The Ethics of Random Clinical Trials

by on December 15, 2010 at 7:20 am in Economics, Education, Political Science | Permalink

In New York City a random clinical trial over a housing program has many people upset (as Tyler noted earlier):

…some public officials and legal aid groups have denounced the study as unethical and cruel, and have called on the city to stop the study and to grant help to all the test subjects who had been denied assistance.

“They should immediately stop this experiment,” said the Manhattan borough president, Scott M. Stringer. “The city shouldn’t be making guinea pigs out of its most vulnerable.”

The controversy brought to my mind this story from Dr. E. E. Peacock:

One day when I was a junior medical student, a very important Boston surgeon visited the school and delivered a great treatise on a large number of patients who had undergone successful operations for vascular reconstruction.

At the end of the lecture, a young student at the back of the room timidly asked, “Do you have any controls?” Well, the great surgeon drew himself up to his full height, hit the desk, and said, “Do you mean did I not operate on half the patients?” The hall grew very quiet then. The voice at the back of the room very hesitantly replied, “Yes, that’s what I had in mind.” Then the visitor’s fist really came down as he thundered, “Of course not. That would have doomed half of them to their death.”

God, it was quiet then, and one could scarcely hear the small voice ask, “Which half?”

Dr. E. E. Peacock, Jr., quoted in Medical World News (September 1, 1972), p. 45, as quoted in Tufte's 1974 book Data Analysis for Politics and Policy.

Hat tip for the quote source to Raw Meat.

Dan December 15, 2010 at 3:29 am

isn't there a huge temptation for social workers and govt. officials to help/hinder trial members, in the hope of promoting whichever option they favour?

This isn't an area where a double-blind trial is possible (prove me wrong?), and I'd imagine outcomes involving homelessness being partly dependent on how much those providing the front-line services go beyond their official duties to help people.

Patrick Dennis, MD December 15, 2010 at 4:34 am

What a pathetic travesty. In addition to the two words cited by Dan – double blind – here are two more: informed consent.

Robert Speirs December 15, 2010 at 4:35 am

To be clear, then, it is impossible to use the scientific method in the provision of "social services". What standard is used instead? Who whines loudest? Bribery? Naked force?

Peter December 15, 2010 at 5:46 am

This criticism is understandable, but I think a bit misguided. In many experiments, there is a simple rationale for forming a control group. Often, a government or NGO does not have unlimited funding–they cannot afford to provide a program for everyone–therefore to a fair way to allocate is through randomization (this could even be through targeted randomization–prioritizing those with the greatest need). In addition, as they do in medicine for potentially life-saving treatments, it is possible to statistically account for stopping rules that allow for the experiment to end early if it is successful so that it can be provided to people in any desired way, assuming there is enough funding (this is not mentioned in the pithy quote above–does the doctor have time to perform this procedure on every person afflicted in the world? And if the procedure is successful, it can be quickly determined as such).

New York City does not have an infinite supply of money, and they want to find out how to help people the most. You have 23 millions dollars you're spending on a potentially lousy housing program; what if you could do a lot better for the homeless if you spent that money in a different way?

Flossmoor Joe December 15, 2010 at 6:41 am
Jeff Ransom December 15, 2010 at 6:54 am

I love spaghetti

Joel W December 15, 2010 at 7:01 am

Simple pareto optimality should lead to the moral conclusion here.

Six Ounces December 15, 2010 at 8:36 am

I don't know much about medical research, but I can tell you there's something awry. Whenever they list possible side effects and say, "Only 2% of people taking the drug report these effects," I am in the 2% category 100% of the time. I think that many of the respondents don't respond or don't report. If they were asked explicitly, do you have this side effect, they'd say "Yeah, as a matter of fact I do. It's not too bad and I thought it was unrelated to the drug." But the drug companies don't want them to report side effects, so I don't think they dig very hard to find the truth.

As for the efficacy and the ethics, the solution is fairly obvious. All those who can pay and want the drug, get it. Grants can pay for some of the drugs for those who can't or won't pay. Anyone who can't pay enters the drug/placebo lottery.

It's better to save half than save none. A 50% chance of living is better than zero chance of living.

The drug might be worse than no drug, so it might be a contest you want to lose.

No one has a right to free use of resources.

We ALL die, so if your only argument for free care is prolonging your life, that's pretty weak.

What better use of your own earnings than to prolong your own life? If you're illness is terminal and you can pay for an experiment, you should get it. The only job of the FDA is to make sure there is at least some theoretical basis for efficacy, i.e. Not total snake oil.

It would be unethical but scientifically confounding to prohibit those in the trial from seeking other cures. But that could add to the information set too. Treatments in combination are sometimes better.

Kieran December 15, 2010 at 1:37 pm

There are instances, and I'm not saying this is one of them, where the act of obtaining informed consent would bias an experiment. The ethical solution to this is to obtain informed consent after the experiment has been done. If a participant does not give their informed consent then you remove them from the trial and delete all data collected on them.

Ricardo December 16, 2010 at 1:45 am

The reason double-blind randomization exists in medicine is that medical researchers are studying chemical and biological processes: you need to control the standard of care given by the physician in order to test the effectiveness of a drug.

In social policy, that's not really the case. Suppose that social workers give an extra bit of help to the treatment group: so what? Why wouldn't they do the same thing in the real world even in a non-experimental context? If they give extra help to the control group relative to the treatment group, that's really only a problem if someone thinks that social workers can work magic in curing problems like poverty and homelessness. History suggests that isn't and if there is an effective tool against these problems, it won't be something that social workers can easily implement on their own to fudge the results.

RCTs aren't perfect but there aren't very many other good options. Without RCTs, we are stuck with the op-ed approach to public policy where people impose their own narratives on the very limited facts that are available.

Andrew December 21, 2010 at 9:55 pm

I've always thought trails should be done on people who don't know they are in a trail – just as long as I'm not one of those people.

Comments on this entry are closed.

Previous post:

Next post: