Interesting comment about medical research vs. physics research

[bobaustin]

I caught this in the NY Times:

http://community.nytimes.com/comments/www.nytimes.com/2009/06/28/health/research/28cancer.html?sort=oldest&offset=4 ​

The article focuses on cancer research, but the conclusions can be applied to all medical research.

As the article correctly pointed out, medical research is done in very small and "safe" steps. Big visions and bold models, irrespective of how well they might explain reality and how far and how fast they could bring us, are routinely ridiculed and shot down.

To illustrate, let me give a very simple example. Let's say that physicists and medical researchers are going to research a flight of stairs. The medical researchers start first. They apply for a grant and get funded to measure the first step. They measure the first step, then apply for a grant to measure the second. And so on until they reach the last step. In the end they have a conference during which they discuss their findings and try to guess the total length of the stairs. They come up with a number, but that is just an assumption, so they apply for another grant to measure the whole thing. In the end they get to the truth after years of research and much time and money wasted. The physicists do things a little differently. They eyeball the stairs and estimate the height of one stair and then count the stairs. They multiply the number of stairs by the estimated height of one and get to the approximate height. Now it is time to check if this is correct, so they measure the height of the stairs. If the model and theory didn't match the result, they come up with another explanation for the height they measured, and then they come up with another way to test the theory. They get the result much faster and many times by deduction.

What shocked me again and again while sitting in audiences on medical conferences is the degree of destructive skepticism that dominates all discussions. Many people in medical research build whole careers on just commenting other people's work in the following manner: "interesting, but there is insufficient evidence to support this".

If physicists were medical researchers, they would still argue if in fact the Earth rotates around the sun, and some of them would demand that this be verified by someone leaving the solar system to observe for a year if in fact our model of the solar system is the correct one.

Fortunately, physicists are not medical researchers, so we now drive cars and fly planes, use iPods and the internet. In medicine we are still at the level of the steam engine, which is admittedly much better than walking or a horse-drawn carriage, but still far from where we could be.
Recommend Recommended by 9 Readers​

The comment was in response to a NYT article on cancer research and how after $105 billion in taxpayer money, we have little to show for it because the grant system is corrupt and doesn't encourage new ideas (http://www.nytimes.com/2009/06/28/health/research/28cancer.html)

 

[Mapes]

What a specious analogy to compare the uncertainty involved in typical medical research with the uncertainty involved in the Copernican system. Two-body orbital rotation is completely solved; even 3- and >3-body interaction is amenable to simulation. A better analogy would be to the field of global warming, where there is in fact much debate and pragmatism over purported solutions prior to investing mammoth amounts of money (compare this to the debate and pragmatism exercised before risking human life on underdemonstrated treatments). If you look at it this way, medical researchers and physicists working on sufficiently complex systems don't look that different -- and why should they?

 

[Andy Resnick]

I agree with the article, and would go further to claim that nearly all grant-based research is of the incremental type.

Part of the problem is the nature of peer-review: grant reviewers do not, in general, review projects they are experts in, and the tendency is to rank projects high if the reviewer can clearly understand the (proposed) benefit. These are projects that are cleanly embedded in the existing body of knowledge (for example, "this has been shown on system 'a', we propose to study the same thing in system 'b'"), not projects that have a disruptive effects.

Most agencies have some specific policy about this- either they have some special program for "especially innovative"/ "high risk" proposals, or the culture is such that unusual ideas are encouraged (NASA used to be like this). But getting these awards is significantly more difficult than getting 'normal' awards.

Compound the need to have a project easily explainable with tight funding overall, and the result is less and less innovation and more and more "by the numbers" data generation.

 

[Dale]

I agree with Mapes, and also point out that medical research is bound by many times the legal, ethical, regulatory, and government oversight requirements of any other research domain. A physicist who incorrectly measures the mass of a particle gets corrected by the next physicist with very little long term damage. A medical researcher who incorrectly measures the risk of a new medical device kills people, gets sued, and possibly goes to jail. Not only are the systems themselves more complex, but the research environment itself is more complex.

The incremental nature of medical research is not necessarily a bad thing, and I hardly see the grant system as a big problem. Also, the statement that we have little to show for the research money spent is absurd. Life expectancy is way up, and mortality from targeted diseases is way down. The incremental approach gives results and keeps from killing too many people. New ideas aren't always good ideas.