Each mammography exposures you to radiation.
False positive mammography results lead to biopsies, which have some risk of infection and other adverse results.
New Mammogram Guidelines
- Thread starter Puppycow
- Start date
Each mammography exposures you to radiation.
False positive mammography results lead to biopsies, which have some risk of infection and other adverse results.
Where are you getting the idea that this is the rationale behind the current recommendations?
Linda
The question to be answered is whether more tests are fruitful.
Generally, the risks of screening are:
Loss of 'health' - you extend the period where you are 'ill' at the expense of the time that you are 'healthy'.
False sense of security over a false-negative which may delay the possibility of early diagnosis in the future.
Any risks and complications associated with the screening test.
Any risks and complications associated with the diagnostic tests.
Loss of resources available for the use of improving or maintaining health.
Specific to mammography - only about 1/3 of those who are diagnosed early show prolonged survival, cumulative radiation exposure associated with the screening test leading to other cancers, additional radiation exposure and the risks/complications associated with biopsy with the diagnostic tests.
A risk-benefit analysis considers the risks and benefits. This is relatively easy when the outcomes for the risks and benefits are the same (i.e. deaths as a result of screening vs. deaths prevented), but when they are different, they are either simply compared qualitatively or some sort of quantification of harm is applied to different outcomes.
ETA: The 'cost' of obtaining the desired outcome is the harm from obtaining the diagnosis.
Cost analyses can look at the cost of obtaining a monetary benefit (cost-benefit) such as productivity or avoidance of the cost of more expensive treatments through prevention, the cost of obtaining a useful outcome (cost-utility) such as Quality Adjusted Life Years, or the cost of obtaining a specific effect using one treatment vs. the cost of obtaining that specific effect using a different treatment (cost-effectiveness).
In the case of mammography, the recommendations seemed to be based on a qualitative risk-benefit assessment. The problem seems to be that there is substantial variation in how individuals will qualitatively compare the harm of anxiety vs. the harm of death.
Linda
Last edited:
ponderingturtle
Orthogonal Vector
- Joined
- Jul 11, 2006
- Messages
- 54,545
bruto;5335103 I understand the balance problem said:
bruto
Penultimate Amazing
What I hear so far on #1 is that the necessity of more frequent testing is not sufficient to outweigh the potential harm from either the tests or the treatments they lead to. Those I've heard defending the guidelines assert very vigorously that cost was not a factor in this weighting.
The answer to # 2 seems to vary with the test in question. There is said to be an actual cumulative risk from the radiation involved in a mammogram, so this must be balanced against the likelihood that it will do any good. Reducing the number of mammograms you get in a lifetime reduces the individual risk of damage from the test during your lifetime. So if testing every other year gives about the same rate of benefit as every year, you do well to reduce your exposure to the radiation. By contrast, we read that the risk from a pap test is not the test itself but the risk that a positive or ambiguous result will lead to unnecessary treatment (this is also an added factor in the mammogram guidelines, but not the only one). In this case, the benefit of changing guidelines seems murkier. Obviously if you run a risk, every time you get a test, of being harmed, then the fewer times you take the test the fewer times you'll expose yourself to the risk, but the risk on each instance is unchanged. If you're shooting loaded dice, the fewer times you shoot, the fewer times you'll lose over all, but the dice are still loaded, and the odds still the same every time you shoot. It is here that epidemiology and individual risk diverge. If half as many people take the test each year, then of course half as many are harmed by it, but if the risk remains the same every time you take it, then the statistic is not useful as a basis for individual choice.
The question then, it seems, becomes one of whether there is any better way to fight the loaded dice than simply playing less frequently. If the risk of the test is that the policy for interpreting and treating the results is indadequate, then the guidelines are at least questionable, if not bogus, and simply defer the risk for a year. If the results of the test are so difficult to interpret even by the best practice that a high risk of overtreatment is inherent in the process, and harm from the best available treatment still outweighs the likelihood of benefit, then perhaps there is no better way, but then it raises the question of whether one should have the test at all. Some experts I've heard have made exactly that argument with regard to PSA tests for men, for example. What I have yet to hear here is an argument for the new guidelines that makes that distinction clearly and justifies the decision as something other than a statistical dodge for an unaddressed problem.
bruto
Penultimate Amazing
Yeah but the obvious solution is to improve what is done with the results not to simply fail to obtain useful results.
Linda
bruto
Penultimate Amazing
If your interpretation requires the assumption that a group of professionals are acting like morons, wouldn't it occur to you that maybe your interpretation is off?
Linda
bruto
Penultimate Amazing
I don't recall saying that professionals were acting like morons, and I can guarantee that I did not use that word, but if I implied it I apologize. The articles I've seen and radio interviews I've heard state "overtreatment" as one factor in the mammogram guidelines, and the major factor in the pap guidelines. How you interpret the word and its implications is up to you.
What remains unanswered, at least to me, is the question of whether "overtreatment" is avoidable through better decision making on the part of doctors when positive results are found, or whether there is simply no good way to judge when further treatment is warranted. The implication from what I've seen is that it is, at least partially, the former, especially for the pap test, in which the overtreatment cited has to do with how HPV infections are handled.
edit to add: interesting as this discussion is, I hope to be able to read further responses, and obtain any further enlightenment that can be had. I will willingly accept correction if you can come up with some further information to counter my suspicion that the epidemiological rationale for the new guidelines, sound though it may be, has little to say to the question of individual choice. However, I'm about to go on a trip, with likely very limited and spotty internet access, and limited time even when the access is there, so please forgive me if I cannot continue the discussion except, perhaps, to grunt occasionally, or (you hope) to say "point taken, you win."
Last edited:
Niggle
Basement Cat's, Evil Twin Skippy
- Joined
- Nov 17, 2008
- Messages
- 674
Is there any way to quantify the harm caused by the annual exposure to mammogram-level radiation? For instance, does the mammogram itself increase the risk of breast cancer due to the radiation exposure?
Recently, some doctors have begun recommending ultrasounds in addition to mammograms for women whose breast tissue is very dense (my own, for instance; it has taken as many as eight tries to get clear pictures of both breasts during a regular mammogram). Are ultrasounds reliable enough to be used instead of mammograms?
Which is cheaper, a mammogram (X-rays) or an ultrasound?
Recently, some doctors have begun recommending ultrasounds in addition to mammograms for women whose breast tissue is very dense (my own, for instance; it has taken as many as eight tries to get clear pictures of both breasts during a regular mammogram). Are ultrasounds reliable enough to be used instead of mammograms?
Which is cheaper, a mammogram (X-rays) or an ultrasound?
Delvo
Дэлво Δε&#
All it takes is one ionization in the wrong part of the wrong molecule, so we know that the risk from even a single photon has to be there. But with mammography, like X-rays in general, the amounts of radiation are so low that the magnitude of the risk is similar to or less than the margin of error, which has made it impossible to separate from background noise in the data so far.
I've been wondering why ultrasound isn't the standard tool all along myself. Usually, ultrasound is not only risk-free but more informative for soft tissue without bones or gas inside it.
I'm not talking about your interpretation as to whether "overtreatment" is a concern, but rather your interpretation of how the professionals chose to deal with overtreatment. This is what you said:
"The guideline is based on the expectation of over-treatment. And the solution put forth is to skip the test, but is that a rational approach to the problem? If a certain percentage of the women getting the test will test positive and be overtreated, there's little to be gained by deferring the error for a year. It's true that fewer women per year will be harmed, simply because there will be half as many tests, and presumably if a woman who has HPV but the ability to get over it unaided happens to go through the entire process in a non-testing year, she'll dodge the harm, but if she gets it at any other time, there's no reason here to expect she won't be over-treated just the same as before, because the cause of the harm is not really being addressed. In the meantime, a person who has something more serious will lose a year of early detection.
I understand the balance problem, I think, but a person is not a population, and a plan that reduces the number of persons harmed in a given period may look good, but if it does not also reduce the likelihood of an individual being harmed, it's a hollow victory."
You ask whether this is a rational approach, and of course, it would not be. If screening intervals affected both groups equally - those who would benefit and those who would not - then to choose simply on the basis of making the population numbers look good at the expense of individuals would be unreasonable. If you are willing to assume that the professionals who made these recommendations are unreasonable, then it wouldn't occur to you that perhaps the assumption ("screening intervals affect both groups equally") was wrong. If you are not willing to assume that the professionals are unreasonable, then it would occur to you to consider whether your assumption is wrong.
The reason that intervals can be lengthened in response to over-treatment is because lengthening those intervals preferentially affects those who would not benefit from screening in the first place. It changes how the dice are loaded.
Changing the underlying case prevalence by decreasing false positives without decreasing true-positives or increasing false-negatives is one way to improve decision-making.
Linda
Ivor the Engineer
Penultimate Amazing
- Joined
- Feb 18, 2006
- Messages
- 10,588
For those interested, the wiki page I linked to has a nice graphic which explains why this is the case.
Lengthening the interval between tests moves the line to the right, which results in a bigger increase of the area (i.e. probability) under the true-negative (TN, dark blue) curve than the increase in area under the false negative (FN, light blue) curve.
bruto
Penultimate Amazing
Quickly checking in here and a quick response before my battery dies, I don't think the problem is false positives, but true positives and how they are responded to, at least in the case of the pap test, but I haven't had time to research further. What I read is that it is true posives involving HPV infection that lead to aggressive treatment, loosely termed by the media "overtreatment." If this is the case, then I would have thought that the odds of overtreatment would be the same in any individual instance of the test, though of course the fewer instances, the fewer chances for it to happen.
Puppycow
Penultimate Amazing
1. No, even if you are super rich it would not be worth it if you are asymptomatic and have no special risk factors and if you value things the same way as they did for purposes of the study. Since the risks and benefits are kind of apples and oranges, you have to make some arbitrary value judgments.
2. Mainly the risk of a false positive and unnecessary treatment including biopsies, anxiety, more time spent in a hospital or clinic (possibly in close proximity to people with communicable diseases). Secondarily the radiation from the mammogram, although that's probably not a big factor. A day at the beach would probably expose you to more harmful radiation.
I'm sorry that wasn't clear. I referred to two groups - those who would benefit from screening and those who would not. The former would represent true positives and negatives, while the latter would represent false positives and negatives. An individual who was treated on the basis of a pap test, but who did not benefit from that treatment (i.e. the outcome would be the same either way if the abnormality would have resolved on its own) would represent a false-positive.
Linda
bruto
Penultimate Amazing
Ok, that makes sense if you assume from the start that "overtreatment" is integral to the test. If there is only one rational way to deal with the results, and the whole thing is a bundle, then yes, a lack of benefit from treatment might be said to equal a false positive. I have yet to see this addressed.
That is exactly what I was addressing and what the recommendations from professional bodies are addressing.
"Overtreatment" represents a situation where the patient does not benefit from the treatment. A positive test is the finding of an abnormal pap test. If you are using the test to identify those people who will benefit from the treatment, then Overtreatment represents a test failure - a false positive. The true-positive screening test is not the finding of an abnormal pap test. It is the finding of an abnormal pap test in someone who enjoys a better outcome (the early treatment of progressive disease) because of the result of that test. The finding of an abnormal pap test in someone who has the same outcome with or without the treatment (recovery in either case) is Overtreatment or a false positive.
If you extend the screening interval, you will reduce the number of positive results (true and false) obtained overall. You seem to assume that the number of positives in each group (true and false positives) will be reduced by the same amount. But that's not what happens. Instead, the number of true-positives remains essentially unchanged. It is a reduction in the number of false positive which accounts for the overall reduction. So now your proportion of true-positives to false positives has become much larger, which allows for better decision-making in the face of a positive result.
Linda
Last edited: