We’ve touched on some of the many disturbing things that happened during the clinical trial on which Dan Markingson committed suicide. In my first post, I asked how a psychotic, homicidal patient who was involuntarily hospitalized in a psychiatric hospital could give an informed consent for participation in a clinical trial. There appeared to have been abuse of a vulnerable patient and extraordinary coercion—participate in this trial or be committed to a psych hospital seems to have been the bottom line.
In my second post, we looked at investigator responsibilities, delegation of authority, and Good Clinical Practice tenets, all of which were violated with no consequences.
Now we turn to the need to disclose conflicts of interest (COI), again a basic clinical research ethics principle that was violated. There are so many obvious conflicts of interest that it is hard to know quite where to start.
The most obvious and egregious COI was that shown by Dr. Stephen Olson, who acted as both Dan Markingson’s treating physician and as Principle Investigator on the CAFÉ study. As Dr. Harrison Pope, a Harvard expert, concluded in his testimony, Olson “failed to meet the standards for good clinical practice both as a principal investigator and as the study physician for Mr. Markingson.” He failed his ethical responsibilities to Dan by:
—Enrolling him in a clinical trial he was incapable of consenting to, particularly over the objections of Dan’s mother, Mary Weiss.
— Not dropping Dan from the protocol when Dan was clearly not showing improvement.
— Not examining Dan closely and regularly. “Dr. Olson’s signature appears only twice throughout all of the study documents” from 12/8/2003 until his death 5/8/2004. (p 30, Pope testimony). According to Ms. Weiss, Dan had told her “that Dr. Olson would, at most, ‘stick his head in occasionally at study visits. The medical records confirm that impression.’” Olson likely would have had to examine Dan more frequently if he were not on a study, as he would not have had the additional staffing funded by the study to have a surrogate see his patients.
—Improperly acting as both Markingson’s study physician and treating physician
The conflict between Dr. Olson’s roles as Dan’s personal physician and as researcher is further detailed by Dr. Pope. For example, it likely kept Dan from getting a second opinion and from getting additional medications. The protocol also prohibited Olson from measuring blood levels of the anti-psychotic, which would have shown if he were being compliant in taking his meds.
Olson reportedly mislead the IRB, leading them to believe that Markingson’s enrollment had been agreed to by his case manager, when he hadn’t yet been seen by Mr. Pettit. He also failed to inform them of information about Dan’s inability to consent and that the judge had granted a stay of commitment for Dan if he followed recommendations.
The NIH website summarizes the problem succinctly: “If an investigator is also the personal health provider of the potential research participant, there may be an additional conflict of interest. A physician's duty is to honor the best interests of the patient. An investigator must do what is best for the study. These two objectives are not always consonant. Further, potential participants may be reluctant to question the advice of a health provider on whom they depend for care.”
Olson’s involvement, and that of the department chairman and coinvestigator, Dr. Charles Schulz, appear to have been significantly driven by financial interests, though they had other incentives, too, such as increased prestige and publications.
Financial Conflicts of Interest
There were major financial incentives for both Drs. Olson and Schulz, as well as for the university.
Dan Markingson—aka Subject 13—was worth over $15,000 to the university, had he completed the study. The CAFÉ study not only yielded $327,000, but helped generate more attention (and probably more trials) for the university’s schizophrenia program.
This particular study was structured so that patients had to complete the trial in order for the full payment to be received by the institution. Thus, dropping the patient because the study drug was ineffective, or adding additional medications, was prohibited and would have resulted in significant financial penalties—and this was not explained in the “informed consent.” Because the study period was one year, Olson had to keep Markingson on the trial to get the full payment; I suspect this contributed to his recommendation that Dan’s stay of commitment be extended to keep Dan a captive participant.
Dan was far more valuable to the University and psychiatrists as a study patient than as a regular psychiatric inpatient. As Dr. Pope noted, above, ancillary personnel could do many of the assessments. And Dan was destitute, so any care they provided him outside of the study would have had minimal reimbursement—we often didn’t feel that the payments from Medical Assistance even covered the cost of billing, let alone the medical care provided. (Note: not all of the grant money goes to the investigator. Much of it then has to be disbursed to the university for direct costs and for their administrative overhead charge, often 50% of the total grant. Physicians are also expected to generate monies to cover much of their salaries.)
According to the Pioneer Press in 2008, “Olson received $220,000 from six companies since 2002, including $149,000 from AstraZeneca, according to the state records. Schulz received $562,000, including $112,000 as a researcher and consultant to AstraZeneca.”
Yet these significant COIs were not disclosed to study participants. When Dr. Schulz replied to anguished Ms. Weiss’ 3rd letter (her first two went unanswered), he did not disclose that he was a principal investigator, nor did he forward her serious concerns to the IRB.
I was shocked to read Dr. Schulz’s testimony that he hadn’t even read the consent form for a study in which he was a co-investigator (and was listed on the FDA Form 1572 as being a responsible party). (p 156 Schulz deposition).
“At another point, plaintiff attorney “Barden read an excerpt from a bioethics book arguing for the importance of informing patients about a doctor's financial ties to drug companies.
"Have you had any training in biomedical ethics?" pressed Barden.
"I've taken the courses at the University of Minnesota that are required for us to participate in clinical research."
"And isn't this part of that training?"
"I'm not aware," said Schulz. "I don't recall that."
Contract research organizations (CROs) are intermediaries hired by a sponsor (generally pharmaceutical) to administer a clinical trial. This used to be done by the pharma company itself. As business has gotten more competitive, they generally farm that work out now, to avoid having to have their own staff. Quintiles, responsible for the CAFÉ study, is the largest CRO, capturing 14 percent of the lucrative $11.4 billion global market.
CROs put a lot of pressure on sites to enroll. Delays in completing the approval of a new drug cost from $684,931 to $1 million per day and slow recruitment is the major factor. So they try to motivate their sites to enroll patients, sometimes with encouragement (holding them up as examples to other sites or similar recognition) or promises of further studies, sometimes by threats of having the trial taken away. They employed both tactics to manipulate the study coordinator, Jeanne Kenney, illustrated by their e-mails here and here.
While CROs can often provide helpful tips from their experience doing trials, I was flabbergasted to read, “In the CAFÉ study, for instance, a Quintiles study monitor suggested that each of the CAFÉ study site coordinators try recruiting subjects at homeless shelters.” This is so coercive and unethical, it almost defies belief. . .but the longer I am in this field (and in medicine), the more disillusioned I have become. I should no longer be surprised by anything, I’m afraid.
It’s not just individual researchers who benefit from industry-sponsored research. The entire university feeds happily at the trough. At University of Minnesota alone, “In fiscal 2010, business and industry sponsored $35.4 million in research spending at the U -- 5.4 percent of total research expenditures of $653.6 million.”
The CAFÉ study initially suffered from poor enrollment—so much so that Quintiles, the CRO administering the study, had placed the site on probation and threatened to terminate it.
So the university established a 16 bed specialty psychosis unit at Fairview hospital, known as “Station 12,” “in part to enhance the hospital's startup schizophrenia program and meet the U's mandate to bring in more research dollars.” According to a CAFÉ Study Coordinator Teleconference, on this unit “All patients are reviewed for possible research candidacy. Research staff are in contact with nurses, case managers, and attending psychiatrists daily Research staff attend morning report before inpatient rounds take place. The focus is to identify any possible subjects that may be eligible for studies. This is also focused on building a repertoire [sic] with psychiatry residents, who are often much easier to approach than attendings.”
So we have a University department chairman demonstrating leadership by being unaware of basic research tenets and by saying that disclosure of financial COIs would just cause “confusion.” What an exemplary role model.
Of course, there are other COIs and pressures as well, in that successful enrollment begets publication, more fame, and more research grants from other sponsors as well. It also seems to bring considerable protection from those in power at the university.
The IRB holds a fair measure of blame in this tragedy, as well, though they have admitted no wrongdoing. The IRB is paid by the study to review the protocol and consent and, in theory, to provide oversight.
IRBs have had their own share of bad press. For example, the Johns Hopkins’ IRB received scathing criticism for failures of oversight in an asthma study that led to the death of a healthy volunteer, Ellen Roche. The Office for Human Research Protections (OHRP) temporarily suspended all federally funded research involving human subjects at Hopkins.
Even back in 1996, alarms were raised by the GAO that the workloads of IRBs were too heavy and precluded thorough review. The load on IRBs is well described in Trials and Tribulations.
Funding for IRBs comes from the study sponsors—in this case, AstraZeneca likely paid the U. Minnesota IRB fee. And while the fees are thousands of dollars, the workload of most IRBs results in just minutes spent on each protocol. (Note, individual members of a committee are not paid for their reviews. The funding goes to the institution for administrative overhead).
There are also nonfinancial ethical conflicts of interest involving IRBs. These may be due either to excessive personal involvement or prejudgment by the experts on the board, or increasingly, from competition between the PI and IRB members.
Academic institutions have an incentive to approve studies, so that they can tout that it is providing “state-of-the-art” medical care and that it has been selected as a research site over its competitors by a leading pharmaceutical company.
But IRBs have certain responsibilities. As noted even on the U Minnesota website, “The IRB reviews research projects which involve human subjects to ensure that two broad standards are upheld: first, that subjects are not placed at undue risk; second, that they give uncoerced, informed consent to their participation. The IRB works with investigators to modify projects to ensure adequate protection for its subjects' welfare and right of self-determination. The University's process for protecting human research subjects reflects federal regulations developed in response to such cases as the Public Health Service syphilis study and the U.S. government radiation experiments.”
Yet Moira Keane, the director of the IRB (and now a Board member for PRIM&R), responsible for overseeing the CAFÉ study, was asked in her deposition, "So it's not the Institutional Review Board's purpose to protect clinical trial subjects, is that what you're saying?"
In the Markingson case, there are apparent conflicts of interest on multiple levels. The investigators had significant financial incentives to enroll patients—and to not alter therapy, even if a patient was not responding to therapy. Shockingly, the investigators claimed ignorance of basic ethical principles, including informing potential subjects of financial conflicts of interest.
They all claimed shocking levels of ignorance about basic research ethics, although holding positions of responsibility and authority. And, at each level, they absolved each other of any responsibility for this young man’s death.
With failures at multiple levels of supposed safeguards, and knowing that some of the same staff provided the assessments and care for patients on the AstraZeneca study as on a major NIH trial, the CATIE study, it seems essential that OHRP and an independent review examine not only the CAFÉ study, but take a careful look to see if similar breaches tainted the NIH trials.
Who will step up?
[I will continue this series in early January, looking a bit at the NIH and CATIE concerns and the University's troublesome response to this case. As Matt Lamkin just noted, "It can be difficult to keep up with the research scandals at Minnesota’s Psychiatry Department over the years," so we still have lots of questions to bring up.]
Carl Elliott, A referenced summary of the Dan Markingson case
Jeremy Olson and Paul Tosto, Critics say drug firms' payments to doctors are conflict of interest
Dan Markingson photo courtesy Mary Weiss
Origami dollar heart -Thomas Hawk/Flickr
One Flew Over the Cuckoo's Nest - Portland Center Stage
Molecules to Medicine banner © Michelle Banks