Wednesday, February 13, 2013

Has the Department of Psychiatry been faking research subject records? New evidence comes to light.

Two weeks ago, I posted about a curiosity in the medical file of Dan Markingson.  University officials have repeatedly argued that Markingson was fully competent to give his informed consent to the study, pointing to an “evaluation to consent” form located in his chart.  According to this form, which was signed and initialed by Jean Kenney, Markingson had given rational, informed answers to all of the questions he had been asked before he gave informed consent.  The curiosity was that Markingson’s file actually contained two slightly different versions of the same form (here and here.)  I asked readers to help me figure out what could account for the differences.

This request produced a startling response. Within hours, several people had posted on Facebook and written by email to say that the charts of their family members contained exactly the same “evaluation to consent” form.  Today, one family has agreed to let me post the form that was given to them by the Department of Psychiatry. 

In most ways, this form looks identical to one of the forms that allegedly came from Dan Markingson.  The signature of the study coordinator, Jean Kenney appears identical; the signature of the witness, Elizabeth Lemke, appears identical; and the rubbed-out marks located at the top of two of the documents appear identical.  Most alarming of all, the “answers” supposedly given by the research subjects appear identical.  But the new form has no date, no subject ID, no subject initials, and no mention of Unit 12, the ward in Fairview Hospital where Dan Markingson was hospitalized.

At the risk of provoking a legal action against me by my employer, let me speculate about what this could mean.  Judging from this admittedly circumstantial evidence, it appears as if someone in the Department of Psychiatry may have been using a generic, photocopied form with pre-determined answers as documentation that mentally ill research subjects were competent to consent to research studies.  Such forms would have been particularly important for the CAFÉ study, because they would have allowed the enrollment of acutely psychotic subjects without getting proxy consent from their family members.  For example, Dan Markingson was enrolled in the CAFÉ study over the objections of his mother, Mary Weiss.

It is worth pointing out that these “evaluation to consent” forms have been submitted to official research oversight bodies, which have relied on them to make judgments about the conduct of the studies.  For example, in a March 2004 letter to the University of Minnesota IRB, Dr. Stephen Olson wrote that Dan Markingson had been given an “evaluation to consent” and had answered all the questions satisfactorily.  Also, when the FDA investigated Markingson’s suicide in 2005, it cited the “evaluation to consent” form as evidence of Markingson’s competence (see page 9.)  Submission of false evidence to the FDA can lead to disqualification as a clinical investigator.

I want to be cautious here.  I have only seen three of these “evaluation to consent” forms: this one, and the two that appeared in the file of Dan Markingson.  And although I have no reason to doubt that the families of other research subjects have found the same forms, I have not seen those forms for myself.  That said, I find these documents so alarming that they demand an investigation. How many research subjects in the Department of Psychiatry have this same form in their charts?  Were any of those subjects in studies subject to federal oversight, such as the CATIE study? Did any of those subjects suffer serious injuries or death?


  1. Leaving the Markingson case aside for a moment: the presence of an Evaluation form, with no sign of ID or date, in the file of a trial subject would itself be grounds for a complaint to the IRB and subsequent investigation. If your correspondent in fact has a form with this level of incompleteness, he or she should provide an original to the IRB as soon as possible.

    Submitting the original would also dispel any lingering concerns over the authenticity of the new form (somewhat natural, I think, when shared anonymously), and facilitate comparison with other evaluation forms already in the IRB's possession.

  2. Of course, that is assuming that the IRB is not complicit.

  3. Great detective work. You are like a dog with a bone. Keep it up!

  4. This comment has been removed by a blog administrator.

  5. Thanks, I appreciate that. (For some reason your comment was posted twice -- possibly my error. I deleted the second one.)

  6. It would appear that a generic version of the Evaluation to Sign Consent Form was created which would then only require insertion of the individual subject's details and dates for the evaluator and witness pre-populated signatures. Further, it would appear that the pre-populated form was used in the Markingson case (identical redaction blob). Not sure how the slightly different blob could come about - OK to be deeper, but I think it's also shorter! - but the dates to the signatures are identical including the over-emphasis/correction to the '1' of 21 without this being initialled and dated. Whatever the implications for the Markingson case or for other subjects in the CAFE or any other study, procedures at this site were not in compliance with international standards of Good Clinical Practice.