Proove tests of DNA captured by swabbing inside a patient’s cheek were so unreliable that many physicians disregarded the results. There was scant evidence, said the company’s former chief scientist, that the tests improved patient outcomes. (yokeetod/iStock)
For doctors, the brochure from a California medical laboratory sounded like easy money: $30 for every person enrolled in a study of genetic tests meant to help select the best pain medication for each patient. A typical physician could make $144,000 a year in “research fees.”
But the clinical trial was largely a ploy to boost Proove Biosciences’s revenues, and many of the doctors who signed up did no actual work, say current and former employees.
Proove has grown rapidly by tapping into the public angst over surging opioid addiction. It is one of many companies touting personalized DNA-based tests backed by little or no credible scientific data showing their reliability. That’s because a regulatory loophole has left huge swaths of the multibillion-dollar genetic testing industry largely free of government oversight.
A STAT investigation found that Proove employees stationed in physicians’ offices pushed unnecessary tests on patients — a practice called “coercion” by one former manager — and they sometimes completed research evaluation forms on behalf of doctors, rating the tests as highly effective when they weren’t. In fact, Proove tests of DNA captured by swabbing inside a patient’s cheek were so unreliable that many physicians disregarded the results. There was scant evidence, said the company’s former chief scientist, that the tests improved patient outcomes.
The company, located in two modern buildings in a burgeoning life sciences corridor in Irvine, Calif., also billed insurers in misleading ways to bump up payments, according to current and former employees and copies of insurance forms.
Promoting itself as the “leader in personalized pain medicine,” Proove claims its tests predict how patients will react to pain medicines and who might become dependent on opioids.
The company has enjoyed glowing television news coverage suggesting it can help combat the opioid crisis. Current and former employees and doctors who worked with Proove called such portrayals the product of hype generated by CEO Brian Meshkin, who lacks a background in science and is known for his consummate sales abilities and talent for self-promotion.
Meshkin is “in it for the money. He wanted to swab as many people as possible,” regardless of medical need, said Beth Stevens, a former Proove manager who left the company last year after clashing with the CEO. “He wanted people who had a cold to be swabbed for opioid risk.”
Meshkin declined an interview request and responded to only one of many written questions. In an emailed statement, he said, “Proove is acting within the confines of the law … [and intends] to follow both the letter and spirit of the law.”
But legal experts told STAT that the way the company pays physicians involved in two ongoing Proove trials — both key to the firm’s monthly revenues of roughly $2 million — might violate anti-kickback laws.
STAT began its examination of Proove late last year, after being contacted by current and former employees who had read an article about the dearth of evidence supporting the company’s flagship opioid risk test. The investigation included interviews with 12 current and former employees and four doctors who worked with Proove. STAT also obtained thousands of pages of internal memos, billing and insurance records, and reports of test results.
Separately, the FBI and the Office of Inspector General of the Department of Health and Human Services are investigating possible criminal wrongdoing by Proove, according to a former and a current employee who said they were recently interviewed by federal agents. They said the agents were focused on possible kickbacks involving payments to doctors. The FBI and HHS declined to comment.
The government has been concerned about companies marketing tests under the guise of research for several years. In 2014, HHS issued a fraud alert warning laboratories that some ways of paying doctors for “research” might violate laws against kickbacks for lab test referrals. The alert followed a spate of cases involving allegedly improper payments to doctors by laboratory companies.
In response to the HHS alert, Proove’s legal counsel drafted a memo for executives and sales staff stipulating that company practices complied with the law. Among other arguments, it said Proove compensates doctors only for research they personally perform. But current and former employees and doctors who worked with Proove have since disputed that claim.
Dr. Titus Taube, a family practitioner in Warner Robins, Ga., said in an interview that when he signed up as a Proove researcher, “It sounded like it was the wave of the future.” He said he quickly became disillusioned and concluded that the pay was for referring patients to Proove for tests because the research tasks, including filling out questionnaires and assessing the tests’ efficacy, were done by a Proove employee at his office.
If doctors who order tests receive improper financial inducements through a research program — such as compensation without working or payments based on the volume of referrals they make — the company and doctors might be violating federal and state laws against kickbacks or other fraud, said Dr. R. Gregory Cochran, a physician and attorney with expertise on those statutes. Promoting inaccurate or unreliable tests could also leave a company vulnerable to claims of fraud and abuse, he said.
Proove conducts “laboratory developed tests” — designed and used inside a single company or hospital lab — that can be sold as long as they are ordered by a doctor. Last year, the Silicon Valley blood-testing firm Theranos became Exhibit A of what can go wrong with these tests: Its claims of rapid diagnosis of dozens of diseases from a tiny vial of blood were exposed as overblown. Theranos is now the subject of a federal criminal probe.
The Food and Drug Administration warned in 2015 that some laboratory developed tests harm patients by leading to erroneous diagnoses and treatments. But the agency withdrew proposed regulatory guidelines for these tests late last year, deferring to President Trump’s yet-to-be named FDA commissioner and Congress to decide what, if any, rules to put in place.
“The current gaps in oversight … pose a challenge for the continued evolution of precision medicine,” undermining patient and physician confidence and potentially slowing the adoption of validated tests, said Andrew Fish, who directs AdvaMedDx, a trade group for leading diagnostic-test developers.
For now at least, Proove and similar companies can sell tests without having to show that results are valid and meaningful. STAT obtained Proove laboratory reports with gene-test results for six patients that included numerous incompatible clinical recommendations.
In one case — called typical by former and current employees — a report cautioned that a patient was highly susceptible to breathing problems when given the powerful synthetic opioid fentanyl, yet suggested a normal dose. The records do not show if fentanyl was prescribed, or if the patient suffered harm from the recommendation.
Christopher Coston, who managed Proove research assistants in western states until his job was eliminated last March, said “a lot of doctors looked at the reports and said, ‘I don’t think so. This doesn’t make sense.’”
Dr. Oleg Gavrilyuk, a San Diego physical medicine specialist, defended Proove. Tests of how his patients would respond to particular drugs validate his clinical judgment most of the time, he said. When they don’t, he ignores the results.
Ford created the first car,” he said. “Of course it was not perfect. Now Tesla is the state of the art. You cannot choke development based on the idea that it’s not perfect. It’s the first step.”
Proove’s first clinical study subjected to peer review, based on data from 134 patients, wasn’t published until late January in the Journal of Psychiatric Research. It found that doctors said most patients improved after treatment decisions guided by a Proove pain-perception test.
Dr. Eric Fung, a biotech executive who briefly served as the firm’s chief scientific officer in 2015 and coauthored the study, said in an interview that it emerged from an effort to mine Proove’s database of tens of thousands of patient records for evidence of some kind of testing benefit. He acknowledged that the lack of a comparison group, made up of patients who did not get the test, weakens the findings.
Fung added that he left the company over doubts about the utility of its tests.
“I could not find a good statistical or clinical benefit for these tests,” he said. “I didn’t feel comfortable with the science being done at Proove.”
A push to swab every patient
The questionable science hasn’t stopped Meshkin from building the privately held Proove into what Deloitte Consulting and Inc. magazine call one of the nation’s fastest-growing businesses. His formula is simple: recruit doctors as paid “researchers,” providing them with a financial stake in ordering tests for their patients.
Here is how it works, according to current and former employees: Proove conducts large trials — two in progress aim to test 150,000 patients — that ask clinicians to rate how much various Proove tests help patients.
Company sales representatives pitch research contracts to doctors, often targeting financially troubled practices that need a new revenue stream. They promise steady income for light work. Meshkin said last year that about 150 doctors had signed on, out of 400 who order Proove tests. But there has been substantial turnover as dissatisfied doctors drop out.
Proove places some of its 300 employees inside doctor’s offices, and instructs them to try to collect cheek-swab DNA samples from each patient and recruit them for clinical trials. Patients are informed that trial participants will not be billed for test fees, according to a current Proove employee. They also are supposed to tell patients their doctor has insisted on the tests.
Even patients without pain, seen at family medicine and OB-GYN practices, were enlisted for pain- and addiction-related tests, Coston said. It seemed illogical, the former regional manager said, but Meshkin became angry when research assistants failed to sign up every patient.
Many doctors “would test every single soul who walked in,” said Coston. “But they didn’t do research. The doctor is steadily getting paid, and did nothing to deserve it except ‘rent’ space” to Proove.
Research assistants completed all research tasks at least half the time, said Stevens and other managers who supervised the assistants. Asked whether doctors profited from work they didn’t do, she said, “If it quacks like duck and it walks like a duck, it is.”
Stevens said each research assistant had a quota of at least five cheek swabs a day. Documents obtained by STAT show they were offered escalating cash bonuses for meeting numerical goals above the quota.
Stevens called Meshkin’s tactics to push tests on patients who didn’t need them “coercion.”
To avoid disrupting the flow of patients into exam rooms, the research assistants fly through patient surveys and consent forms. This leaves some patients feeling bullied, several former and current employees said.
Some research assistants grew uneasy with the high-pressure tactics and quit. “People got tired of trying to get patients to participate who didn’t want to participate,” Stevens said.
Recruitment of patients became increasingly scripted and sales-oriented, said a current employee who spoke anonymously for fear of reprisals. Research assistants were told to avoid the word “study” when signing patients up for a trial, the employee said, after some objected to being used as “guinea pigs.”
The title “research assistant” was even changed to “patient engagement representative,” and Meshkin told managers to find attractive young people to fill that role, according to a former manager who spoke anonymously, citing a confidentiality agreement.
Crucially, Proove urges all doctors — whether participating in a trial or not — to sign a “standing order” for all their patients to get any Proove genetic tests, such as those for risk of opioid abuse, pain sensitivity, and response to opioids and other drugs.
Many doctors have been reluctant to sign. But by 2015 about 70 percent had done so under pressure from Meshkin, said the former manager. Examples of orders obtained by STAT authorize Proove research assistants to collect DNA from any patient under the doctor’s care.
Meshkin emphasized that under a standing order, the decision to be tested “is not up to [the patients], it’s up to the doctors,” Stevens said — that patients “don’t have a choice.”
Standing orders are common in medicine, although not in such circumstances. Doctors often issue standing orders to streamline care when patients need regular tests or treatments for a known chronic condition. Such orders rarely apply to a wide range of genetic tests for all patients in a diverse practice.
Standing orders “are often a red flag, particularly in a situation, like here, where the patients have varying diagnoses and needs,” said Cochran, associate director of a joint health policy and law degree program at the University of California, San Francisco, and UC Hastings College of the Law. “There’s absolutely a kickback concern if the doctor knows he’s going to get a check based on the number of patients who sign up and get tested.”
In his written statement, Meshkin said, “Proove seeks to ensure that each participating clinician independently determines the medical necessity” for tests, “and that every patient signs an informed consent to have the testing performed.”
Meshkin spins a good story
The man behind the Proove enterprise has an elaborate personal website, featuring flattering photos of himself and his family. Effusive language praises his character, business acumen, and civic accomplishments as a former school board member, and the site lists his accolades, including an “Entrepreneur of the Year” award from the Orange County Business Journal in Southern California.
“Volunteer. Public Servant. Entrepreneur. Coach. There are many words used to describe Brian Meshkin. But if you ask Brian, he’s just Brian — a guy who works hard, dreams big, and rolls up his sleeves to help others,” the site’s unsigned biography begins.
He also calls himself a “published researcher,” although he lacks any formal scientific training, and says nothing about the legal problems of a previous genetic testing venture.
That “nutrigenomics” company sold genetically customized dietary supplements through Las Vegas spas and directly to consumers. In an interview last year, Meshkin said he sold that company, Salugen, in 2009, soon after California authorities sent a cease-and-desist letter for improper marketing.
Former and current Proove doctors and employees said the site aptly captures a man who deeply believes in himself and grasps the power of spinning a good story.
Meshkin has been listed as a coauthor on study abstracts, Proove posters presented at scientific meetings, and the recently published study. His involvement in research was one of the reasons that Dr. Daniel A. Schwarz, Proove’s former research and development director, said he decided to resign.
A Michigan pain specialist, Schwarz said he left Proove in 2015 over concerns about the company’s lack of scientific integrity. “You can’t have a CEO trying to run the science who doesn’t have a science background,” he said. “It was marketing. It wasn’t science, and I couldn’t be a part of that.”
Schwarz said scientific control became a key issue when he began to doubt the validity of Proove’s work and wanted to repair the company’s research and development. Meshkin “wasn’t listening,” he said. “He didn’t take the advice, and I felt that he just wanted to do it his way.”
Schwarz also grew alarmed when coworkers said patient research records were not stored securely. Another former Proove executive, who requested anonymity due to a confidentiality agreement, said data sheets languished for long periods, stacked in heaps at Proove’s main office.
“The patient comes first, and if you have a product where you are not doing the right thing by the patient, I can’t be a part of it,” Schwarz said.
According to Stevens, company insiders hoped for vast improvements after Fung, a respected biotech executive who earned his MD and PhD degrees at Johns Hopkins University, was hired with fanfare in 2015.
“I remember one of the clinical people who worked underneath [Fung] said to me, ‘If he leaves this company, we all need to go,’” Stevens said.
Fung quit after three months.
‘Can I trust the data?’
Taube, the Georgia family doctor, said he saw firsthand how untrustworthy Proove’s tests and studies are. He signed on with Proove about four months ago after he was approached by a sales rep. Taube said he was intrigued by Proove’s ideas for advancing precision medicine and impressed by the support of Medicare and other insurers, described by Proove as study “sponsors.”
At first, he filled out the research forms himself, including an “investigator intervention evaluation” used to assess whether Proove tests helped patients or affected their care. But he said Proove managers told him, “No, no, [the Proove employee] will take care of it.”
This practice was widely adopted after some doctors filled out research forms partially or not at all, but still expected to be paid, said a former Proove executive who spoke anonymously due to a confidentiality agreement.
Taube then would sign the evaluations prepared by a Proove research assistant without giving them a close look. But he began to view the clinical trials as mostly marketing and looked more closely at the evaluations he signed. He was startled to see that Proove tests always received top ratings, though they were having little impact on his practice.
The results often were refuted by a patient’s experience, he said. For example, they showed that several of his patients should not take losartan, a blood pressure medicine, because they metabolized it poorly. But they had taken it successfully for years.
Taube recently stopped signing up new patients for Proove, saying he wondered, “Can I trust the data?”
His experience was typical, according to current and former Proove employees. They said Proove managers told representatives always to rate tests a 4 or 5 on a 5-point scale of benefit to the patient, even when a test had no bearing on clinical care.
Another former Proove executive, who also signed a confidentiality agreement, said even the principal investigator of a 50,000-patient Proove clinical trial — Dr. Daniel R. Kendall, a McLean, Va., osteopath for National Spine & Pain Centers — didn’t rely on the tests. The former executive said an audit of 1,500 patient charts from Kendall’s practice showed no changes in care based on Proove tests.
Via email, Kendall disputed that conclusion. He would not provide the audit and referred all questions to Proove. In his email statement, Meshkin said National Spine & Pain Center patients showed “significant improvement in treatment,” but did not provide any details.
Profit innovation: ‘unbundled’ charges
Clinical trials helped increase testing, but insurers often rejected claims, leading Proove to devise another way to pump up its bottom line: creative billing.
Insurers deem many Proove tests experimental due to a lack of evidence that they could aid diagnosis or treatment of the patient. One consequence of the denials, said some former employees, is that Proove periodically stops paying many sales reps and even fires them.
Proove also has failed to pay many physicians, including Taube, for research services — even doctors who actually did the work. Coston called it “the main thing I was getting yelled about by my doctors.” Taube said the company blamed slow insurance reimbursements.
By 2014, Proove could complete numerous tests using a single cheek swab. But insurance payments were less than what the company thought it could get for the same tests if they had been billed based on several different swabs, said current and former employees.
So Proove adopted a new billing strategy in 2015. Documents show that instead of submitting an invoice for multiple genetic tests from one date of service — when the patient’s cheek was swabbed — Proove “unbundled” the charges, dividing them up among several invoices keyed to the dates lab reports were issued. Current and former employees said this process has become the norm.
The company might bill insurers for its “opioid risk test” on Monday, for its “pain perception” test the following Monday, and so on — suggesting that the patient was tested on several occasions. Combined billings were often in the range of $7,000 to $10,000.
A Southern California doctor, who would only comment anonymously, said he kicked Proove out of his practice. “[T]hey billed the insurance four times more than what they told me they would bill initially. … Several patients of mine became very upset at the unreasonable amount of money that they were charging the insurance company,” the doctor said in an email. Such reactions were common among physicians, said current and former employees.Michelle Cavanaugh, an expert with the California-based medical billing firm Kareo, called unbundling “working the system.”
“What is the clinical justification for assigning it different dates?” said P. Thomas Hirsch, president of Laboratory Billing Solutions in Portsmouth, N.H., after hearing a description of Proove’s practices. “If they can’t give one, it’s just a scam.”
This story was originally published by STAT, an online publication of Boston Globe Media that covers health, medicine, and scientific discovery
Get the best of KQED's science coverage in your inbox weekly.