PSA: Medicare's poster child for a bad screening test (but what about all the other bad screening tests??)
January 23, 2016
PSA has always been a poster child of cancer screening. On the one hand, it does prevent the spread of prostate cancer in some people screened. In fact, a recent study showed that since doctors have been ordering PSA tests, the incidence of metastatic prostate cancer has declined. (Welch, NEJM, 10/15) However, there has never been evidence that PSA screening has reduced survival at all, and there are a tremendous number of false positive PSA’s leading to over-testing and over-treatment. (See BRCT showing that out of 1000 men treated for prostate cancer that may have been discovered incidently by PSA testing, 600 will develop impotence.) After fewer PSA were ordered starting in 2012 (due to recommendations against PSA), the number of early stage prostate cancers has also declined. (JAMA, 11/15) This has led to fewer false positives, fewer treatment-induced complications, and no increase in prostate cancer death.
What is wrong with detecting early stage prostate cancer in larger numbers? Isn’t that the whole point of screening: find the cancers before they spread? The reality of PSA screening is not that simple, a fact common to many other screening tests. While 28,000 men die of prostate cancer annually, there is no difference in that death rate between people screened and not screened with PSA tests. In fact, the vast majority of lethal prostate cancers have already spread beyond treatment before a PSA test could detect them. Contrarily and equally concerning, PSA tests pick up many prostate cancers that will never cause any harm. Almost half of men over the age of 60 actually have prostate cancer, but virtually none of them will die of the disease. Yet if they get PSA screening they will be told they have prostate cancer and be subjected to treatments that are unnecessary and harmful. Says Dr. Welch, who has studied prostate cancer: “PSA treatment has led a lot of men—our 2009 estimate was over one million since the test was introduced in 1987—to be treated for a cancer destined to never bother them. And treatment frequently leads to impotence and can cause incontinence and bowel problems.” (NYT, 1/7/16)
Currently many organizations, including the US Preventive Services Task Force, recommend against PSA screening for the reasons stated above. Now Medicare has declared that, through its quality indicators, it may punish doctors who order the test. But is that the most reasonable approach? Dr. Welch, who has done more than virtually any researcher to demonstrate the dangers of PSA testing, wrote a very convincing article in the New York Times (http://www.nytimes.com/2016/01/07/opinion/why-doctors-shouldnt-be-punished-for-giving-prostate-tests.html) imploring Medicare not to penalize doctors who order the test. Citing the virtues of shared decision making, he suggests that doctors and patients together should make a decision about testing based on accurate and well communicated data. Some patients may still want to get a PSA despite the clear dangers of testing, some may decide to forgo it; it is really up to each patient. Rather than create another “quality indicator” upon which doctors will be judged, doctors should be rewarded if they take the time to discuss the merits and difficulties with PSA screening, something that Medicare already does with lung cancer screening.
Ironically, PSA screening is not much different than most other screening tests, but most other tests actually are endorsed and paid for by Medicare. A screening test is performed on someone without symptoms in the hope of finding and fixing problems before they become severe, but virtually no screening tests actually adheres to that neat script; most people who are screened face the same perils as do men getting a PSA. Negligible significant survival benefit (sometimes just the opposite is true), a lot of false positive, and an excessive amount of over-treatment for “diseases” that if left alone would never hurt the patient; this is the story of virtually all screening. Mammograms also prevent little or no death, have a lot of false positives that lead to unnecessary tests and treatments, and yet Medicare punishes doctors who do not order mammograms. Stress tests, carotid scans, bone density testing, dementia testing, CT scans for lung cancer, most components of the physical exam, virtually all lab tests; when done as screening tools all of these share the same downsides of PSA tests, and yet all are paid for and approved (sometimes required) by Medicare despite no data showing their value and a lot of data showing their harm. Why is Medicare picking on PSA when virtually the entire screening repertoire is just as bad?
Shared decision making and patient centered care seems to be the answer to what ails us in Medicare. Rather than encouraging some screening tests and discouraging others, rather than building contrived quality indicators that reward and punish doctors based on a manipulated script, Medicare should encourage, even require, accurate discussions between patients and doctors about these tests before they can be performed. The data must be presented in absolute risk/benefit terms, ideally using a visual aid such as the theaters we utilize in Interpreting Health Benefits and Risks. Once patients are able to choose their plan of care wisely, it is very likely that the number of unnecessary and harmful tests will be curtailed, with great saving for Medicare, and better health for our country’s elderly.