by bridgetohealth » Fri Feb 27, 2015 10:34 am
Just shoot me now for getting involved in this discussion.
If you compare women, who, after a mammography and breast cancer diagnosis choose treatment or don't choose treatment, you are comparing in large part the types of women who choose and don't choose treatment -- it's not necessarily the treatment itself you are comparing. Poor women, for example, are more likely to forego treatment, but also tend to have worse health to begin with. This is the reason to do randomized tests.
Mammography has been tested more than almost any other screening or treatment. They randomly have women get a yearly mammogram or no yearly mammogram. The difference between these two groups of women is ONLY whether they get yearly mammograms. This is the nature of randomized trials, and anyone who doesn't understand that should read up on it in regards to issues that aren't so emotionally charged. Within the mammography group, there are breast cancer diagnoses, and within those there are women who choose and don't choose treatment. But that doesn't matter, because the trials look at whether group A, yearly mammogram receivers, is different from group B, no yearly mammograms. And -- the difference shows that IF there is a significant difference in breast cancer mortality, it is extremely negligible, and is counterbalanced by all-cause mortality -- i.e., the same number of very few women who may be saved from dying from breast cancer die in group A die from something else instead (i.e. unnecessary treatment) in group B. In addition, in group A there is the emotional toll of false positives in the mammography group, plus unnecessary lumpectomies etc.
If you want to look at treatment, which is a totally different question, then you would RANDOMLY place women in the treatment or no treatment groups. Therefore, you would not be comparing women who are the type of person to choose or forego treatment, but instead you are comparing the actual treatment itself. This is obviously highly unethical -- telling people to forego cancer treatment -- and isn't done. So what they do instead is compare different kinds of treatment, or try to make assumptions from the type of data some are suggesting here, but those would only be assumptions, and have absolutely zero to do with the greater question of whether mammography saves lives and has benefits that outweigh the negatives.
An unrelated example: one reason researchers have falsely believed that supplements have a health benefit is because the type of person to choose to take a supplement is different from those who don't. They are richer, more concerned about their health, often eat less junk, etc., etc. If you just say who here takes a supplement? And compare the two groups, you are not just comparing supplements vs. no supplements -- you are comparing supplement choosers vs. non-choosers. And it turns out (with actual randomized trials) that for most supplements, the supplements don't help and often hurt. But that difference is overcome (if you just have people raise their hands) by the fact that supplement choosers are healthier people for reasons having nothing to do with supplements. The suggestion to compare breast cancer treatment choosers vs non-choosers is the same idea.
Again, randomized tests of mammogram vs. no mammogram, are telling us about mammograms, and the treatment has nothing to do with it -- the fact that they are randomized means there are choosers and non-choosers in both groups in the same proportions as they exist in the general population.
