Reconciling the findings
So what was going on in Canada? From the perspective of readers of the scientific literature at least, these were the most carefully conducted randomized trials of mammography.
And the reporting was meticulous: the randomization process, the comparability of the characteristics of women in the control and intervention group, the diagnostic procedures performed, how different cancers were found, the number of lymph nodes involved, the process used to determine cause of death, and the causes of death themselves were all described in detail. In fact, the level of detail reported undoubtedly increased the opportunity for criticism (there being more methodologic details to raise questions about).
In the context of what was known about breast cancer screening, were the findings really so heretical? Not especially. Let's consider all the randomized trials I've discussed, starting with the best-studied age group: women age 50 to 70. Think about what is being compared to what.
The HIP study compared clinical breast exam plus mammography to no screening and found that doing something was better than doing nothing. The four Swedish trials compared mammography to no screening and found that mammography was better than doing nothing. The Canadians compared clinical breast exam plus mammography to clinical breast exam and found that mammography didn't add anything. There's nothing inconsistent in these findings. All are perfectly plausible.
Now how about Canada 1 and women in their 40s? There it was mammography versus no screening and the finding was that mammography was no better than doing nothing. Are there reasons mammography might not work in younger women? Probably. Younger women tend to have more aggressive cancers. Because they grow fast, they often become palpable during the interval between mammograms, rendering the mammogram useless. Furthermore, everyone agrees that it is harder to interpret mammograms in women with denser breasts, a condition more common in younger women.
But something else is different about younger women: they are less likely to get breast cancer plus they are less likely to die of it. That's good for the women, but a challenge for researchers trying to figure out whether screening works for younger women. Of the 50,000 women age 40 to 49 followed for seven years in Canada 1, for example, only 64 died of breast cancer. Not to sound crass, but that's not a lot of deaths for researchers to work with. Maybe mammography works, but not very well. In any event, the effect, if it exists, is small enough that we just can't reliably detect it.
How good were the studies?
In the fall of 2001 a systematic review of all of the mammography trials to date was completed by a group known as the Cochrane Collaboration, an international nonprofit organization that regularly reviews the evidence of health care interventions sort of Consumer Reports for medicine.
The article appeared in the Lancet and, soon after, on the front page of the New York Times. Based on the rules of evidence that the Cochrane Collaboration uses for all interventions, many of the mammography trials were considered substandard some not even worthy of inclusion in the review. This time the Canadian study stood out as among the bestconducted studies, while the Swedish and American studies were told they have the major flaws.
I must admit, I was taken aback at the criticism of the American study. Maybe it was a little bit of misplaced nationalism on my part. Maybe it was because the HIP was a "classic" study that I have always admired (it really was the first large-scale trial of cancer screening). Maybe it was because the findings of the HIP fit my beliefs about cancer testing: namely, that early detection in this case mammography works best when it's not too sensitive (that is, when the mammograms don't reveal every tiny abnormality of the breast).
So I wanted to critically examine what the Cochrane Collaboration had to say about the HIP study. Some of their complaints struck me as nitpicking. They were concerned, for example, that the number of patients in the mammography group was different in different publications (variously 30,131, 30,092, and 30,239), and likewise for the amount of patients in the control group. While I find this kind of inconsistency irritating, I can't believe it has much effect on the study's findings. However, two problems were raised that definitely do call the validity of the study's findings into question problems corroborated in the book written by the primary investigator of the HIP study. I include them here to provide some flavor of the complex issues that arise in doing research on early cancer detection.
Our website is not responsible for the information contained by this article. Webworldarticles.com is a free articles resource thus practically any visitor can submit an article. However if you notice any copyrighted material, please contact us and we will remove the article(s) in discussion right away.
This article was sent to us by:
Fiona Landman at
08152010
1. Mesothelioma Treatment
All articles in this directory are property of their respective authors. Additionally, read our Privacy Policy
© 2010 WebWorldarticles.com - All Rights Reserved. Partners: Gunblade Saga