Some Implications Of the New PSA Recommendation

DrRich | October 24th, 2011 - 7:05 am


The United States Preventive Services Task Force created another hub-bub recently when they released their latest, updated recommendations on whether men should routinely have PSA testing for the early detection of prostate cancer. The USPSTF’s recommendation was simple and straightforward: No.

News reports on this new recommendation have fairly accurately portrayed the arguments on both sides. Proponents of PSA testing are in an uproar because prostate cancer kills many men, and its early detection makes it easier to treat. Without PSA testing, the early detection of prostate cancer is difficult and often impossible. But those siding with the USPSTF point to randomized clinical trials showing no significant reduction in mortality in populations of men who have had PSA screening, and further, that men who have PSA screening end up having a lot of very unpleasant and expensive medical procedures which can leave them with life-altering side effects.

DrRich is by no means an expert on prostate cancer or PSA testing, but as it happens he is an American male who is within the age group addressed by this new recommendation. So he indeed has a legitimate interest in whether the USPSTF has made a wise decision or not.

To help him decide whether this new recommendation is a reasonable one, DrRich has gone to the source: to the document published by the USPSTF itself in announcing its new recommendation. Helpfully, the USPSTF has laid out in detail the specific clinical studies it relied upon, and the rationale it used, to synthesize the results of those studies into a concrete recommendation.

The USPSTF document points out two major conclusions which can be gleaned from the medical literature on PSA screening. First, when PSA screening is applied to large populations of men, it is difficult to demonstrate a reduction in mortality. Of two large clinical trials comparing men randomized to PSA screening to those randomized to “standard care,” one found that PSA screening yields a relatively small but statistically significant reduction in cancer-related deaths, but the other showed no mortality benefit. So, given a large population of men eligible for screening, doing PSA testing appears to yield a benefit that is either small or non-existent. And as a result, from a public health standpoint a recommendation to do widespread PSA screening is simply not justifiable based on current evidence. And this finding accounts for the USPSTF’s new recommendation.

But the second major conclusion that is revealed by the medical literature is that, for men in whom screening has actually detected early prostate cancer, subsequent treatment significantly reduces mortality. This result addresses one of the big questions often raised about early detection of prostate cancer, namely, whether the cancers detected by PSA screening actually require treatment. Many of these early cancers apparently never cause death, so many have speculated that “watchful waiting” might be a reasonable course of action rather than aggressive prostate treatment. But the USPSTF’s review of the relevant studies shows that when early-stage prostate cancer is identified, the best clinical trials available show a significant reduction in cancer-related death and all-cause mortality with either surgical prostatectomy or radiation therapy.

As the backdrop for these two major conclusions, the USPSTF strongly emphasizes the drawbacks of PSA screening. This screening often leads men to experience some very bad outcomes from prostate biopsies, or from therapy for prostate cancer. The very nasty complications resulting from these procedures are all too frequent, and are very difficult to even think about let alone experience. Furthermore, pursuing all those  positive PSA tests is extraordinarily expensive for the healthcare system. The reasoning offered by the USPSTF in making their new recommendation relies heavily on the price which men must pay, in terms of complications, in pursuing the results of a positive screening test.

DrRich has long been disturbed by the state of the art of both prostate cancer screening and prostate cancer treatment, by the lack of obvious progress in improving these things, and by the seeming complaisance with which many urologists seem to accept the status quo. PSA screening appears far too sensitive (too many false positives, leading to too many biopsies). Prostate biopsies often yield both false positive results (detecting cancers that are probably clinically meaningless) and false negative results (missing cancers that are clinically important). And the numerous treatments available for treating prostate cancer (all of which are very unpleasant) have not been rigorously compared, leaving the various “camps” of urologists to argue that their pet treatment is the best one, and all those other urologists have their heads up their ass.

All this confusion and uncertainty places the patient faced with the prospect of whether to have a PSA test, or worse, with newly-diagnosed prostate cancer, in a complete quandary, and apparently with no objective means to resolve what he ought to do next. But despite all these shortcomings, the urology community has aggressively turned PSA screening and the cascade of uncertainties (and resultant procedures) that flow from it into a burgeoning industry, to the extent that one must wonder how badly these specialists want to clarify the current muddle. And for this reason, it is difficult to take the loud objections being made by the American Urological Association against the USPSTF’s new recommendations very seriously.

So from a public health standpoint, the USPSTF recommendations on PSA screening seem reasonable to DrRich.


DrRich keeps coming back to the second major conclusion from the USPSTF’s analysis of the medical literature on prostate cancer screening: Even with all the drawbacks associated with PSA screening, and even with all the conjectures about whether these early prostate cancers really need to be treated after all, it turns out that if prostate cancer is detected by some screening technique, then treating that cancer saves lives. And DrRich notes that while the USPSTF dutifully describes this result in the body of their report, they do not mention it in the Abstract of their report, and they do not seem to have given it much weight, if any, in their final recommendations.

But it seems to DrRich that this is an important result, and ought to be taken into account. It should not be simply brushed off as irrelevant, or unworthy of notice. It begs to be explained.

How can it be that, on one hand, offering PSA screening to a large population of men does not seem to result in much overall mortality benefit, whereas on the other hand, if you do find prostate cancer when you screen for it, then treating that cancer significantly reduces mortality?

Most likely the explanation lies in the dilution effect. The moderate (but statistically significant) benefit of treating early prostate cancer is washed out when those patients are included in a much larger population of men who are eligible for screening, and who may or may not have prostate cancer, which may or may not be detected adequately by current screening techniques, and if it is detected may or may not be treated.

To see how such a dilution effect might operate, let’s consider seat belts. Everyone knows that seat belts save lives. So let’s do a study to prove it. One way to do this would be to compare the mortality rates of people who are in automobile accidents, according to whether they were or were not wearing seat belts. Odds are it would be fairly easy to show a mortality benefit with seat belts. But now let’s compare the mortality rate of all drivers over a 5 or 10 year period according to whether they were wearing seat belts, regardless of whether they were ever in an automobile accident. DrRich suspects you would not be able to demonstrate a mortality benefit with seat belts in this second study.

The PSA screening studies that the USPSTF relied on to make their PSA recommendations are analogous to this second seat belt study. The prostate cancer treatment studies that did show a mortality benefit are analogous to the first seat belt study.

Please note that DrRich is not comparing PSA screening to wearing seat belts. Wearing seat belts does not lead to a lot of unnecessary expense, nor does it create life-altering side effects. PSA screening, given the state of the art, is neither inexpensive nor benign.

But despite its major drawbacks, PSA screening does detect early prostate cancer. And if you measure outcomes from the point where the prostate cancer is actually diagnosed (instead of from the point where you decide to do PSA testing), survival is measurably increased by its early detection and treatment.

So the dichotomy is explained. From a public health standpoint, where you have to decide what the result will be on a large population of individuals if some screening test is implemented, it does not make sense to do PSA screening. But if you are an individual who might have prostate cancer, in whom the early detection of that cancer might save your life, then it might make sense to do the PSA screening. (Whether it does or not depends on how you, the individual, assign relative weights to the notion of dying from prostate cancer vs. the inconvenience, expense, pain, and possibly horrible side effects from PSA testing and what it might lead to.)

So while from a public health standpoint it would be a mistake to recommend widespread PSA screening, from an individual standpoint either decision – to have or forgo PSA screening, depending on how you yourself weigh the tradeoffs – would be entirely reasonable.

But individuals are not allowed to decide this for themselves. This is no longer the kind of decision which individual doctors and patients are supposed to be making any more. In fact, it is now illegal to do so.

And this, Dear Reader, describes the problem with the USPSTF decision on PSA screening. For, in fact, the USPSTF is no longer making mere “recommendations,” which doctors and patients might take into account if they wish as they decide whether some preventive healthcare measure is right for an individual patient. Rather, the USPSTF rulings now determine whether you and I, as individuals, will or will not receive that preventive measure.

Obamacare, which is now the law of the land, makes the USPSTF the final arbiter of which preventive services are to be covered by private insurers (Section 2713), by Medicare (Section 4105), and by Medicaid (Section 4106). Only those that have achieved a grade of A or B by the USPSTF will be covered. And if you believe you will be able to purchase for yourself PSA screening (or any other medical service which Obamacare has decided not to cover) you have not been paying attention. Perhaps you can do so today (if you’re not on Medicare or Medicaid), but probably not for long.

What all the news outlets have forgotten to mention, in their coverage of the PSA controversy, is that the USPSTF has been officially converted from a panel that simply makes recommendations which doctors and insurance companies can take or leave alone, into a panel that determines definitively what is covered and what is not – and indeed, into the chief tool by which our leaders will seek to withhold expensive preventive services.

And while in the particular case of PSA testing, he is not particularly sorry to see the new USPSTF recommendation, DrRich submits that, given the general nature of medical screening tests, it is child’s play to set up a clinical trial that would “prove” (given the expense of the test, the false positives, the false negatives, the side effects of the test itself, the side effects and expense of the follow-up tests needed to see whether a positive screening test is truly positive, the expense and side effects of the treatment that will be used if the diagnosis is actually confirmed, the relative efficacy and inefficacy of that treatment, not to mention the dilution effects of having to screen a large number of individuals to find the relatively few who actually have the condition of concern and will benefit from its treatment) that there is no preventive screening test you could name that produces an overall benefit to the population.

DrRich has long predicted that the brilliant people in our news media will be continually “surprised” each time some heretofore sacred medical screening test is declared by the all-powerful USPSTF to be, after all, useless.

This being the case, can we just stop pretending that Obamacare is all about prevention, disband the USPSTF altogether, stop funding any screening tests whatsoever and any research being done to develop new ones, and call it a day? That would be much more transparent, not to mention cheaper, than stifling preventive medicine in the painfully slow and deceptive way we are doing it today.

5 Responses to “Some Implications Of the New PSA Recommendation”

  1. WarmSocks says:

    How can it be that, on one hand, offering PSA screening to a large population of men does not seem to result in much overall mortality benefit, whereas on the other hand, if you do find prostate cancer when you screen for it, then treating that cancer significantly reduces mortality?

    What a surprise that treating cancer reduces mortality! Some doctors order screening tests but never look at the results. Consider Lonnie, diagnosed with prostate cancer. He phoned all five of his brothers and suggested that they see their doctor for a PSA test. One had just had a “complete physical” two months before, but went in anyhow just to placate his brother. “Oh, yes,” his doctor said, “your bloodwork included a PSA. Since we didn’t contact you, everything’s fine.” But the brother pushed and wanted to know his exact results, wanted to send a copy to Lonnie to prove he was okay. First doctor ever who deemed a PSA of 487 not significantly elevated.

    Perhaps the fact that sometimes tests are performed but the results are not evaluated accounts for the findings that in the general population, PSA testing doesn’t lead to a reduction in mortality. It’s only when the discovered cancers are treated that we see a reduction in mortality.

  2. Bob Clements says:

    I get a kick when I hear that tired old refrain “you should discuss this with your doctor”. In my experience a thorough discussion of the risks and benefits of this screening takes a good 5 or 6 minutes. It can take 10 if they have questions. Of course a single such session should be enough for any one patient, but rarely do they remember the complex arguments. Especially since they’ve been bombarded by the different cancer societies’ drumbeat of “get your screening”, “don’t die of embarrassment”. So you need to give them a refresher course every year. This is just one tick of the long check list we need to do every year. I just can’t understand why everybody is avoiding primary care.

    • DrRich says:


      Fear not. The USPSTF is from the government and they’re here to help. Thanks to them, instead of spending all that time sorting through the pros and cons of screening tests with your patients, you can just say no.


  3. Jan Krouwer says:

    Your post expresses very nicely what I would have liked to say. But …

    Regarding: “And the numerous treatments available for treating prostate cancer (all of which are very unpleasant) have not been rigorously compared, leaving the various “camps” of urologists to argue that their pet treatment is the best one, and all those other urologists have their heads up their ass.”

    The treatment, proton beam therapy, is not unpleasant and has minimal side effects. And with > 200,000 men diagnosed every year in the US, a survey based retrospective analysis would provide better information than any RCT.

  4. phwest says:

    The other reason the two results are possible is that both studies are balancing positive and negative outcomes, but while the same number of positive outcomes (Successful treatment) occur in both populations, only some of the negative outcomes are included in the “cancer present” study. To the extent that a false-positive PSA result drives further testing and treatment with potentially negative outcome, this is a cost that is excluded from the later cancer present study.

    Not that this would change the conclusion – individuals are the best judge of the relative risks and rewards of any diagnosis and treatment plan. Particularly with something like prostate cancer, where there are so many tradeoffs.

You can leave a response, or trackback from your own site.

Leave a Reply