There are many disturbing things that happened during the clinical trial on which Dan Markingson committed suicide. Besides the issue of consent, or lack thereof, which I raised in my last post, one of the most disturbing aspects to me has been the lack of accountability and the apparent violations of clinical practice standards, with no consequences.
Zero accountability is how I have come to understand the outcome of this CAFÉ trial.
Clinical Research Standards—in Theory
Despite sensationalistic trials gone wrong, there are actually principles for “Good Clinical Practice,” and numerous guidelines for the ethical conduct of clinical trials. “Good Clinical Practice” (GCP) refers to International Guidelines on Harmonisation (ICH), E6 Guideline for Good Clinical Practice 1.24. This defines GCP as “A standard for the design, conduct, performance, monitoring, auditing, recording, analysis, and reporting of clinical trials that provides assurance that the data and reported results are credible and accurate and that the rights, integrity, and confidentiality of the trial subjects are protected.”
As part of this, there are extensive guidelines outlining the responsibilities of investigators and delegation of authority. The 2009 Guidance for Industry—Investigator Responsibilities, explicitly focuses on the supervisory responsibilities of investigators, stating, “When tasks are delegated by an investigator, the investigator is responsible for providing adequate supervision of those to whom tasks are delegated. The investigator is accountable for regulatory violations resulting from failure to adequately supervise the conduct of the clinical study.”
The FDA also notes that inappropriate delegation includes:
“Screening evaluations, including obtaining medical histories and assessment of inclusion/exclusion criteria, conducted by individuals with inadequate medical training (e.g., a medical assistant).”
Informed consent obtained by individuals who lack the medical training, knowledge of the clinical protocol, or familiarity of the investigational product needed to be able to discuss the risks and benefits of a clinical trial with prospective subjects.”
Form FDA 1572—Statement of Investigator
In addition to the FDA guidances, all investigators on investigational studies have to sign an agreement with the FDA, called the Form FDA 1572. The 1572 is a binding contract between the investigator and the FDA, whereby the investigator makes certain commitments. Investigators are required to sign a Form FDA 1572 for any phase 1 through phase 3 study; they are up to the discretion of the sponsor for Phase 4 (post-marketing) studies. Dr. Olson had signed the 1572 for this study, assuming specific responsibilities. As I explain in my text, the 1572 is essentially a marriage contract between the investigator (PI, or Principle Investigator) and co-investigators and the FDA whereby the investigator vows to fulfill the following obligations:
• Conduct the trial in accordance with the protocol. The PI can only change the protocol if a patient’s safety, rights, or welfare is at stake.
• Personally conduct or supervise the investigation. If responsibilities or assessments are delegated, the PI is still responsible.
• Inform patients that the drugs are being used for investigational purposes.
• Report adverse events (AEs) to the sponsor, within 24 hours if they are serious AEs.
• Read and understand the Investigator’s Brochure, the information already known about the drug.
• Ensure that other staff members who are assisting in the trial understand their obligations.
• Maintain accurate records.
• Report all unanticipated problems promptly to the IRB.
Responsibilities: Theory vs. the Reality of the CAFÉ Study at the U. of Minnesota
Let’s look at the reality of what happened to Dan Markingson on the Seroquel study, and see how it stacks up to basic expectations for the conduct of a clinical trial. Dan was enrolled on a study for schizophrenia, yet the diagnosis was initially “psychosis NOS” (not otherwise specified and to rule out bipolar affective disorder). The diagnosis of schizophrenia was not firmly documented until some time after he was enrolled.
Dr. Olson clearly delegated responsibilities to someone who was not at all adequately trained to do them. A social worker, Jeanne Kenney, did most of the clinical trial assessments. Her deposition and the recent, feeble “corrective action” issued by the Minnesota Board of Social Work as penance for her role, are quite damning. For example:
—although she fully acknowledged lack of any training in biological sciences, let alone medicine, she was responsible for evaluating Dan’s “review of symptoms” and noted, incorrectly and repeatedly, that he had hyperthyroidism, rather than hypothyroidism.
—she acknowledged that Dan had ~15 lb weight loss over a relatively short time, yet did not bring this to anyone’s attention because she didn’t understand that it might be significant.
—at one point, when questioned about inconsistencies in pill counts and Markingson’s erratic compliance, attorneys asked about there having been “106 percent compliance” with medications, she responded as follows:
Q. Did that concern you that this person is getting overmedicated?
A. Not for one day, no. Not for one day.
Q. Because why?
A. I just -- kind of based on our own experience of, you know, my own medications or something, if I took one, I don't think that would be anything to scare me, and I'd never heard that taking one too many pills would hurt anyone. So --
Q. But this was six too many pills.
A. Six too many pills was, really, they were all small doses, so it really only would have been like one pill possibly.
Q. Possibly. But you don't know that.
A. Well, I don't -- at the time, I didn't know which medication…
Q. Did you notify Dr. Olson about this breach in protocol?
A: I don't remember if I told him or not…
The Board concluded that her documentation consistently fell below the minimum standards expected for a social worker. As bioethicist Carl Elliott summarizes in “I was Just Following Orders,” “despite the fact that Kenney had no formal medical training, she apparently administered prescription drugs to research subjects and made formal judgments of the side-effects. She frequently assessed the severity of potentially dangerous adverse events such as akathisia [disturbing feeling of inner restlessness and inability to be still], and sometimes she even assigned the job to a social work intern she was supervising. According to the Board, her record-keeping was “devoid of any clearly articulated, consistent set of treatment goals” and she omitted crucial information relevant to suicide prevention. When Markingson’s mother, Mary Weiss, left “alarming voicemail messages” about her son’s condition, Kenney did not respond adequately. Kenney also made significant mistakes about medication dosage and a medical diagnosis, and these mistakes remained in the chart until well after Markingson’s suicide, when Kenney went back and changed them. To top it all off, Kenney often signed the chart with the initials of a physician.”
A message left by Dan’s mother, Mary Weiss, to the study coordinator, Jeanne Kenney, which warned, “Do we have to wait until he kills himself or someone else before anyone else does anything? ” Yet neither Kenney nor Dr. Olson thought this was worrisome enough to even notify the IRB, let alone to drop Dan from their clinical trial and place him on other medications, such as the Risperdal he had initially stabilized on, or Haldol, an older antipsychotic. In her deposition, Kenney acknowledges that Olson was a vulnerable subject, yet she did not feel it necessary to report his mother’s concerns to the IRB.
Given the severity of the violations of Good Clinical Practice standards and the resultant death of a clinical trial participant, what do you think would be an appropriate punishment?
Ms. Kenney received a carefully worded “corrective action” that emphasized that this was not a disciplinary action—and that only requires her to do 18 hours of continuing education and write a book report to the Board about what she had learned…and it took four years for the Board to reach this conclusion.
This response is disappointing, but not altogether surprising. Unfortunately, there is a long history of punishing low-ranking members of a unit. That they were “just following orders” has never been felt to be an adequate excuse. All too often, the leaders escape scott free.
It’s not hard to find summaries of investigator responsibility. For example, Dartmouth clinical trials website states, “The investigator is legally responsible for all aspects of the investigation and cannot delegate this responsibility to another individual.
It is common practice for investigators to delegate certain study-related tasks to employee, colleagues, or other third parties (individuals or organizations not under the direct supervision of the investigator). When tasks are delegated by the investigator, the investigator is responsible for providing adequate supervision or those to whom tasks are delegated and the investigator is accountable for regulatory violations resulting from failure to adequately supervise the conduct of the study.”
I have spent hours poring over depositions and original documents for the CAFE study, yet my review was, of necessity, incomplete, as there are thousands of pages. I did learn that Dan was told in his consent that his condition would be monitored closely by the study doctor (p 213 of Olson deposition), yet Dr. Olson only saw Dan monthly (p 164 ). Olson was unaware of some notes from neuropsychiatric testing that said that Dan was not on medication at that time (though he was supposed to have been). (p 230-232). He was similarly unaware of clinic notes from March, 2004, that suggested deterioration. (p 243-4). Yet on April 27, he recommended extending Dan’s stay of commitment for an additional 6 months.
It is still unclear to me whether he made dosage adjustments after seeing Dan or if he relied on Kenney’s assessments to modify dosages. What seems clear from the depositions, however, is that Olson deferred most of the visits to others—not exactly my definition of close follow-up by your doctor. (For more about PIs—Partially-Involved Investigators or, more disparagingly, Phantom Investigators, see Elizabeth Moench’s unsettling article).
Olson did state that, at the time of enrollment, Ms. Weiss expressed concerns as to whether medication would be switched if the medication didn’t work. (p 221), and he was aware of her desperate message to the coordinator, though he did not return her call. How does Dr. Olson’s actions fit with the responsibilities he assumed towards his patient and by signing the FDA Form 1572 obligations of investigators?
Physician as Caretaker vs. Investigator
There is another troublesome aspect—that of the conflict of interest of Dr. Olson’s being both the physician responsible for Dan’s care as well as being the Principle Investigator. (While we will discuss Conflicts of Interest in more detail in an upcoming post, I want to raise this here as a particular responsibility Dr. Olson seems to have shirked). Dr. Olson acknowledged, in his deposition, that he was Dan’s personal, and only psychiatrist, and that he did not transfer his care to someone else. He also, surprisingly, admitted not knowing the term “therapeutic misconception.” (p 72)
The U. of Minnesota training module says that “recruitment of participants need to be done in a non-biased, non-power based manner. It is important that none of the participants ever feel they do not participate in the study, they will be penalized.” (p 219)
Dr. Olson failed to act in his patient’s best interests. If a patient is doing poorly on a treatment regime—be it research or not—it is customary and obligatory to look at other treatment options and to change therapies. While some of the staff say they did not detect the deterioration that caused Ms. Weiss to sound alarms, it is clear that Dan was not showing an adequate response to his treatment. Despite 5 months of treatment, Dr. Olson recommended that Dan’s stay of commitment be extended for another 6 months, conveniently allowing Dan to complete the study.
The conflict between being the primary caretaker and a researcher is something many of us struggle with regularly. That is why it is recommended that the research physician not be the patient’s primary physician. I also encountered this conflict with many of the trials I conducted, which often enrolled patients with severe and life-threatening infections. My threshold was always, “What would I want if this were my family member?” and “Can I sleep tonight without regrets?”
The Minnesota Social Work Board’s puny response towards Kenney is not what I find most troublesome. The thing that I am most appalled by in reviewing this tragic case is that nothing—absolutely nothing—has happened to Drs. Olson or Schulz, the IRB responsible for overseeing the trial, or the University of Minnesota.
Photo of Dan Markingson and Mary Weiss courtesy of Mary Weiss
Pills by e-magic/flickr
Molecules to Medicine banner © Michelle Banks
For further information on research ethics, basic courses are available through ACRP (Association of Clinical Research Professionals) and PRIM&R (Public Responsibility in Medicine and Research).