Trial By Error: The “Wired” Account of Cancelled Exercise Study for Long Covid at University of Virginia

By David Tuller, DrPH

On June 1st, Wired published a 7,600-word lollapalooza about mind-body interventions for Long Covid. The article, written by Alan Levinovitz, a professor of religion at James Madison University, posited that angry patients were derailing important research into these and other psychological and behavioral treatments. For balance, Levinovitz quoted multiple critics of these approaches (including me), but at key moments he stacked the deck in favor of his own interpretations. 

One example, which I mentioned in an earlier post, involved the discussion of the PACE trial and related research. Levinovitz ignored clear evidence of serious methodological flaws and presented these disputes as matters of “they said, they said.” In another example, he offered what appeared to be a slam-dunk case of a patient suppressing valuable research, in the process portraying an unnamed University of Virginia law professor as an unreasonable anti-science advocate. But he omitted a key detail regarding the study the professor was protesting, rendering their protests hard to comprehend. Moreover, he did not offer the professor a chance to explain their perspective, and then failed to shield their identity from the public.

Here’s the relevant section of the article: 

“If overcoming opposition feels too burdensome or controversial, researchers may simply abandon a trial. Siddhartha Angadi is an exercise physiologist at the University of Virginia who specializes in rehabilitation. He designed a long Covid exercise trial with a cardiologist and a pulmonologist, both of whom had experience treating patients with severe heart failure. (In rehabilitation medicine, ‘exercise’ means physical activity in service of rehabilitation, so standing up from a chair could qualify.) The IRB approved it. But then a UVA law professor and long Covid patient emailed him and the IRB with ‘grave concerns’ about the trial. There was ‘no way this would be approved if the dangers involved’ were understood, they wrote. Angadi and his coauthors got spooked. ‘We were like, this is just too much,’ he said. ‘So we bailed on it.’”

The proposed study was to feature what is called “high intensity interval training” (HIIT), which involves short bursts of vigorous activity followed by periods of more gentle movement or rest. But some or many Long Covid patients report serious and lengthy relapses after minimal amounts of movement, a phenomenon known as post-exertional malaise (PEM). Engaging in “high intensity” exercise could trigger PEM in such patients. Researchers have an obligation to inform study participants about possible or likely risks.

In discussing rehabilitation trials, Levinovitz offered only a parenthetical example of what “exercise” might entail: “standing up from a chair.” He did not mention that this specific UVA trial involved HIIT, which was the focus of the professor’s complaint and is different from “standing up from a chair.” In short, the article appeared to have misrepresented the context for the professor’s fears about “the dangers involved,” meaning these concerns were more likely to come across as exaggerated and histrionic rather than legitimate and rational. Wired’s reported version of the incident therefore served to support Levinovitz’ thesis. 

(Elsewhere in the article and in online exchanges, Levinovitz has argued that self-report of PEM is unreliable and doesn’t correlate well with objective data. In his view, therfore, screening for PEM in trials is difficult and can even undermine research, and issuing special warnings about it above and beyond general disclosures about overall risk is unwarranted. Others, of course, do not share this opinion.) 

The day after the article appeared, UVA law professor Megan Stevenson identified herself on X as the letter-writing culprit. She also posted an e-mail she had sent to Levinovitz. In the e-mail, she chided him for not having sought a comment from her so she could explain her views.

She further wrote: “You seem to suggest in the article that I got in the way of valuable research…While there are certainly valid medical questions about whether limited exercise within someone’s energy window is appropriate, there is 100% agreement among anyone who knows anything about long COVID that high-intensity interval training bears serious risk for some. At a minimum, failure to properly convey that risk to potential participants would be a serious violation of research ethics.”  
In extended online exchanges with Stevenson and others, Levinovitz doubled and tripled down, disputing Stevenson’s arguments and dismissing as absurd the suggestion that he and Wired did anything wrong. He insisted that her decision to intervene in the matter “harms scientific progress,” and he claimed the moral high ground in not offering her an opportunity to respond. As he posted: “Your name was redacted in the email I received, by the researcher, because it’s unethical to share the name. I didn’t choose not to reach out to you — I didn’t find out it was you until just now, when you emailed me.” 

Because he did not share her name, he rejected the notion that he had “outed” her. In his view, she outed herself. However, as Stevenson noted, she is the only law faculty member who has publicly revealed that she has Long Covid, she mentions the illness in her UVA bio, and a quick Google search would have immediately surfaced her name. Even if no one who identified her could know with 100% certainty that she had sent the letter, she’d be the obvious candidate.

Removing the fact that she, like the investigators, worked at UVA—an incidental detail not essential for purposes of the story—would have made it impossible for anyone to find her. This is basic journalism—delete unnecessary information to protect people’s privacy. Presumably no one at Wired, including editors and members of the magazines’s fact-checking department–conducted a search for “UVA law professor with Long Covid.” That seems like a remarkable lapse. (I have sent Wired some questions, but have not heard back.)

**********

Would contacting Stevenson have been unethical?

Levinovitz insisted repeatedly it would have been unethical to try to contact Stevenson in order to solicit a response. “[T]hinking it’s the journalists business to out a redacted author of an email who goes unnamed is…wild,” he wrote.

Others disagree with that assessment, among them Steven Lubet, an emeritus professor of law at Northwestern Pritzker School of Law, and an expert on legal ethics. (Lubet is a colleague and friend.) Here’s his view: “It can certainly be legitimate for journalists to seek out the identity of an unnamed person, unless they have promised not to, even if they have been granted confidentiality by someone else. That’s called reporting.”

In any event, Levinovitz didn’t have to try to find her himself or reach out to her directly. He could and should have asked his source—presumably Angadi, the investigator—to ask the letter-writer if she would be willing to be contacted. Very simple! Stevenson herself made this point on X. This is standard practice in journalism.

Stevenson noted in the online exchange that Angadi, in his response to her letter, informed her that the investigators had already decided to drop the project. If that’s the case, then her letter had nothing to do with their decision, in contrast to the article’s account. She suggested to Levinovitz that perhaps Angadi and colleagues had made that choice because they had not been aware of the risks and subsequently realized the study could endanger some participants. (I have reached out to Angadi to get his version of events, but have not heard back.) 

Levinovitz scoffed at that notion. “Megan, you really don’t know what you are talking about,” he wrote at one point on X. He further noted: “no, what happened is you drove extremely talented researchers — who know much, much more about cardiopulmonary rehab and risks than you do, and treat extremely at-risk patients — away from an underresearched condition…it sucks, and it harms scientific progress.” 

In a separate thread, he added this: “she’s a law professor though — if it’s such a big problem, she’s welcome to pursue legal action.” 

**********

The bottom line

The bottom line is that Levinovitz did not provide enough information for readers to accurately assess Stevenson’s concerns, did not seek out her views through the standard approach deployed by journalists everywhere, and then provided enough detail so that anyone could find her name in a quick search. He doesn’t see any of that as a problem. And Wired doesn’t seem to see it as a problem either.

Levinovitz and Wired owe Stevenson a big apology. And a right of reply, if she wants one.

**********

An important programming note

I have never hidden the fact that I crowdfund to support my position at Berkeley, and that most of the donations—which go directly to the university, not to me—come from patients. That is clearly a reportable conflict of interest. I have routinely been criticized for it and accused of being a shill—not least in this 2019 PACE propaganda piece from Reuters. (In that instance, as with the Wired article, I immediately alerted the publication that my academic position was not properly described.)

I assume this vein of criticism, as well as other accusations, might resurface in the course of this debate, along with the Reuters article itself. My response has always been this: Tell me what facts I got wrong, and I’ll correct them.