Does screening mammography work?

SDN Members don't see this ad. (About Ads)

I came across this and thought it was an interesting read. Curious to see the thoughts of other people.

http://www.thennt.com/screening-mamm...ducing-deaths/

Here are a few important quotes from the commentary to consider:

While overall mortality remains unchanged, in this analysis breast cancer mortality is reduced by approximately 15%. If accurate, this would represent a NNT of approximately 2000. However, breast cancer mortality is not as important as overall mortality, because individuals deciding whether to undergo screening mammography will typically want to avoid death, rather than simply avoiding death from one possible cause.

Do you agree w/ that, or do you believe that breast cancer deaths are somehow worse than deaths due to other causes? In our society, I worry that we've created a hierarchy of death, in which a cancer death is much worse than one due to any other cause.

Importantly, overall mortality may not be affected by mammography because breast cancer deaths are only a small fraction of overall deaths.

Something we, especially women, so often forget.

Finally, if it is true that breast cancer deaths are reduced it has been estimated that for every one patient who avoids death from breast cancer approximately 10 to 20 women are treated unnecessarily as cancer patients, typically receiving surgery, radiation, and chemotherapy.3,4 This additional estimate of harm is not represented in the NNH statistics listed above.

A very hard truth to consider--screening creates far more cancer patients than it saves. Is that a worthwhile trade-off? Also consider this when you hear someone say "screening mammography saved my life".

"however an equal number of individuals appear to lose their lives due to mammography"

I think I would take strong exception to this comment.

so if we already don't recommend self breast exams and clinical breast exams have no mortality benefit so if we stop doing screening mammograms then cancers are going to be picked up by what? patients presenting with gross abnormalities suggestive of later stage of disease?

I haven't gone through the primary literature but my concern is this:


If the same number of people are dying in the mammography and control groups but fewer are dying of breast cancer in the mammography group, then I can think of only two possibilities. Either the data is faulty and breast cancer deaths are being classified as something else (which biases the study to supporting screening) or, even worse, the patients are dying as a side-effect of receiving mammography and the follow-up treatment.

The former means we're wasting money on a useless intervention and the latter is a far more concerning situation where we are simply stopping patients to irradiate, mutilate, and bill them on their way to the grave.

This isn't to argue that mammography is some evil. It's just a test which provides information we may not be capable of properly utilizing. Maybe this guy's wrong but he paints a convincing argument that it's at least worth considering that maybe we got caught up in a parade of pink ribbons and early-detection dogma.

The rest of medicine is full of examples where we realized that we lack the predictive tools to make it a good idea to aggressively track down asymptomatic problems...


As I said, I haven't read enough of the primary literature to feel capable of making an informed opinion. I though the guy made an interesting argument and wanted to see if there is an informed rebuttal out there.

Originally Posted by Miss Alyssa

so if we already don't recommend self breast exams and clinical breast exams have no mortality benefit so if we stop doing screening mammograms then cancers are going to be picked up by what? patients presenting with gross abnormalities suggestive of later stage of disease?

Yes. You do not get to implement a poor screen just due to sentiment ....

Depends on how dense your knockers are.

Their endpoints in the review are too close. They're looking at 10-13 year survival studies of people who presumably don't have cancer at the time. Also there's not enough data about the studies to draw other conclusions.

Furthermore, here's a follow-up study which has interesting findings.

http://www.ncbi.nlm.nih.gov/pmc/articles/PMC102363/

Originally Posted by drizzt3117

Their endpoints in the review are too close. They're looking at 10-13 year survival studies of people who presumably don't have cancer at the time. Also there's not enough data about the studies to draw other conclusions.

Furthermore, here's a follow-up study which has interesting findings.

http://www.ncbi.nlm.nih.gov/pmc/articles/PMC102363/

That paper is again is only about breast cancer mortality. What good is being done if there is no change in over-all mortality? A program of early and aggressive decapitation in breast cancer patients would decrease breast cancer death to nearly zero with no change in over-all mortality. But that's not exactly a miraculous triumph of modern medicine...

If anything, that paper draws attention to the concerning fact that even the impact on breast-cancer mortality is still under legitimate debate.

Not familiar with mammography lit, but there's pretty strong evidence that routine PSA's do more harm than good (at least as it was under the study conditions at the time - not sure if protocol has changed). I think the author is a Kaplan or some such

Originally Posted by loveoforganic

Not familiar with mammography lit, but there's pretty strong evidence that routine PSA's do more harm than good (at least as it was under the study conditions at the time - not sure if protocol has changed). I think the author is a Kaplan or some such

Gotta say that I disagree with this (I'm only responding because I've been following the debate since it's blown up this past week). I think those studies show that PSA work-up can cause harm, but the entire body of evidence shows an overwhelming benefit for the screening.

There's been 50% reduction in prostate cancer mortality since PSA became available. Yes, we should screen smarter and more conservatively, but PSA is the best biomarker available for prostate cancer and it saves lives, without a doubt. That cancer is insidious and will metastasize silently, any biomarker we have is extremely valuable when used correctly.

I think there are two big issues - mortality isn't the only useful measure here and "when used correctly" is key. Prostate cancer occurs at high rates in older populations without significant impairment to the individual, who will often die of something unrelated. The 'holy crap I have cancer' jumps people into proactive treatment, with the associated significant sequelae, rather than watchful waiting

Originally Posted by loveoforganic

I think there are two big issues - mortality isn't the only useful measure here and "when used correctly" is key. Prostate cancer occurs at high rates in older populations without significant impairment to the individual, who will often die of something unrelated. The 'holy crap I have cancer' jumps people into proactive treatment, with the associated significant sequelae, rather than watchful waiting

Right on both counts.

But I think the bolded is where the topic gets murky, and incidentally where the USPSTF has placed a lot of their focus. It seems that the only people in the conversation who are comfortable with watchful waiting are the epidemiologists. At the end of the day, I think a lot of physicians would've been more satisfied if the panel said "we need more data" instead of "stop screening."

Originally Posted by loveoforganic

I think there are two big issues - mortality isn't the only useful measure here and "when used correctly" is key. Prostate cancer occurs at high rates in older populations without significant impairment to the individual, who will often die of something unrelated. The 'holy crap I have cancer' jumps people into proactive treatment, with the associated significant sequelae, rather than watchful waiting

Some people don't have significant impairment; others get bone mets and terrible pain. How can you predict who gets what? Gleason score and... PSA.

http://www.ncbi.nlm.nih.gov/m/pubmed...054145/related

Originally Posted by DeadCactus

That paper is again is only about breast cancer mortality. What good is being done if there is no change in over-all mortality? A program of early and aggressive decapitation in breast cancer patients would decrease breast cancer death to nearly zero with no change in over-all mortality. But that's not exactly a miraculous triumph of modern medicine...

If anything, that paper draws attention to the concerning fact that even the impact on breast-cancer mortality is still under legitimate debate.

There we go, nice find. I need to read the full article when I get a chance...

Originally Posted by KinasePro

Right on both counts.

But I think the bolded is where the topic gets murky, and incidentally where the USPSTF has placed a lot of their focus. It seems that the only people in the conversation who are comfortable with watchful waiting are the epidemiologists. At the end of the day, I think a lot of physicians would've been more satisfied if the panel said "we need more data" instead of "stop screening."

I think we're pretty much in agreement, and I think the comment about the epidemiologists is salient, for the reasons below:

Originally Posted by drizzt3117

Some people don't have significant impairment; others get bone mets and terrible pain. How can you predict who gets what? Gleason score and... PSA.

I agree that PSA has utility in theory, but I think the results of PSA too often lead to poor medical decision making. Patients generally aren't comfortable with letting cancer be. Accepting death isn't something that comes easy for most in our society, but you really have to accept it to be ok with letting known cancer go untreated. Physicians, from a medical malpractice standpoint and a "not wanting the patient to die" standpoint, aren't going to be too keen on the idea either. So we end up with incontinent patients without sexual function who live, until they die of MI, stroke, etc.

Again, theoretically, PSA has utility, but I'm not sure how we can implement it in a way to yield better outcomes (factoring in quality of life).