A Lesson in the World of Screening for us All
One of the most surprising aspects of modern medicine is how hard it is to create a screening test that works well.
After all, what could be simpler than screening? Screening just involves measuring something in everyone and finding those few whose measures are abnormal. And then, knowing which of us has the problem, we can cure or eliminate it before it does any harm.
This is, in fact, the idea behind screening. Find the problem early, reap the benefits of taking care of the problem before it causes any harm.
Some screening test methods do just that. The best of them is the newborn screening test for a condition called PKU. It was the first of all the newborn screening tests and it has all the features of a good screening program:
- The test is simple, inexpensive, and extremely safe.
- Essentially everyone with PKU tests positive, and everyone without PKU tests negative.
- If you are found to have PKU in the newborn period, there is an action that can be taken that will ensure no harm will come to you. If you have PKU and no one knows it, the intervention, which is simply to change your diet, cannot be offered and severe mental retardation will result. The test result makes a dramatic difference.
- There is a system that allows all Americans to be tested at birth, for all positive results to be reported to the family and their doctor, and for those with PKU to receive the life-saving intervention.
The dirty little secret of screening, however, is that most techniques to do this do not work very well. With few exceptions, most screening tests used in medical practice have a tremendous overlap, that is, people with positive results often do not have the problem. Also, too often there is no clearly defined intervention or treatment that leaves the person reliably better off.
And yet, there is an explosion in the number of screening tests being developed. As a result, every American is asked to be screened for a widening variety of problems, often without a good screening technology in place. The test is available, but the test is often not a very good performer, not like the PKU test at all.
This has all come to our attention with the recent announcement that the screening test for prostate cancer, PSA, does not work very well. This was the finding of the nation's leading authority on screening tests, the United States Preventive Services Task Force (USPSTF). The USPSTF exists for only reason, to study screening tests and see if they work.
How could a screening test fail? Well, it turns out that when you study them, they nearly all fail in some way or another. The bottom line is that a screening test fails if, when all is said and done, it is likely not to help you.
How could a screening test turn out to cause more harm than good? Here are some ways:
- The test could be so expensive, painful, and/or dangerous, that you are left more hurt than the problem you are looking for could have made you.
- The test could tell far too many people that they have the problem or disease when in fact they do not. Imagine taking a test that told you had cancer when in fact you do not. This turns out to be the most common problem with screening tests.
- The test could tell far too many people that they do not have the problem or disease when in fact they do. This is a potentially devastating outcome, but fortunately, one problem current screening tests tend to avoid most of the time.
- When you look at everyone with a positive result and see how their problem eventually pans out, you might find they fare no better than those who skipped being screened who have the same problem. That is, early detection may not make any difference.
- For some tests, there is no treatment available to help. This makes the finding out early about the problem a rather futile exercise.
Figuring out how well a screening test actually performs turns out to be a tricky and complex task. That is why the United States Preventive Services Task Force (USPSTF) deserves nothing but gratitude. The USPSTF is a very thoughtful and careful group. They review thousands of studies and look at screening tests for years before coming to a decision. In the case of the PSA screening test for prostate cancer, they spent a very long time looking at the data, came to a conclusion in 2009, and then decided to spend two more years looking at the data again, to be sure they were advising us as accurately as possible.
What the USPSTF found was that the PSA test for prostate cancer is really a miserable failure as a screening test. About half the men who are told they have prostate cancer as a result of this test turn out to not have any sort of cancer that would ever cause them harm. Worse, of any group of men who have a positive PSA, no one can tell which of these men have actual, dangerous cancer, and which do not. That dooms all men with a positive PSA to endure the burden of thinking they have cancer, and undergoing serious and potentially dangerous treatments.
The USPSTF found that when all is said and done, no lives have been saved by the PSA screening test, but tens of thousands of men have been rendered impotent, incontinent, or both, for no very good reason.
It is the response to this finding that exposes the power of the screening test in today's modern medical world.
In a more rational world, one would expect the findings of an organization like the USPSTF to be accepted with tremendous interest and even gratitude. Years of careful review and study should be.
But the medical world has reacted to this report with a howl. Leading the charge are the urology associations. Their response is clearly emotional and contains little data. They state that they refuse to stand by and let their patients die by not having them screened with the PSA. Of course, they have no information that demonstrates dropping the PSA test would actually lead to anyone dying. Rather, they present stories of individuals whose lives appear to have been saved by the PSA test turning positive.
Now, we all know which story is most convincing:
A harrowing story of a 58 year old man who thought he was well, only to find his PSA level was elevated, getting a prostate biopsy that shows he has prostate cancer, and then having his prostate removed and cured of a disease that could have killed him if he hadn't been screened.
A highly technical scientific report looking at the sensitivities, specificities, negative predictive values, and positive predictive values of the PSA test. And in that same report, numbing tables and graphs looking at the statistical patterns of outcome for men screened with the PSA.
And yet, that really boring scientific report demonstrates that the 58 year old man in the first story has a very good chance of being misled. The odds are so good that he did not really face any threat from prostate cancer, that he would have been just as likely to live out a healthy life without the PSA test!
Turns out facts do matter.
Sadly, as an outstanding group of epidemiologists at Dartmouth have found (and reported in their book Overdiagnosis, an outstanding book that I have blogged on earlier), money often trumps facts. And the money element of screening is that it can turn a whole nation of healthy people into paying patients.
I present this story of the USPSTF, of PSA, of prostate cancer, to illustrate why we at Advanced Pediatrics take the science of screening so very seriously.
We look at every screening test recommended and look for data that groups like the USPSTF have on its performance. If the test is a lousy performer, we do not recommend it. We will never recommend a screening test simply because it screens for a serious illness, the test must actually deliver some benefit for us to tell you it's a good idea.
In the meantime, do trust the USPSTF. They are careful, deliberate, and quite wise.
Dr. Arthur Lavin
*Disclaimer* The comments contained in this electronic source of information do not constitute and are not designed to imply that they constitute any form of individual medical advice. The information provided is purely for informational purposes only and not relevant to any person's particular medical condition or situation. If you have any medical concerns about yourself or your family please contact your physician immediately. In order to provide our patients the best uninfluenced information that science has to offer,we do not accept samples of drugs, advertising tchotchkes, money, food, or any item from outside vendors.