Kevin, M.D. - Medical Weblog

if we can't convince doctors of EBM then there's certainly no convincing the general public.

what especially impresses me is siegel's article in wsj as a major editorial, and this reply today:


Merely asserting that "my patients want to know if they have cancer" is not a valid way to deal with this frustrating puzzle. Dr. Siegel's use of one illustrative patient also flies in the face of evidence-based medicine: the plural of anecdote is not data.

http://online.wsj.com/article/SB117505385170051530-search.html?KEYWORDS=siegel+lung+cancer&COLLECTION=wsjie/6month
 

I don't know kevin. If something showed on your wife's mamogram, would you want her to wait 6 months for a retest?
 

How come Kevin just takes these pot shots in his blog entries, but never actually participates in the ensuing discussion in the comments?
 

Kevin's quote from my post is the prevailing attitude among women, and women doctors. Breast cancer detection and EBM just don't dance well together these days - there in lies the crux of our confusion and worry.

I've had a few atypical mammograms all before 40 and I can tell you my husband would be the first to demand further testing - not just for the emotional need to know, but because there are few reassuring answers to questions about waiting. For EBM to work, we have to have enough evidence to justify the "not testing" path.
 

I think the main problem is that most women and even some doctors are not aware of the numbers behind the endless promotion of testing.

The chance of having at least one false positive in 10 years of yearly testing is often downplayed. I saw websites that say - "oh it's only 10% first time less thereafter", forgetting to mention that this is after a single mammogram and that the cumulative chance of at least one false positive after 10 yearly mammograms is around 50% (less in Europe). About 25% of these abnormal results end in biopsy, most of which will be negative. I took these numbers from PDQ NCI website, by the way. Not to mention that only in a small subset of true positives earlier diagnosis would make a difference.

If every woman with an abnormal or ambiguous result is sent for a biopsy, we'll see 50% of women having biopsies. And these have risks as well. Even if all of them sent for ultrasound and/or MRI, the number of biopsies will increase, probably overdiagnosis as well.

Biopsies have risks, overtreatment even more so. These might be small, but if we look at the actual probability of somebody's life being prolonged, even small risks may be important.

The women are also overestimate the potential benefit. When the benefit is expressed as relative risk reduction it sounds much higher that it really is. USPSTF estimates probability of mammograms saving one's life is 1/1200 after 10 yearly mammograms for women in their 50s; 1/1700 for women in their 40s. Even if we ignore CNBSS studies and take really optimistic estimates, say 1/500 for women in their 50s, the risks that come from false positives results and overtreatment don't seem that small.

I wonder if women would be just as enthusiastic about investigating every ambiguous results (or even mmmograms themselves) if they had been aware of the numbers.
 

"For EBM to work, we have to have enough evidence to justify the not testing path. "

True, one needs to really look at available research before jumping to conclusions one way or the other. The evidence, however, is clearly in favor of waiting, as was shown anecdotally to be the right approach in brin's case as well (although she'll never understand that):

(i hope this link works)
http://www.utdol.com/utd/content/abstract.do?topicKey=genr_med/43947&refNum=46-48

by the way, my wife is a pediatrician (and a woman) and she agrees with the evidence-based approach.
 

One point worth making is different breast types dictate different levels of aggressiveness. And, of course, family history drives some of the advice we are given.

What advice we follow is also heavily dependent on how that advice gets communicated. Clearly this discussion uncovers many aspects of breast health that can be improved upon beyond EBM and communicating statistics.
 

I thought Kevin said that physicians did all the unnecessary testing because of liability "concerns". Now it's the patients who are requiring it?

It must be difficult for him to stay consistent.