Resources: Articles

Editor details the challenges of covering genetic testing companies that make dubious claims Date: 11/15/17

Charles Piller

By Joseph Burns

In December 2016, Charles Piller (@cpiller), the west coast editor for Stat, reported that a genetic test to identify patients who could be prone to addiction lacked a firm scientific basis. With an eye-opening headline, Called ‘hogwash,’ a gene test for addiction risk exploits opioid fears, the article raised important questions about the Proove Opioid Risk test from Proove Biosciences in Irvine, Calif.

Published Dec. 13, it was the first of four articles about Proove that Piller wrote over eight months that questioned the validity of the test and the company’s marketing practices. In this Q&A, Piller explains how he did the stories and how he and Stat reacted to charges that the reporting was inaccurate.

Q: How did you identify the topic, meaning how did you hear that Proove had a potential problem with “questionable” science?

Piller: Stat closely covers the opioid crisis as well as genetic testing. Proove claimed to have a vital solution to the opioid crisis – a simple test that could determine a person’s likelihood to become addicted. If true, this could be a wonderful tool for physicians. But we were skeptical, based on our understanding of genetic testing and the complex social dimensions of addiction. Leading experts on the genetic components of addiction suggested that the company’s claims were at best greatly overstated, and at worst “hogwash.”

Q: How did the first story about “questionable” science lead to what seemed like an even bigger story: how the test was marketed to physicians? Did a source tip you off about these sales practices?

Piller: After my first story, many sources – former and then-current employees – got in touch with me. Several had key insider information about the testing process, contractual relationships with doctors, company strategies and finances, and other aspects of the firm’s operations. After I was able to gain their trust, the sources supplied original documents that confirmed their claims of inaccurate and unreliable testing, apparent financial malfeasance, dubious research practices, and possibly illegal payments to doctors that some called “kickbacks.”

Q: Can you describe how you gained sources’ trust?

Piller: When it comes to trust, each source is unique. Some were pretty relaxed about talking, while many current and former Proove employees said they were terrified about being fired or sued by CEO Brian Meshkin, who has been litigious in the past. Others were concerned that if they spoke publicly, they would never see money owed to them by Proove. In general, for those who were providing documents or other information, I followed these practices, which resulted in gaining the confidence of many sources, named and anonymous:

• Reassure them that I’d never mention a source’s name to any other source, unless given specific permission to do so.

• Offer various levels of secure communications (such as private email and encrypted email and text) to meet their level of concern.

• Have the long view: Be willing to offer reassurance about confidentiality repeatedly as needed, while going for on-record quotes whenever possible.

• Help sources understand the importance of their cooperation for coworkers and the public, so they appropriately can feel good about coming forward.

• Press sources to be named, and walk them through each scenario to allay their fears if possible. But reassure them that I would never name them in a story without permission.

Q: How long did the reporting take and how did you do it?

Piller: I reported the two investigative stories over about three months, but did a couple of other stories during that time. It involved reaching out to dozens of employees via LinkedIn, email, or personal referrals, gathering and verifying the authenticity of documents, and double checking the veracity of claims made. I got numerous, very similar reports from doctors, medical assistants, former Proove executives and scientists, and others. None of them knew who else I was talking to, so it seemed clear that the highly consistent reports were credible. I was able to get plenty of on-record quotes – further evidence that my sources were not simply disgruntled employees with an axe to grind.

Q: Given that health insurers and consumers struggle to know which lab companies’ tests are based on questionable science, what strategies or questions to ask would you suggest other journalists pursue when doing similar stories?

Piller: This company claimed 97 percent predictive accuracy for its testing. Most experts scoff at such claims, given that the science of predicting behavior based on genetics and simple surveys is very much in its infancy. In general, such certainty is elusive for most genetic testing, so that should be a red flag for reporters. Insurance billing can be complex, sometimes allowing payments for tests that show a certain genetic result without that result having clear predictive or diagnostic value. Only experts in the field who are not conflicted by commercial interests can offer a credible reality check on these matters.

Q: Were you surprised by any of what happened after your reporting, the raid of Proove’s offices, the criminal probe, or the asset sell-off? Did you think any of these actions came in response to your work, at least in part?

Piller: This was an extreme case of a company with business practices that were obviously illegal or improper – or very close to crossing that line. Honestly, I expected law enforcement to get involved. I was surprised that Proove stayed in business even for a few months after my second story appeared, because at that point most of their doctors, key to their business model, cut ties with Proove. I know the feds were reading my work closely and were influenced by it, because some of my sources interviewed by those investigators said they were asked about it.

Q: How did the phone interview with Proove former CEO and founder come about? Did you call him for a comment about the asset sale or did he call you?

Piller: I contacted him for comment, and he’s a loquacious guy who wanted to get his views across. This was a little surprising, given that he would not talk when it really counted – at the time I asked him all the hard questions about his business practices.

Q: When Proove pushed back on your reporting, how did your editors and the publisher at Stat react?

Piller: A few months after my second story, a public relations firm for Proove sent a long letter demanding a retraction and various corrections. We were 100 percent confident in the truthfulness of our reporting, but naturally take such claims seriously as all journalists should. After careful review by myself, my editor, and our attorney, we concluded that there was nothing in the letter that in any way showed any inaccuracies or unfairness in our reporting. Shortly before we were going to respond to Proove’s PR agency, the FBI and HHS raided the company, and the letter of complaint about our coverage was withdrawn. Proove’s complaints on its website were similar to that letter, but we felt that our reporting spoke for itself, and chose not to reply to the website statement.

Q: How did you learn that the letter of complaint was withdraw?

Piller: The PR agency said it was withdrawing the letter, which it wrote on behalf of Proove. Later, Proove's attorney sent Stat a "cease and desist" letter that was meant to prevent us from writing further stories about the company. Our attorney reviewed both documents, of course, but they had no effect whatsoever on our coverage.

Q: Any advice for reporters who cover genetic tests or might want to begin doing so?

Piller: This field is rife with overstated claims and outright shysters. It’s a rich area for reporting. One tip: Look at the company’s clinical trials, if any. If the trials are large and financed by payments from patients or their insurers (as was the case with Proove), that could be a sign that the company is using the trial process as a way to disguise purely financial motives.

Q: Did any other publications do follow-up articles that helped your reporting or lead you to other angles to pursue about Proove?

Piller: No.

Q: Did you get any help from Stat colleagues such as other reporters or data experts?

Piller: Not for this story, aside from normal editing.