Thursday, April 5, 2012

The truth about cancer screening

Cancer sucks.

Unfortunately, we suck equally bad at detecting cancers in their early, curable stages.

Cancer is the second leading cause of death in the United states.  In 2009 it killed roughly 560,000 people.  There are an astonishing 200+ types of cancer.  The table below is a rough estimate of the number of deaths caused by the leading killers.

Let's pick the low-hanging fruit first - check out lung cancer.  Can you believe how bad it is compared with the others?  Few people survive it.  Thank you, big tobacco, for all the pain and suffering you've caused.  

So with all of the deaths from lung cancer, you'd think we would screen for it pretty aggressively, right?  After all, it's potentially curable if found early.  So we probably screen smokers starting at a young age so we don't miss any, right?  

Wrong.  Truth is that we don't even screen for it at all.  No joke.  Shocking, right?  Most people will find that appalling and even unacceptable.  Before you get your undies all in a bunch, read on.  It's not because we don't want to cure lung cancer - it's because we just don't have a good way to screen for it, safely diagnose it, and effectively treat it.  Lung cancer is a bitch.  It's tricky.  And that's why it's the number one killer among cancers.    

And what about breast cancer?  There was national outrage last year when the US Preventive Task Force changed recommendations for the age at which breast cancer screenings should begin.  For those that are unaware, the recommended age for routine mammography was bumped up from 40 to 50.  Public outcry ensued.  It's still a very controversial issue. 

Much of the outrage stems from a misunderstanding of screening tests.  Health screenings are, in general, a very misunderstood science.  The misunderstanding stems from the two major fallacies of healthcare.  See my previous blog post for a detailed explanation.   In short, people like to be screened for different things because they assume that 1) screening tests are accurate, and 2) there are beneficial treatments for the whatever is being screened for.  I'm not sure when these fallacies became popular belief, but they certainly are out there and they are deeply ingrained into the American psyche.  

There's an old adage in medicine - you better be careful what you order, because you might find something you don't want to.  The public will hear this and be outraged - how can they not want to find something?  They want to know everything that is wrong.  Doctors, on the other hand, understand that most of the time the "something" that you incidentally find is rarely anything serious, but it will take you on a long, wild goose chase that leads to more and more tests, some of them being potentially harmful.

Of all the 200+ cancers out there, the US Preventive Task Force only provides strong recommendations to screen for three of them.  Three!  Colon, breast, and cervical cancer are the ones that we routinely screen for.  They are the only cancers that we 1) have good, accurate tests for, and 2) we have good treatments for.  For all of the other cancers out there, we either suck at detecting them early, or we suck at treating them.  If you can't reliably detect or treat cancer, there's no sense even screening for it.

Let me break down the four reasons why cancer screening is not as simple as you would think it is.  If you understand these principles, you will understand why we do things the way we do.

1) We don't have many good cancer screening tests

Recall the first great fallacy of healthcare - not all tests are accurate.  It would be nice if we could just take a simple blood sample and tell with near 100% certainty that you do or don't have a certain cancer, but unfortunately such tests just don't exist.  Maybe someday they will.  But for now, we don't have them.  And the few blood tests that we do have (e.g. prostate cancer screening test - PSA or prostate specific antigen) are not very accurate.  They are notorious for providing either false positive or false negative results.  The few exceptions to this rule are the tests we do for colon, cervical, and breast cancer.  Colonoscopy, pap smear, and mammography are all very accurate tests.  That is one of the primary reasons that we only test for these cancers - because we have good tests for them.

2) Screening may cause more harm than good

Sounds blasphemous, right?  Recall that no test is without potential harm.  Let's look at esophageal cancer.  The only way to reliably diagnose it is by doing an EGD (small scope that goes into the esophagus).  EGD has the potential to cause perforation, or a tear in the esophagus.  It's a serious complication that can lead to death.  The Mayo Clinic put out a study in 2010 showing that their EGD perforation rate over an 8 year period was 0.033%.  Doesn't sound like a lot, but if every male in America over age 50 (roughly 50 million) was screened on a yearly basis for esophageal cancer, it doesn't take long before you have caused more perforations (0.00033 x 50,000,000 = 16,500) than have died from the disease (0.04 x 292,540 = 11,701)!  If you are going to do an invasive screening test, it has to do more good than harm - and by a significant amount.  Colon cancer screening is done primarily by doing a colonoscopy (done with a scope just like an EGD), but the prevalence of colon cancer is so much higher that we help many more people than we harm.   We know that a small portion of people will die from the screening procedure, but we save many more people from colon cancer than we kill doing the test, so the benefit outweighs the risk.  Kinda weird to think about it that way, but that's the way it is.

3) There isn't proof that screening for cancer leads to less deaths

This is also difficult to believe, but for some cancers it's true.  Recall that not all treatments are beneficial.  We've learned a difficult lesson over the past 20 years with prostate cancer.  Back in the 1990's there was a push to screen for it more regularly.  Guess what happened?  A lot more cases of prostate cancer were diagnosed (duh).  The real kicker, however, is that despite the increase in diagnosis, the rate of prostate cancer related deaths has not declined.  If you are going to screen for a cancer, it makes sense that the treatment for it should do some good.  But as strange as it sounds, detecting prostate cancer doesn't equal a higher cure rate.  So if you're going to test for cancer, you need to make sure you can treat it well enough to make a difference. 

4) It doesn't make financial sense

This one might be the most controversial.  Some of you will advocate that financial concerns shouldn't even be part of the equation - testing and treatment for cancer should be done regardless of the cost.  Remember the term TINSAFL you learned in high school economics (There Is No Such Thing As a Free Lunch)?  It applies to medicine as well.  If we were to screen every person in America for brain cancer with an MRI every year of their life from age 20 - 80, the cost would be staggering.  That test alone would probably bankrupt Medicare.  The truth is that we just don't have the money to do expensive screening tests on a mass scale.  You can hate it and disagree with it, but it's just the way things are. 

So for a screening test to be worth anything, it has to be cheap, accurate, cause minimal harm, and lead to a good outcome.  It's amazing that we have any tests that fit this criteria.  New tests are constantly being developed, and hopefully someday we'll have the technology we need to cut down on the number of cancer-related deaths.  But for now, all we have is all we have, and it's the best we've got.

I wish that things were different, and someday perhaps they will be.  Maybe 200 years from now people will look back at the year 2012 and call it the dark ages of cancer screening.  Maybe by then we'll have simple blood tests to accurately pick up all cancers - or perhaps even we'll be able to screen DNA and accurately predict what kind of cancer you will be susceptible to.  And hopefully we'll be much better at treating it.  But for the time being we'll do the best we can - even if it means only screening you for three types of cancer.

In summary, there is a whole science behind cancer screening that most people really don't understand very well.  It't much more complex than I've outlined here.  There are very smart people that look at this stuff for a living, and we have to trust that they are making the right decisions.  Next time you hear about a change in cancer screening recommendations, hopefully you'll wait a bit before jumping to any critical conclusions because now you understand the science behind it all.  

This post turned out to be more of a downer than I thought it would.  I'm really not a pessimist.  A realist, yes, but not a pessimist.  In order to try to add at least a bit of humor, I've added an amazingly funny, entertaining music video that was created by a fellow internist on cancer screening.  Enjoy, and stay tuned!


  1. I don't think your post was a downer at all, just realistic, as you said. I worked for 6 months as the only triage nurse for all outpatient pulmonary referrals for a large urban area (1 million) plus huge enchachment area and was shocked to see how many people were referred in for suspected lung CA. One woman I called who had been smoking a pack a day for 30 years (she was 50) was quite upset that I could not give her a next day appointment (instead of the 7 days I offered her) because as she told me "my GP told me that this is going to kill me". My two thoughts were: yes, it will kill you, so 7 days is not going to make a difference AND after smoking that much for that long what the fuck did you think that 3 month cough was going to show up as?
    I had to change jobs because it was making me very angry.

    1. Ahhhh yes the patient who abuses their body and then expects incredibly timely, overly comprehensive workup and treatment for a disease they could have prevented. Even better are the patients who claim "well my doctor never told me cigarettes are bad for me, so I never quit." Ummmm ok.