Not better than the alternative: an informal experimentation tragedy

Police are reinvestigating the 2007 death of Yolanda Cox, a woman who collapsed in anaphylactic shock after being injected with an experimental drug by her sister, a GP. The drug was developed by their mother, originally intended to treat diabetes but apparently believed to extend lifespan. After testing on diabetic patients and apparently themselves without any apparent side effects the mother and sister gave it to the woman three times, with tragic consequences the third time. Was it a failure of medical ethics, research ethics – or plain psychological bias?

The whole case is on the borderline between medical research and medical treatment. Normally, drug trials are surrounded by elaborate rules to protect test subjects and to prevent various interests from biasing the results. However, in this case the family denies it was a real drug trial. The drug was given to treat an ovarian condition they had diagnosed Yolanda with. That they thought it was safe and beneficial seems likely. Whether they had any scientific reason to think it would be helpful seems harder to tell from the statements in press. The patent filing makes some references that might support this, but it is not a scientific argument. Similarly it is unclear to me whether there were good scientific reasons to think it to be safe, although an earlier drug trial with 31 subjects apparently went without a hitch. However, according to Camden News the dosage was either not measured or three times the trial dosage (the article is contradictory). That suggests a problem with the science: that the drug works for one condition doesn' imply that it works for another, and that a certain dosage works doesn't imply that a significantly higher dosage works better. That the drug seems to be good for almost anything is another warning sign.

A key concept for participating in research is informed consent: do the participants understand the facts of the experiment and its potential consequences well enough? Yolanda appears to have known much about the drug, so it is likely that she had good information. However, none of the news stories suggests she gave a written statement of her consent.

But even if she did, there would be reasons to think that she might have been influenced by her family – she "grew up" with the drug which her mother developed, her sister and mother diagnosed her condition and claimed the drug would likely help, they appear to have been devoutly religious and prayed before injections – all this could contribute to a situation where her choice was not totally free. The normal groupthink occuring in a group with high investment in a particular outcome and desire for cohesion can bias their decisions to ignore contradictory evidence. It is not inconceivable that this is exactly what happened here.

Was it OK for the sister to inject the drug? Many doctors do treat family members, but it imposes heavy burdens: it
might be hard to disclose bad news, ask about intimate medical history
or to be objective to judge what amount and what kind of care they
need. In this case it seems that the biases may have led to bad
decisions. While doctors should use science and experience to guide
their actions,
sometimes experimental or nonstandard treatments are necessary.
However, in this case it seems likely that the drug was not the best
known treatment for the condition – it might potentially be better, but
that can not be determined by a single trial or from what was known at the time. Hence the sister may have not been giving the best care.

What about the mother's role? Once upon a time self experimentation was common in medicine. The reasons have been manifold, from genuine altruism to a desire for recognition to nationalism, and it has often been criticized for being more adventuring than science. Self-experimentation has advantages in terms of reliability, access to observations, first-hand experience, reduced red-tape and especially informed consent. The research is firmly grounded in the researchers autonomy. But there are always concerns about excessive risks, pathological motivations, lack of objectivity and the risk of extrapolating from the findings in a single individual. I do not think these arguments against self-experimentation are conclusive in general, and that in many cases (in particular experiments with extending human capabilities) self-experimentation can still contribute much to science and human exploration.

But while self-experimentation might be acceptable for the researcher, it is less clear that experiments on others can be done informally. One of the critics of self-experimentation, John K. Davis argues that the teamwork nature of science means that self-experimentation could pressure colleauges into dangerous activities. While I am sceptical of the general applicability of this argument, the current case might be a prime example of where this may actually have occured.

Group-think is treacherous, since it prevents the criticism that is often needed to even become aware of it. The family should have recognized the conflict of interest and risks of reduced autonomy, but didn't. The unnecessary risks they then took looked sensible to them. This is why larger and less closely connected groups are useful for research and medicine: they have a chance to bring up uncomfortable questions. The presence of formal structures, rules or an ethics review board are less important than the ability to ask why people think a course of action is the best, objective thinking and a clear option of saying "no".

  • Facebook
  • Twitter
  • Reddit

2 Responses to Not better than the alternative: an informal experimentation tragedy

  • Dennis Tuchler says:

    What did the patient (sister) know when she was injected? Was the (psychologically) competent to consent? Was she aware of the risks? Did the treatment preclude the use of other treatments that were available, affordable, and likely to be helpful?

  • Good questions, Dennis. I would suspect that the treatment did not preclude other treatments (insofar the condition even needed a treatment).

    As I see it, the consent issue is the intriguing thing in this case. In terms of information, she was probably well supplied. But the *interpetation* of this information (in terms of risk/benefit) could very well have been skewed by the social situation. Competence would likely have been diminished somewhat by this, but it is very hard to guess to what extent, including whether it was serious enough to really impair the competence to consent.

Authors

Subscribe Via Email

Affiliations