A suicide draws attention to the ethics of psychiatric drug testing

This is a really important long read that we all need to pay attention to. It concerns how we treat people with who are suffering from paranoid delusions — and how we treat people whose families worry that they are a threat to others. It concerns the relationships between doctors and the pharmaceutical industry. It concerns the ethics of clinical trials — the risks we run as we test potential treatments that could help many, or hurt a few, or both. If we want to reform mental health care, this needs to be part of the discussion.

In 2004, Dan Markingson committed suicide. The story behind that death is complicated and depressing. At the Molecules to Medicine blog, Judy Stone documents the whole thing in three must-read chapters. Many people find help in psychiatric drugs, and credit those drugs with making their lives better. (Full disclosure, I'm one of them. I have used Ritalin for several years. I am temporarily on an anti-depressant.) But we have to pay attention to how those drugs get to us. This isn't just about treating people. It's about the process that gets us there. Because, if that process is compromised, the treatments we get won't be as effective and lives will be lost along the way.

Markingson began to show signs of paranoia and delusions in 2003, believing that he needed to murder his mother. He was committed to Fairview Hospital involuntarily after being evaluated by Dr. Stephen Olson, of the University of Minnesota. He was subsequently enrolled on a clinical trial of antipsychotic drugs—despite protests from his mother. This study was a comparison of atypical antipsychotics for the treatment of first episodes of schizophrenia (aka the CAFÉ study), sponsored by AstraZeneca. The study’s structure was that of a Phase 4 randomized, double-blind trial comparing the effectiveness of three different atypical antipsychotic drugs: Zyprexa (olanzapine), Risperdal (risperidone) and Seroquel (quetiapine), with each patient to be treated for a year.

After about two weeks on study treatment in the hospital, Markingson was discharged to a halfway house—again over his mother’s objections. Over the coming months, Dan’s mother, Mary Weiss, continued to express concerns about her son’s deterioration, even asking if her son might have to kill himself before anyone else would take notice…then, in fact, her son violently committed suicide on May 7, 2004, mutilating himself with a box cutter. The University of Minnesota and their IRB have maintained that the study was conducted appropriately and that they have no responsibility for Dan’s death. Dan’s mother and bioethicist Carl Elliott believe otherwise.

We’ll explore some of the major issues of contention in this case over several posts, as illustrative of basic clinical research principles, including adequacy of informed consent, IRB oversight, conflicts of interest, and coercion, including threats to a bioethicist whistleblower.

Read the first part of the story

Read the second part: How clinical trials should be done and how they were done in this case.

Read the third part: Conflicts of interest between the researchers and the pharmaceutical industry.

Image: Pills (white rabbit), a Creative Commons Attribution (2.0) image from erix's photostream



  1. When my father was diagnosed with cancer, some of us urged him to try experimental treatments being studied at NIH.  He joked, “I wouldn’t want to wind up in the control group,” but I think he had a point.

    The ethics around studies like this seem to be getting better, but there’s still a risk factor.

    1. Apart from the obvious (transparency, consent etc.) how can it get “better”?

      Somebody has to try the new drug, and somebody has to end up in the control group receiving the placebo. One group is guaranteed to die, and the other carries the risk of test driving if it actually works and any side effects that nobody knows about.

      It’s basically a suicide commando, biochemical kamikaze, somebody has to do it, but ethically? Whom but the very desperate will? And what about bias? People afflicted in a way that compels them to partake in such a study are likely not a very representative test group for the general population.

      Maybe it’s time to start thinking about brain-dead human clones, gotta be better than mice, cellcultures and pigs.

      1.  If a treatment exists, even if it’s not fully effective, it would be unethical to design a study that tested an experimental drug against a placebo. The usual way of testing new drugs compares them to existing treatments, not to no treatment/placebo treatment.

  2. There was no consent. How did they get away with this? The patient couldn’t consent because he was involuntarily institutionalized. How does the doctor get to consent to unproven treatment without the mother’s approval? 

    1. It sounds like the hospital reserves the right to rent out its residents for use as guinea pigs.  What could possibly go wrong?

  3. For more discussion of the Dan Markingson case and other ethical controversies with the University of Minnesota Department of Psychiatry, visit the Facebook page Community Alliance for Ethics in Minnesota Psychiatry.


  4. I’m always a little wierded out when I see a medication that has “may cause suicide” or the like on it’s list of potential side effects.

    I have to wonder what the precise legal limit is in terms of the percentage of users killing themselves, for example.

    1. In some (but not all cases), suicide is proposed to be an effect of the efficacy of the drug. For example, doctors are warned to closely follow a patient starting an antidepressant, because if the medicine begins to work the patient may have the energy to carry out a suicide plan he or she had already formulated, but previously been literally too depressed to act on.

  5. Marsha Linehan gave a lecture a couple of years ago on the ethics and difficulties with performing research on people at risk for suicide. She also explained how the studies can be misleading due to the exclusion of high risk individuals. Most psychiatric drugs are tested on people who are in a lower risk category. The big problem is that since most ethical guidelines don’t allow research on high risk individuals, doctors can only assume that a psychiatric drug will help a high risk person. And without a clinical trial that will likely endanger the health of the high risk individuals, it will be extremely difficult to develop drugs or other types of treatment that will work for them.

Comments are closed.