What tests are early "screens" for cancer?
May 13, 2008 8:01 AM Subscribe
If we always hear, "They could've beaten his cancer if they'd only found it sooner," then why aren't we all getting screened for cancer long before the doctor orders the tests? What tests can I get now (MRI? CT?) to check for such things? I assume that money is the issue (isn't it always?), but if I'm willing to spend the cash, what do I need to do for me and for my family members?
This is what expensive private health care is for. You will be screened regularly or whenever you want, but you'll pay a lot for it. Many high profile people have caught their cancer's very early because they can afford the best health care in the world. Robert De Niro springs to mind.
posted by fire&wings at 8:10 AM on May 13, 2008
posted by fire&wings at 8:10 AM on May 13, 2008
Well, for one thing, CT scans are associated with a small increased risk of cancer, so some folks suggest it's wiser to not get them unless there's a reason to get them. Other tests may have similar tradeoffs. Or not.
posted by aramaic at 8:17 AM on May 13, 2008
posted by aramaic at 8:17 AM on May 13, 2008
why aren't we all getting screened for cancer long before the doctor orders the tests?
False positives.
posted by knave at 8:17 AM on May 13, 2008 [1 favorite]
False positives.
posted by knave at 8:17 AM on May 13, 2008 [1 favorite]
Yeah, false positives. Also, not all cancers show up on scans in the early stages. Talk to your doctor about this, they should be able to explain the details of this.
There are also certain ages where certain tests are most appropriate. Some of them are invasive, so it's not really recommended to get them unless necessary.
posted by fructose at 8:30 AM on May 13, 2008
There are also certain ages where certain tests are most appropriate. Some of them are invasive, so it's not really recommended to get them unless necessary.
posted by fructose at 8:30 AM on May 13, 2008
Don't discount that for many of the people this is said about, they were either not getting regular checkups or under the care of a less than great doctor. People have a strange tendency to ignore things until the last possible moment. They don't go to the doctor when they first notice that they can't walk up a set of stairs without getting winded. They wait until their heart feels like it's going to explode. Same thing for that dull ache or "weird birthmark".
posted by gjc at 8:43 AM on May 13, 2008 [1 favorite]
posted by gjc at 8:43 AM on May 13, 2008 [1 favorite]
Money isn't always the issue. My dad's cancer wasn't caught because his doctor was basically incompetent and he wasn't really vigilant about checking his own test results. Although there was a huge jump in certain test result numbers over a two year gap, they were still within the "normal" range. This "normal" range was meant to cover men between the ages of 50-75, and my dad jumped from being at the low end of normal one year at age 49 to being at the very highest end of the threshold at age 52. Any reasonable person would have noticed something amiss, but the doctor didn't even show him the results, just checked off "normal" and sent him on his way.
So, in addition to always going in for physicals, insisting on regular check-ups, blood work, pap smears, mammograms, and anything else that's recommended, make sure you know and trust your doctor and keep records of your own test results, and check them yourself for inconsistencies. If anything looks amiss, ask about it, and get a second opinion if you're still not satisfied.
posted by booknerd at 8:46 AM on May 13, 2008
So, in addition to always going in for physicals, insisting on regular check-ups, blood work, pap smears, mammograms, and anything else that's recommended, make sure you know and trust your doctor and keep records of your own test results, and check them yourself for inconsistencies. If anything looks amiss, ask about it, and get a second opinion if you're still not satisfied.
posted by booknerd at 8:46 AM on May 13, 2008
Forgot to add- as technology increases, the detriments to getting more regular scans are reduced.
For example, my dentist recently changed over to an electronic x-ray setup. Instead of film, he sticks a sensor in your mouth and the x-ray shows up on a computer screen. He said the electronic sensor requires far less exposure to x-rays than film. Something like one tenth.
posted by gjc at 8:52 AM on May 13, 2008
For example, my dentist recently changed over to an electronic x-ray setup. Instead of film, he sticks a sensor in your mouth and the x-ray shows up on a computer screen. He said the electronic sensor requires far less exposure to x-rays than film. Something like one tenth.
posted by gjc at 8:52 AM on May 13, 2008
This is what expensive private health care is for.
No, it's really not.
False positives are a concern, but they are caused by the quandary of pre-test probability and likelihood ratios. The only way we know how to interpret tests in medicine (labs, scans, etc) is by testing the tests to see how they work out. We compare a test result to some sort of "gold standard" -- in a perfect world, a tissue sample -- to say, "Okay, this person we know has cancer, let's see in what proportion of people we know to have cancer does this blood test also differ for people with cancer."
So when your doctor orders a test, he or she has a pre-test probability that you have, say, cancer. Each test then has a certain ability to detect a cancer, if there is one, and each test also has a certain likelihood that it will have a negative result if there's NOT cancer, too. And after some math, out pops the post-test probability. I actually found a really great illustration of this with a little interface here.
As you can see, if your pre-test probability of having prostate cancer is 0.1% (say, a 5-year old), even a test that detects with 99.99% probability prostate cancer will make it only a little bit more likely that your 5 year-old actually HAS prostate cancer.
So here's our problem--you could test everyone by doing a whole body scan, but unless you have symptoms and/or risk factors, the odds that you have disease X are vanishingly low. So these whole body scans would a) cost a lot b) not find much c) irradiate everyone and cause actual cancer and d) would create a lot of false positives. If we're talking CT scans, often times there are incidentalomas--random growths that cause no harm and you would have never known they existed--that end up getting scanned and worked up. This causes incredible expense, worry, and risk. And they're essentially just all false positives.
See also: lead-time bias, which is when you have a new, "better" test that can diagnose a disease earlier, but doesn't affect a person's survival: they basically just learn earlier that they have disease X, but it isn't caught early enough to impact whether the disease can be cured/stopped/removed.
posted by gramcracker at 9:12 AM on May 13, 2008 [6 favorites]
No, it's really not.
False positives are a concern, but they are caused by the quandary of pre-test probability and likelihood ratios. The only way we know how to interpret tests in medicine (labs, scans, etc) is by testing the tests to see how they work out. We compare a test result to some sort of "gold standard" -- in a perfect world, a tissue sample -- to say, "Okay, this person we know has cancer, let's see in what proportion of people we know to have cancer does this blood test also differ for people with cancer."
So when your doctor orders a test, he or she has a pre-test probability that you have, say, cancer. Each test then has a certain ability to detect a cancer, if there is one, and each test also has a certain likelihood that it will have a negative result if there's NOT cancer, too. And after some math, out pops the post-test probability. I actually found a really great illustration of this with a little interface here.
As you can see, if your pre-test probability of having prostate cancer is 0.1% (say, a 5-year old), even a test that detects with 99.99% probability prostate cancer will make it only a little bit more likely that your 5 year-old actually HAS prostate cancer.
So here's our problem--you could test everyone by doing a whole body scan, but unless you have symptoms and/or risk factors, the odds that you have disease X are vanishingly low. So these whole body scans would a) cost a lot b) not find much c) irradiate everyone and cause actual cancer and d) would create a lot of false positives. If we're talking CT scans, often times there are incidentalomas--random growths that cause no harm and you would have never known they existed--that end up getting scanned and worked up. This causes incredible expense, worry, and risk. And they're essentially just all false positives.
See also: lead-time bias, which is when you have a new, "better" test that can diagnose a disease earlier, but doesn't affect a person's survival: they basically just learn earlier that they have disease X, but it isn't caught early enough to impact whether the disease can be cured/stopped/removed.
posted by gramcracker at 9:12 AM on May 13, 2008 [6 favorites]
Curing cancer isn't the big picture, it's merely a means to an end, that end being quality of life. Excessive medical intervention in someone already healthy will result in false positives and mistreatment, as well as the finding of misc. conditions that can be treated but (in terms of the big picture) are better left untreated, but which almost certainly will be treated once people become aware of them. The net result can be reduced quality of life. So as with everything, it's about striking a balance. Here's-a-blank-cheque-I-want-every-test-every-month is a poor balance. If you have an overactive imagination and the tests give you peace of mind, then that's also part of the equation, and a higher outlay on testing and the resulting greater chances medical misadventure, may be worth it to you. Or seem worth it... initially. :-)
posted by -harlequin- at 9:39 AM on May 13, 2008
posted by -harlequin- at 9:39 AM on May 13, 2008
If I understand your question correctly, the answer is because there isn't a super-machine that we can pop into and quickly have it tell us everything that's wrong.
The real answer to your question is that we are being screened. That's what a pap smear is, a mammogram, digital rectal exam, PSA test, etc. If you're at risk and visiting your doctor, you'll be screened. As far as imaging methods go, there are limitations. Most of the devices/modalities currently in use have a maximum resolution of about 1 cubic millimeter. That means a tumor smaller than a cubic millimeter won't even be detected (that's about 1 billion cells, hardly small). There's the radiation exposure from CT as mentioned. As for MRI, you can't just hop and see what turns up. It isn't that simple. You need to know what you're looking for before you start the scan. You can't stick your head in a see if you have cerebral edema, a tumor or MS. Just doesn't work that way. There are a lot of different programs, settings, etc.
Maybe in the future (think jetsons/star trek era) there will be a handy device that can do everything, but not yet.
posted by sero_venientibus_ossa at 10:42 AM on May 13, 2008
The real answer to your question is that we are being screened. That's what a pap smear is, a mammogram, digital rectal exam, PSA test, etc. If you're at risk and visiting your doctor, you'll be screened. As far as imaging methods go, there are limitations. Most of the devices/modalities currently in use have a maximum resolution of about 1 cubic millimeter. That means a tumor smaller than a cubic millimeter won't even be detected (that's about 1 billion cells, hardly small). There's the radiation exposure from CT as mentioned. As for MRI, you can't just hop and see what turns up. It isn't that simple. You need to know what you're looking for before you start the scan. You can't stick your head in a see if you have cerebral edema, a tumor or MS. Just doesn't work that way. There are a lot of different programs, settings, etc.
Maybe in the future (think jetsons/star trek era) there will be a handy device that can do everything, but not yet.
posted by sero_venientibus_ossa at 10:42 AM on May 13, 2008
Best answer: A book that covers these issues in some depth is Should I Be Tested for Cancer? It really helped me when I was deciding about whether to get mammograms in my 40s.
posted by not that girl at 11:59 AM on May 13, 2008
posted by not that girl at 11:59 AM on May 13, 2008
One thing I found interesting in the book I mentioned is that the better our scans get, the more people we find have itsy bitsy cancers somewhere in their body. But one thing that isn't known about these tiny cancers is how many of them actually proceed to causing symptoms. It's possible, for instance, that lots of us are walking around with tiny cancers that will resolve themselves, become encapsulated and never advance, or advance so slowly that we'll die of something else before they start causing symptoms. So one risk of cancer screening is not just false positives, but true positives that lead to unnecessary treatment that is more toxic and damaging than that particular cancer ever would be.
posted by not that girl at 12:06 PM on May 13, 2008 [1 favorite]
posted by not that girl at 12:06 PM on May 13, 2008 [1 favorite]
Gramcracker answered this perfectly. The only thing I would add is what one of my Medical School profs used to always say: "Any test you order that isn't medically indicated, will be positive." It's a curse that says in essence, if you're not using your brain when you order a test, you are sure going to use it to deal with the result. It probably took me the first 5 years of my career to realize this; if a patient is asking for some test, I always want to be a nice guy and order it and we'll deal with the fallout later. The problem is that if your pretest probability is really low, ie the 5 year old with a positive screen for prostate cancer, you *still* have an abnormal test to deal with. In the legal climate that medicine exists in, no doctor is going to just sit on a test like that even though they know it's not real. So you start doing things that are really bizarre and have *no* medical basis, like doing prostate biopsies on 5 year olds.
posted by Slarty Bartfast at 2:17 PM on May 13, 2008
posted by Slarty Bartfast at 2:17 PM on May 13, 2008
I FPPed some links that led to a good discussion of these issues last year. Essentially Gramcracker has it nailed though.
posted by roofus at 4:46 PM on May 14, 2008
posted by roofus at 4:46 PM on May 14, 2008
This thread is closed to new comments.
posted by genial at 8:07 AM on May 13, 2008