Theranos Trial: United States vs. Elizabeth Holmes, Day 8 (Three Card Monte)

Elizabeth Holmes' trial feels like a three-card Monte game in which the prosecution desperately tries to keep the jury focused. Holmes has the advantage because very few people understand a lab's inner workings, and fewer still the journey from chemical reaction to FDA approval. At this juncture, even if we concede at least one lab director was negligent or incompetent, insufficient evidence connects Holmes to an intentional cover-up. Problematically, the government has failed to provide a timeline describing 1) exactly what Holmes said; and 2) to whom she made knowingly false statements. 

Lodovico: "Where's that snake? Bring the villain forward." 

Othello: "I’m looking to see if you have cloven hooves like the devil. But that’s just a fairy tale. If you are a devil, I won’t be able to kill you." -- Shakespeare's Othello, Act 5 Scene 2.

The government's failure to identify Holmes' fraudulent statements has allowed her lawyers to spin a tale of outward kindness, hard work, and steady ambition. More specifically, by co-opting evidence of misconduct, defense attorney Lance Wade has dulled the prosecution's blows. Remember outliers manually removed from QC testing? A Theranos document today cited an "appropriate standard for outlier removal," proving Theranos was transparent about its QC process. Remember Theranos' inability to handle a majority of common blood tests? Wade showed a 2011 email in which Holmes is told progress is occurring on 95% of covered blood tests. Remember the government's allegation Holmes had zero in-the-field military contracts? On November 16, 2012, Holmes sent an email informing teams of an upcoming meeting with the Department of Defense. Meanwhile, Wade is taking every opportunity to argue had Theranos been allowed more time, it would have succeeded.

Lance Wade: "In R&D, sometimes you have to fail before you can succeed, right?" (September 21, 2021) 

Yet, everything Wade is introducing is irrelevant or incomplete. The DoD email? It doesn't prove a meeting or a contract actually occurred. The 95% of covered tests? Holmes was referring to assays in research and development, not patient testing (i.e.,  the CLIA lab, which must validate assays before implementation). The outliers? Inconclusive. No expert indicated an acceptable number of removals, and I never saw FDA approval of Theranos' QC processes. (Is the defense deliberately showing lab documents and patent applications out of order to confuse the jury and disrupt a cohesive timeline?)

When not introducing lab documents out of context, Wade continued blaming Ramesh Balwani and lab directors. Balwani's hard-charging, abrasive style was contrasted with Holmes complimenting employees, and a federal regulation confirmed the lab director--not the CEO--must ensure assays are qualified and validated. (I wouldn't be surprised to hear if a lab was on fire and Holmes knew about it, she had no responsibility to call the fire department.)  

2/27/2006 email, Elizabeth Holmes to Surekha Gangakhedkar: "Great job. This is beautiful." 

In seeking to make transparent Theranos' deficiencies while diverting blame from the CEO, the defense walks a fine line. It admitted much of Theranos' "real work" was validating assays on the Advia, a Siemens-made machine. In 2013, Theranos was developing "Advia immunoassays optimized for small sample volumes," i.e., fingerstick assays on a non-Theranos machine. Thus, when Holmes was soliciting investors on or before 2013, she knew Theranos was running blood tests on a foreign company's device. Did she disclose this fact to investors, including Walgreens and Safeway? If not, did she tell prospective investors Theranos used homegrown hardware for the majority of its patient blood tests? If so, why hasn't the government introduced such statements? Instead of direct evidence, the government appears to be promulgating a theory Holmes should not have solicited investors by marketing Theranos' blood tests as reliable and accurate when she knew the Edison was anything but. 

One can't help but sympathize with the government. Imagine a CEO claiming she's developing semiconductor chips that use liquid nitrogen to reduce heat by 50%. Then imagine learning the chips expend less heat because of a new vent design and a fan being reverse-engineered from a German company's existing product. A reasonable investor would assume innovation greater than reverse engineering an existing product, especially if the CEO is soliciting hundreds of millions of dollars. But what if the company argues its vent design was the first step in revolutionizing the fabrication process and, given more time, it could have displaced Intel Corporation and Advanced Micro Devices? How do you prove the company wouldn't have succeeded when its lawyers can show "real work"? How do you prove a negative?

On the issue of specificity, the government has not been entirely feckless. On redirect, it countered Wade by showing a top scientist raised reliability concerns directly to Holmes regarding the homegrown Edison 3.0 and 3.5; that any successful assay development reports required further validation before being useful; and that Theranos never reached the point of utilizing assay development in the Edison 3.0. In other words, at the time Holmes was soliciting investor funds, she knew Theranos' homegrown machine was incapable of producing reliable or accurate patient blood tests. 

To further disprove Wade's earlier claim of Theranos avoiding patient harm, the government called two witnesses, an Arizona-based nurse practitioner and her client. Their testimony confirmed Theranos' fingerstick blood tests were wildly inaccurate in October 2014. Theranos' HCG values--which indicate pregnancy--not only varied dramatically compared to traditional lab results, but also generated the exact same value in two tests taken two days apart, except one value misplaced the decimal point by a factor of a hundred. To its credit, Theranos sent a corrected report several days later, but the damage was done: an expectant mother who'd suffered three prior miscarriages was told she might not be pregnant. Theranos followed up, blaming "post analytical human errors" and saying errors "are extremely rate" [sic]. Oddly enough, the same office--but not the nurse practitioner--allegedly ordered 300 more tests from Theranos in 2015 and 2016, but the data is incomplete: we don't know whether those later tests were run on the Advia or the Edison, or whether they used the fingerstick method.

At the end of Day 8, the question remains: where should we draw the line between ambitious self-promotion and fraud? If the government does its job, we--and future entrepreneurs--will find out. 

© Matthew Mehdi Rafat (2021) 

ISSN 2770-002X 

Bonus I: the following example might help delineate the issues. During a break, a cameraman was conferring with E.K., a local law school professor. I've seen E.K. lecture and do not believe her expertise is relevant to Holmes' case. However, my opinion is not definitive. Another person may view E.K.'s profile and deem her experience useful and valuable.

In a free market system, E.K. is encouraged to self-promote to gain attention and status--within limits. Can she advertise as a "criminal law expert" to gain airtime relating to the Holmes trial even if she lacks wire fraud experience? Yes. What if she claims extensive criminal law experience despite not having participated in a federal criminal trial in decades? More pertinently, assume she advertises herself as a "federal criminal expert"; has never litigated federal charges; and then afterwards, is hired in federal court as an expert witness and is scheduled to second-chair a criminal trial. Did she defraud anyone?

Bonus II: if you're interested in law school, don't assume law schools hire the best professors. Tenure and internal politics are as contentious in the ivory towers of justice as everywhere else. Personally, I don't recommend my alma mater, Santa Clara University Law School. My favorite professor moved to Florida State University College of Law, and other than his class, I can't remember learning anything useful.