Theranos’ former lab director has become a central focus of early questioning in Elizabeth Holmes’ criminal trial. Late last week, the defense continued its cross-examination of one of the prosecution’s key witnesses, attempting to paint Dr. Adam Rosendorff as unreliable.
Three facts could damage Rosendorff in the eyes of the jury. Whether those facts diminish his prior testimony depends largely on Holmes’ defense team. Given the damning facts Rosendorff laid out during the prosecution’s questioning, that will be a difficult task. But discrediting key witnesses is something all defenses have to attempt.
Three lines of questioning
First, Rosendorff admitted under questioning by Holmes’ attorney Lance Wade that he had approved of the company’s controversial pseudo-policy regarding outliers. While Theranos had no official policy for how to eliminate outliers—despite there being ample guidance for how to do so—the company did develop a framework called the “six tip policy.” Each sample produced six results, and from that, two were discarded and the remaining four were averaged. The policy says it was OK to eliminate two, but it didn’t help lab techs identify which are the best candidates for omission. So, this is kind of a policy, but it’s one that didn’t cover everything it should have.
Second, Rosendorff seems to have had a stronger hand in developing and implementing lab protocols than his earlier testimony let on. For one, he had signed off on an “alternative assessment procedure” for Theranos’ proprietary Edison devices, which are essentially a set of proficiency tests that differed from those used to validate other diagnostic equipment. And he admitted that lab tech Erika Cheung hadn’t been using Rosendorff’s alternative procedure when she raised concerns about the Edison devices failing quality control.
Rosendorff also worked with Ramesh “Sunny” Balwani, Theranos’ president and COO, to edit a slide deck that was intended to instruct lab techs on the company’s unique proficiency testing procedure. “Theranos tests have no peer groups,” one slide said.
Third, Rosendorff appears to have had the ear of upper management for most of his time at the company. About a month before he quit, Rosendorff had been invited to a meeting with Balwani and other executives to voice his concerns about Theranos’ testing procedures. Rosendorff admitted under Wade’s questioning that there had been no restrictions on what he could say at that meeting.
Out of the loop
Together, those three lines of questioning could undermine Rosendorff’s previous testimony. But other parts of Friday’s testimony appeared to reinforce his earlier statements.
For one, Balwani sent an email to a Theranos vice president in June 2014, several months before Rosendorff quit. The email said, “Adam came by my office EXTREMELY frustrated that as a lab director he is not being kept in the loop.” Furthermore, the jury heard about backlogs for pregnancy tests, which Rosendorff said he was not aware of. Those more or less confirm what Rosendorff told the prosecution—that he had been left out of the loop.
Another of Wade’s tactics may not have landed as he intended. In a series of new emails, we learned that Rosendorff was slow to follow up on complaints. On two separate occasions, the lab director took more than a week to follow up with a doctor about questionable test results.
Now, the jury could interpret this a few ways. Jurors could see Rosendorff as unreliable, which is probably what Wade wants. But they could also take a more sympathetic view—after all, Rosendorff seems to have been dealing with more complaints than a normal lab. On top of that, he was constantly sparring with executives about how to run the lab. Given those fires, it’s easy to see why he might have pushed “reply to customer complaints” further down the list.
Taking a week to address a complaint may not be normal practice in clinical labs, but Theranos was clearly not running a typical lab.
Listing image by Jane Tyska/Digital First Media/The Mercury News
https://arstechnica.com/?p=1800306