an elaborate process where they warmed finger and noted places to draw from."

Sawyer: "That's pretty standard for finger sticks."

Again we discuss that Sawyer agreed to stay at Theranos 6 weeks after her initial contract.

Apparently the paperwork had been processed incorrectly and she was still listed as the lab director for a bit and Daniel Young reached out to her about it.

That error was discovered when Theranos was preparing for an audit of its Clia lab. Sawyer says she wasn't sure if it was in prep for one or as a result of one. And with that Wade is done.

Redirect and re-cross were extremely short and Sawyer is excused.

We're breaking and another former lab director - Kingshuk Das - is up after lunch. I will be getting a ticket to go to the courtroom so I can actually see the projected exhibits!

Kingshuk Das, a Theranos lab director, has taken the stand.

Lowering the lights to show exhibits on a projector in the courtroom feels a bit like jr high science class.

We see CLIA Lab rules that shows CMS (fed regulators) can revoke CLIA certificates for rule violations.

Defense is chiming in along the way in opposition of this document; Judge shoots it down as it is already in evidence.

Wow, Das testifies that when he started as lab director, he was not aware that of Theranos performing any tests that it had developed in-house. No tests on Edison, Minilab or TSPU. No test on finger stick samples.

(He joined officially in March 2016, longggg after Theranos had publicly launched.)

Dr. Das was working on a CMS audit while at Theranos. We discuss a "statement of deficiencies" that was 121 pages long.

In Nov 2015, CMS did an onsite survey of Theranos's CLIA lab. In January 2016, CMS sent Theranos a notice with the heading: CONDITION LEVEL DEFICIENCIES - IMMEDIATE JEOPARDY

There's a list of "conditions" that were not met, including “analytic systems” meaning the instruments and methods used.

Das testifies that he discussed the issues with Holmes and the Theranos team.

The CMS report shows that in numerous months between 2014 and 2015, a number of types of tests showed that 15% of tests were violating a common quality control rule.

Das testifies that he explained the issues to Holmes using the example of female patients returning abnormal PSA levels (a prostate test) because "females should generally not have PSA detectible."

Das testifies that he had to respond to CMS with a patient impact assessment out of a regulatory, professional and ethical obligation.

The response includes such phrases as “High rates of 102s QC rule failures” and “DC CVs far exceeding limits for a stable testing process" which Das simplifies by saying:

“It means there was a lot of imprecision noted.”

The report goes further, noting "a possible patient impact for every test reported from the lab’s TPS 3.5 instruments.”

"The fraction of patient results truly impacted and the nature and magnitude of any effect are unknown."

Das testifies that Theranos voided its tests from that 2014-2015 period. He explained to Holmes that "these instruments were malperforming from the very beginning."

Holmes suggested they give CMS an alternate explanation that the problem was "not an instrument failure but a failure of the quality control / quality assurance program."

Das disagreed - "the validation data had no bearing on the QA program" - and Theranos never resumed testing on those machines.

He concludes: "I found these instruments to be unsuitable for clinical use."

*TSPU machines (from above). *not TPS*

Of course there is also a TPS (Theranos Proprietary System) referred to in this CMS report. But the machines were TSPU. Also Edison. Also Minilab.

Das testifies about something called a "10x warning" on Feb. 25, 2015, when results come in either above or below the proper range 10 times in a row.

He says Theranos continued to report patient results after the 10x warning.