USA vs. Ramesh Balwani: Lies, Damn Lies, and Misleading Questions

In a severed conspiracy case, subsequent defendants can see the prosecution's strategy. Over the past week, I've realized another defense benefit: witnesses testifying twice aren't as peppy and are thus less convincing, which compensates for the prosecution's ability to better prepare witnesses. Compared to the Elizabeth Holmes trial, almost everything in USA v. Ramesh Balwani seems as if it's happening in slow motion, and no matter how many times I glance at jurors, it always takes a few seconds before I accept they're really paying attention. 

Before Nimesh Jhaveri (from Walgreens) was sworn in, the parties debated a proposed mitigating jury instruction. Jurors do not see all the evidence in a criminal case, and limitations preventing bias also apply to questions. Consider a prosecutor asking, "How many times did you beat your wife?" in a burglary case. Even if the defendant truthfully answers "No," a mistrial is probable because the question itself has misled the jury, and no amount of re-direct can overcome the (false) implication. As Balwani's defense counsel explained, "The government must have a basis for every premise they show" the jury, so if it knows Balwani was  not aware certain Theranos devices were inoperable, it cannot ask questions of any witness in which it implies Balwani had such knowledge. Here, the question was whether prosecutors overstepped in their questions regarding the capabilities of some Theranos devices given to Walgreens. If some devices were inoperable with respect to certain blood tests, but Balwani didn't know those specific devices weren't working, can the government pursue questioning implying Balwani restricted access to the devices' location? Though the prosecution argued the defense could rebut any implications on re-direct, how does one disprove a false impression while retaining the 5th Amendment right against self-incrimination? Judge Davila didn't seem opposed to giving a jury instruction, agreeing with defense counsel's plea "to tell the jury the truth." 

Separately, lawyers discussed whether exploring Dr. Adam Rosendorff's post-Theranos' employment would be more probative than prejudicial. Within the Federal Rules of Evidence (FRE), Rule 608 can sometimes conflict with Rule 404. In layperson's terms, the ability to probe a person's reputation for (dis)honesty may obviate that same person's right to avoid propensity evidence. Think: you are a banker who told ten lies a day five years ago, and your CFO is on trial for accounting fraud. Should the defense be able to call ten witnesses who can attest to your reputation for dishonesty or at least ask you about twenty of your (alleged) lies? At what point does allowing a probe into your credibility lapse into impermissible character evidence? Conversely, at what point should a convicted felon with an incentive to lie--hypothetically speaking--be prevented from testifying so as to avoid poisoning the jury? When do credibility questions force a mini-trial about your entire life? And if your credibility is attacked, do you have the right to raise evidence of good character? By now, you should be able to understand why thoughtful lawyers answer most legal questions with, "It depends." 

Here, Dr. Rosendorff's post-Theranos employment happened to involve a company where an executive was indicted for fraudulent insurance billing. Some of that billing related to work done in Rosendorff's lab, but Rosendorff wasn't charged, nor was he alleged to be a co-conspirator. Even so, defense counsel argued the issue was Rosendorff's "reputational bias," i.e., a person investigated, even if not indicted, has an incentive to blame others. Why, then, shouldn't the defense be able to ask questions to see whether Rosendorff once again blamed others rather than himself for lab issues? Prosecutor John Bostic--who continues to shine--convincingly responded that post-Theranos, Rosendorff was terminated because he was unavailable, not because of fraudulent billing. Moreover, "every witness has a motive to protect their reputation," and because it only takes a few questions to cause an improper implication, the government would have to explore all the facts around every job held by Rosendorff. The defense responded by offering to limit its questions to the fact that Rosendorff's subsequent involvement in fraudulent billing meant he had an incentive or self-interest to lie or to blame others to protect his reputation.

Email, Sunny Balwani: "some will always be doubters--happens with every new technology."

If you're wondering whether an indirect connection with fraud within a non-attenuated time period is admissible to place a witness's bias and therefore credibility at issue, in Judge Davila's courtroom, you have an answer: no, because of FRE 403, i.e., it would be more prejudicial than probative. Judge Davila said he would allow the defense to probe bias, but not for the reasons stated; at the same time, if Rosendorff opened the door to post-Theranos issues based on his answers, the defense could inquire further. 

Both Rosendorff and Jhaveri testified in Holmes' trial, and their substantive testimony so far in Balwani's trial hasn't been much different. The only "new" information from Jhaveri's testimony was that Balwani claimed supply chain issues were a major cause of delays in the Walgreens contract. More specifically, a document referred to "learning processes around ordering patterns" from doctors, i.e. ordering data. (Neither artificial intelligence nor machine learning has been mentioned, but I'm curious whether Theranos was attempting to use either and in doing so, make its objectives more complex than necessary.) Another email stood out, Balwani to Holmes on October 16, 2015: "Our CLIA lab failed... all... levels."

A major reason Jhaveri's testimony has been less effective here is because Attorney Jeffrey Coopersmith has mostly limited Jhaveri to short answers that confirm information unfavorable to the prosecution. Coopersmith's use of emails to show Theranos and Balwani were making progress with Walgreens while actively resolving Walgreens' concerns went awry only once, when a number in an email was (presumably) out-of-context. Coopersmith still managed to argue the email confirmed his own understanding rather than Jhaveri's by asking, "That's not what the email says, right?" [Update: Coopersmith later prevailed when he used Jhaveri's prior statements against him.] 

Generally speaking, when Jhaveri--perhaps the prosecution's most well-rounded witness--attempts to add context, Coopersmith interjects with, "I'm going to interrupt you so we don't go off topic." When witnesses realize context is missing and ask questions seeking clarification, Coopersmith responds, "If it's ok with you, I'll ask the questions." When witnesses go off-topic, Coopersmith moves to strike their testimony, which, even if granted only partially, dissuades continued elaboration. 

Though Jhaveri's cross-examination isn't over, Dr. Rosendorff was introduced out-of-order. Rosendorff is a tragic figure. You want to like him, because if everything he says is true, then he deserves sympathy. Yet, he seems unable to avoid foot-in-mouth disease, even when direct-examined by the infinitely patient Bostic. Rosendorff is apt to say, after an already long answer, "I just want to add..."  When not seeking to match wits with defense counsel, Rosendorff seems blind to his own lack of charisma. In response to the question, "What is a lab director," Rosendorff explained a lab director "directs the lab..." My favorite answer was when he described the Edison as "a black box about the size of a Macintosh" to jurors who mostly look too young to have ever seen one.

What rankles me about Rosendorff is that some emails he wrote while at Theranos might have been drafted by an attorney. (Full disclosure: I've used similar tactics when representing employees, but when drafting such letters or emails, I included detailed facts allowing the company's HR department to investigate properly. Some employees lie or conceal information, and I've used letters or emails to HR or managers to trigger an investigation or a detailed justification.) There's nothing wrong with a diligent employee wanting to exit a dishonest or incompetent company, but Rosendorff's confidence in his ability to analyze law is interesting. When referring to an email dated October 16, 2014 in which Rosendorff referred to 42 CFR 493.801, Rosendorff explained, the "CFR clearly says..." Rosendorff must be unaware even lawyers joke that CFRs--Codes of Federal Regulations--are mostly indecipherable except by the lobbyists who wrote them.

Balwani's lawyers must have recognized Rosendorff's overconfidence, because they asked to show the jury CFRs as "demonstrative evidence." The request, made outside the presence of the jury, had a high bar because jurors are fact-finders, not legal experts. Bostic warned that delving into a lab director's legal obligations under CFRs risks confusing jurors about Balwani's actual charges. Moreover, a violation of civil regulations, even if true, doesn't excuse conspiracy or other criminal charges. In response, Coopersmith asked how he could show the jury Rosendorff's duties as a lab director without relevant regulations. Moreover, why should the defense be restricted to Rosendorff's own opinion of his own duties, especially when CMS's report--which the Holmes' jury saw--was based on regulatory violations? Judge Davila established a middle ground, ruling the defense could ask Rosendorff about his understanding of his duties as a lab director and could refresh his memory by showing him--but not the jury--the relevant regulations.

The main difference between Rosendorff's testimony here and in the Holmes' trial is that Balwani's lawyers have provided context. In Holmes' trial, so many emails from different years were introduced helter-skelter, I began wondering if I wasn't smart enough to follow everything. Last week, the timeline was tight and logical. Coopersmith showed emails from August 2013 in which Rosendorff raised "some medical and operational concerns about our readiness." Then, within minutes, Coopersmith introduced an email from Rosendorff to Balwani dated July 19, 2014 where Rosendorff confirmed work was progressing properly: "We have now reviewed the CTN image and all QC is in order." Is Coopersmith cherry-picking emails and understating the complexity of the different issues Rosendorff had to manage? Absolutely, but all Coopersmith has to do is raise sufficient reasonable doubt. He doesn't have to prove Balwani didn't commit wire fraud or didn't engage in a conspiracy--he just has to present Balwani as minimally sympathetic while raising reasonable doubts about the government's case.

From Holmes' case, closing arguments, defense

Coopersmith's job is made easier by the fact that Theranos was attempting to do something good for the public (i.e., lowering healthcare costs) while tackling difficult scientific issues. Coopersmith highlighted Vitamin D's complexity through an email from another lab manager to Rosendorff indicating Vitamin D testing may be common, but it's also complex: Vitamin D assays are "notorious for variability on even FDA-approved tests." This information was not presented well in Holmes' trial, because I got the impression Vitamin D was a test that should not have had the error rates Theranos experienced. 

Rosendorff's most self-damaging statement was that "under law, the lab director is responsible for everything." He meant everything lab-related, but left out the limitation. Rosendorff's most useful statements were that 1) Balwani was involved on a "daily basis" in laboratory operations, sometimes without his knowledge; and 2) he reported to COO Balwani (rather than Holmes). When he said COO, Rosendorff pronounced each letter slowly and deliberately while looking directly at Balwani for several seconds. Balwani had his head down taking notes and did not realize Rosendorff was giving him a death stare. The two men have yet to make prolonged eye contact.

© Matthew Mehdi Rafat (2022)

Bonus: both Holmes' and Balwani's lawyers have seized upon witness's varied reasons for quitting Theranos, which they claim is evidence of inconsistency and therefore a lack of credibility. If you're an employment lawyer representing a whistleblower, consider coaching your client to say, "I didn't leave for a singular reason, but when I saw x or experienced y, that was the last straw." 

Disclaimer: nothing herein should be construed as legal advice or a solicitation for clients. I am not currently accepting new clients in the practice of law.