Working draft:

The best laid plans go awry


In a study at Johns Hopkins University, researchers induced asthmatic symptoms in healthy lungs, to help doctors learn how the body fights asthma. Hexamethonium was used to induce the symptoms to investigate a specific function of the lungs. A 24-year old woman (Ellen Roche) died June 2, 2001.

She was the third volunteer in the study. The first volunteer reported coughing a day after participating in the study, but the cough was resolved within a few days. The second volunteer suffered no ill effects. Before being accepted, Roche signed a consent form and underwent a battery of tests. The tests included lung function tests, and she was deemed healthy and without conditions that would have excluded her from the study. (Roche in fact worked as a lab technician in the Johns Hopkins Asthma and Allergy Center, and had participated in two previous studies. Wouldn't she have unusually good background knowledge? And didn't she know about the reaction of the first subject? Or is this one of those "Trust anything the white coat says" cases?)

The experimenter, Alkis Togias, didn't initially ask the FDA for permission to have volunteers inhale a drug (hexamethonium) that would trigger asthma-like reactions in their lungs, and then changed how volunteers received the drug during the study without notifying anyone. (The change apparently involved dissolving the hexamethonium in buffered distilled water, instead of saline, as originally proposed in the study protocol. The intent was to make the ingredient safer.)

Johns Hopkins officials said the review board initially had not believed they needed approval for inhaled hexamethonium because it was already being used for basic physiology tests. Thus there was a failure to inform participants that the drug they would inhale was technically "experimental."

Togias had more than 10 research studies in progress when Roche died, and the university put all on hold pending review. It also suspended 16 other studies involving substances for which FDA opinions were not sought. Within six months after the shutdown, the university had reapproved almost all the 2,600 halted experiments except those led by Togias. His research has apparently resumed as of April, 2003.


Very sad, especially in hindsight (again). The "errors" seem so minor, and the subject was an insider and thus in a better position than most research participants to know what the risks might be.

It is worth noting that none of the recent add-ons to ethics reviews (e.g., lay people on the committee) would have had the slightest impact on this. Yet amazingly the IRB was later faulted because: "IRB membership lacked cultural diversity." Well, THAT would have prevented this incident for sure! Even the experts on the IRB seem to have been unaware of the potential toxicity of the treatment, so adding "incompetents" to an already heavy workload for the reviewers would only exacerbate their problems. Whatever these regulators are thinking about it has long since ceased to be about public safety. Using this unfortunate death to instigate a moral panic for further irrelevant controls is reprehensible.

In a subsequent report the JHU ethics review committee was faulted because, among other things, "investigators were encouraged to limit the length of informed consents, and as a result, appropriate attention was not paid to the required elements of informed consent (i.e., the avoidance of complex language; consequences & procedures for withdrawal; new findings; termination of participation, etc.)." I believe that's called damned if you do and damned if you don't, sucking and blowing at the same time?

There seem to have been long delays in feedback from the FDA to letters by JHU before the project, but the FDA is even now (over a year later) righteously pursuing Togias. I think the technical term for this is Scapegoating. Or is it Butt-covering? Can it be as bad as this report implies?

I am reminded that in a bureaucracy, which the research ethics industry has become, success is not defined as solving a problem, it is defined as not getting blamed for a problem.

In reading the FDA letter to Togias, I am struck by the byzantine nature of the regulations in effect here. Perhaps if one actually works with these daily they become manageable, but I doubt it. I am of the opinion that "accidents" like this will become more common because of the complex rules, instead of less common. The lawyers will certainly feast on the complexity.

The only reliable effective process for dealing with such complex rules is hindsight -- everyone is an expert in hindsight. However, the only reason to celebrate our cleverness in hindsight is if we do identify how to fix things so the problem doesn't repeat; taking that opportunity to embed nonsense should disqualify you.

Very sad, but using this accidental death to justify further extraneous complications (e.g., lay people and cultural diversity) in the review process is not justified for medical research, much less social sciences.