Wednesday, November 18, 2009

Breast Cancer and the problem with statistics/ utilitarianism...

As you might imagine, I've been doing a bit of reading about the breast cancer study...

I've come to see part of what troubles me about this. They didn't do any NEW investigations, rather they compiled existing data and made the declaration that it's better to start mammograms at 50 and not to bother with self-exams.

For now, I'm going to ignore the possible implications concerning insurance / government paid health care.

One way to look at this is as a problem of interpretation of statistics. What they saw was an increase in the number of biopsies but not a corresponding decrease in the number of deaths from cancer. There IS a decrease, but it doesn't match the increase in the number of biopsies. They also saw an increase in the number (perhaps the percentage) of false-positive mammograms and they saw research concluding that some cancers come and go naturally without becoming invasive/deadly. They concluded that the risk of a false positive/ unnecessary treatment didn't outweigh the decrease in deaths.

They couched this in statistics -- but, the value judgment underlying this farce is a bad interpretation of utilitarianism.

Background - Utilitarianism -- which (gasp) wikipedia has kind of wrong -- is the idea that the morally good thing is the thing which tends to produce an increase in happiness, while the morally wrong thing tends to produce a decrease in happiness.

Wikipeida is wrong -- or, probably more like confused -- when they define utilitarianism as "the greatest good for the greatest number of people." And -- that's what this study does as well. The correct formulation is 'the greatest overall good, all persons considered'.

The problem with the study is that it tries to compare the relatively small harms from false positive mammograms(thus unnecessary biopsies) and unnecessary treatments --which happen for a relatively large number of women -- to the HUGE pain of a smaller number of deaths from breast cancer, which otherwise would have been avoided with necessary treatment.

The study wrongly concludes that the concerns of the larger number outweigh the deaths of the smaller number.

I think they are flat out wrong. In order for this to come out right, in terms of utilitarianism, the number of false positive mamograms/ unnecessary treatment would have to be staggering -- as, the pain of actually dying of any kind of cancer -- breast cancer is no exception -- is huge for the patient... and their families.

In essence, what they're saying is that it's better to avoid the stress of having to do more tests and possibly a mastectomy/chemo/radiation, than it is to avoid the deaths from cancer that would come from screening women 40-50. As someone who has done mastectomy/chemo, I call BS. I think I can speak on behalf of my chemo buddies when I say that we'd rather have had all of our treatment be unnecessary than to have just one more person not figure out they have cancer until it metastasizes.

Even a good formulation of utilitarianism can reach this conclusion --perhaps that's why I'm not a utilitarian.


Anonymous said...

Okay, you totally wrote in a logical and eloquent way what was niggling at me after I heard this reported on NPR this morning. I have a bad feeling this "finding" will lead to nothing good or helpful but to excuses to deny MORE health services to women.

Seeking Solace said...

Just another reason why Wikipedia is a lousy source. I caught that too when I would working on my Ethics notes for my students!