Tuesday, February 28, 2012

Rethinking shared decisions in prostate cancer screening

Last October, the U.S. Preventive Services Task Force provisionally recommended against screening for prostate cancer using the prostate-specific antigen (PSA) test, eliciting a variety of reactions from medical and patient advocacy groups. The New England Journal of Medicine published one of the most thoughtful responses by Mary McNaughton-Collins and Michael Barry, two physicians who have done a great deal of research on the psychological and physical harms related to false-positive tests, an all-too-common occurrence in men who receive periodic PSA testing. They respectfully disagreed with the USPSTF's "D" (don't do) rating for this preventive service, arguing that the rating should have instead been a "C" (don't do routinely):

A grade C recommendation would allow the patient to be involved in the decision to skip or choose a PSA screening test, after a discussion with a primary care provider about the magnitude of the known harms and the potential for some benefit. The patient could then provide his perspective on how he views the trade-off. Weighing the pros and cons to make a decision about PSA screening is an individual process, and different well-informed men will make different decisions. A grade D recommendation removes the patient from the equation and puts the physician in the central position of discouraging use of the test. Uncertainty in medicine is more common than we usually let on, and the way to address uncertainty is to allow patients the central position in decision making. ... We do not believe that anyone but the patient should decide whether the small and uncertain benefits of PSA screening are worth it.

In 2010, the American Cancer Society took a similar stance, recommending that men age 50 years or older make "informed choices" about whether or not to receive the PSA test after being provided with detailed information about the risks and benefits and tools to help them make the decision. On the surface, both of these recommendations seem more reasonable that the USPSTF's. Rather than making a unilateral decision for every patient based on imperfect population-level evidence, why not let each patient make the choice for himself?

The answer is that primary care physicians have already been trying this approach for many years, and it does not work. Numerous studies have shown that no matter how much information men are given in any format, the vast majority still choose to receive the test, and most of them end up worse off because of that decision. In fact, as a new study concludes in the Archives of Internal Medicine, once prostate cancer is diagnosed via PSA testing, the men most likely to receive aggressive ("curative") treatments are those who are least likely to need it - men who would have been better off not getting the test in the first place.

A case in point: in the same issue of the journal, an academic oncologist who exclusively cares for prostate cancer patients tells the story of his own concerning PSA test, prostate biopsy, and subsequent decision to undergo radical prostatectomy at age 56. Five years later, he is cancer-free, but has permanent right-sided weakness as the result of the surgery, which was in retrospect unnecessary given the limited spread and good prognosis of his tumor. He laments: "Even the most informed patient (me in this case) has difficulty making a truly informed decision." And if a physician who spent decades caring for patients with prostate cancer has trouble making the right call, what hope is there for the rest of us?

Physician-researcher H. Gilbert Welch (author of Overdiagnosed, previously reviewed on this blog) recently challenged the consensus that informed decision-making is the best approach to PSA testing:

I worry that excessive fealty to informed decision-making in cancer screening may distract us from more important issues—like what is bothering patients right now. It may also lead physicians to abdicate our role as advisor—physicians simply transfer responsibility to their patients. And the problem appears to be asymmetric: we seem less inclined to recommend against than to recommend for. ... Why are we so reluctant to make value judgments about things not to do to patients? Is it because we are so fearful of potential litigation? Is it that we have become so financially conflicted? Or is it that we fear being accused of caring about cost? Maybe clinicians need the Task Force to make the call for us. And, for PSA testing, they did. If I was pressed to get off the fence, it's the call I would have made.

I couldn't agree more. Shared decision-making in PSA screening is an attractive concept that in practice leads to more harm than good. The Task Force was right to recommend that it stop, and that we instead advise patients not to get the test.


  1. I agree completely. The benefit is minimal to none and there is a clear potential for harm. That's what we have the Grade D for! Somebody has to take a stand, and not let the specialty organizations (and politicians!) bully them into watering down their recommendations. While I am a big proponent of discussing side effects, risks and benefits with my patients, there is a limit to all this. And, really, sometimes our patients just want us to quit all the damned mealy-mouthed stuff and make a clear recommendation.

  2. I, too agree. Your stance and analysis has been consistently correct and on the money.
    This is a good example of how something that seems, on the surface, to be a balanced position is indeed incorrect.

  3. In case the post isn't clear, I do favor shared decision-making in medical situations where there is either substantial uncertainty about the net benefit or a test or treatment (e.g., PSA screening prior to the results of the randomized trials), or for "close calls" where there is no single right answer and patient values of various outcomes are critical. I do not believe that there is enough uncertainty surrounding the net benefits of the PSA test to justify shared decision making, nor is it a particular close call. As Michael Barry writes today in an insightful NEJM commentary on shared decision making (http://www.nejm.org/doi/full/10.1056/NEJMp1109283), "for some decisions, there is one clearly superior path." Like screening for ovarian and testicular cancer, physicians should actively discourage screening for prostate cancer with PSA because it is unlikely to provide any benefits, and almost certain to produce harm.

  4. I am a great believer in informed consent and shared decision making, but I also appreciate this guidance from the PSTF. There have been so many times I have reviewed the risks and benefits of PSA screening only to have the man look at me and ask "What do you think I should do, doc?" or "What would you do?" I will now feel more comfortable answering those questions directly.

  5. This is such an important issue, but how to get the word out about it!

  6. eleven years ago, as a 54 year old, i was treated for aggressive PC. although i experienced morbidity's, I'm here to write this comment. two acquaintances, not tested, are no longer here. i wish in 2001 i had dr. bennett's favorable pathology report, unfortunately, I didn't. doctor or not, don't mistake one man's regret for wise policy. mm (also private email sent to dr. lin)

  7. I followed my doctor's recommendation, approx. 4 years ago to have a PSA. The results were such that he sent me for the second test. The resulting biopsy confirmed that I had aggressive prostate cancer. I doubt that I would be around today, if it weren't for the PSA.
    It wouldn't bother me to see the "DRE" digital rectal exam thrown out.

    1. I have to wonder about the phrase "confirmed that I had aggressive cancer". It is my impression (via Nortin Hadler's "Worried Sick") that aggressive cancer can't be identified with any useful degree of certainty, and that is why the criteria of the Gleason test is very unreliable. Some, like Hadler, believe that the mostly deformed cells in the prostate of someone over 65 can be described as "cancerous"; knowing which, if any, will become aggressive is unknown. If so, it would follow that your surgeon just wanted to make you feel like the surgery was really worth it.

    2. "approx. 4 years ago to have a PSA"

      "I doubt that I would be around today, if it weren't for the PSA."

      Even the ERSPC results show there is no visible likehood of PSA screening making any difference to prostate cancer mortality 4 years after PSA screening commences. Basically, there is actually little, if any, doubt that Anonymous would be around at the time of his comment if he had not had PSA screening 4 years ago. Claims such as "I doubt that I would be around today" demonstrate the false beliefs that many people are capable of and which make it so difficult to get people to act in a rational way.

  8. I agree that "aggressive" cancer is a retrospective diagnosis that can only be confirmed by observing what happens to patients in the absence of treatment. Naturally, most men prefer to be treated, especially if the pathologist and/or surgeon tells them that not doing so will result in premature death. But longitudinal studies of men who chose "watchful waiting" show that even men with prostate cancers with high Gleason scores are much more likely to die of something other than prostate cancer (e.g., heart disease), and, therefore, won't benefit from cancer treatment.

  9. dr. lin, in my case, i had 6 of 8 cores positive, with a very high percentage, and a high gleason score. on the other hand, your doctor bennett was a classic case for watchful waiting; one core at 5% , gleason 6 or lower. you have reinforced your recommendation against psa testing because of a doctor (bennett) who obviously has remorse about his decision to have himself treated, against established guidelines. furthermore, because you now are contracted to lecture on the anti-psa point of view, you are no longer an honest broker for this discussion. no offense, the "aggressive" case from march 3, 11:38am.

  10. Dear Anonymous - If by "honest broker" you mean someone with no opinion on this topic, of course I don't qualify. But please don't imply that I'm financially conflicted about it. I was invited to give a talk about the benefits and harms of PSA screening, not paid to promote any particular point of view. (Incidentally, the chief medical officer of the American Cancer Society has estimated that every "free" PSA offered by a hospital produces $5000 in revenue associated with payment for the subsequent workups for abnormal test results. That's the type of financial conflict we should all be concerned about.)

    Finally, I'm not saying that you aren't alive today because your primary care clinician ordered a PSA test. There's no way for me to know that, and I don't question your personal decision, which was clearly the right one for you. What I'm saying is that not all men in your situation will benefit from treatment, due to competing causes of mortality. That fact has been established by multiple population-based studies, for example: http://www.ncbi.nlm.nih.gov/pubmed/15126804

  11. dr. lin, my apology for any implication of financial gain, it was not my intention. as stated in my first comment, i was 54 at the time of diagnosis, so yes, i benefited from the psa test. although you may be justified by statistics, i know several men who have died from lack of a timely psa test. in the last 10 years the concept of watch waiting has been greatly expanded, providing an alternative to unnecessary treatment. i wonder if this direction isn't gender based; I couldn't imagine the medical profession telling women not to get mammograms. they even give them to women in their eighties. i suppose men should just man up! thank you for printing my comments. mm

  12. Dr. Lin:

    I continue to argue that a rational person could quite reasonably decide to have a PSA test. If so, your position that no doctor should give any choice to the patient is unreasonable.

    Why is the choice of a PSA screening defensible as a rational choice by a patient?

    (1) If you think the European study is either superior to the U.S. study, or at least deserves substantial weight, then this study suggests that as of a 12 year follow-up, PSA screening-induced diagnosis of prostate cancer reduces the risk of mortality from prostate cancer by 5%. This must be compared with a 50% risk of serious side-effects of treatment. A tough choice, but one that many patients would regard as a reasonable alternative to consider. d
    (2) The real issue is not screening but how to treat prostate cancer. On this front, the data suggest that if active surveillance is combined with aggressive treatment once biopsies suggest the prostate cancer might be getting aggressive, then prostate cancer treatment which might follow treatment will appreciably reduce the risk of death from prostate cancer. Again, the tradeoff is a high risk of serious side-effects. And different men will view these tradeoffs differently. So, if the real issue is treatment choices, why so much focus on screening? It's not really the issue.
    (3) Finally, doesn't it cause you any pause to argue that men should not be even presented with this choice , given than many men disagree with you? You're implicitly arguing that we're all irrational, ignorant, or foolish, at least those of us who aren't urologists, who might have financial motives. I certainly don't think you're a fool. I think you may be a bit over-confident that your judgment of the evidence, and your values about the tradeoffs, should trump those of all other perspectives.


    Tim Bartik

  13. Dr. Lin,

    I like the perspective. I'm not an attending yet, so until I am in your shoes, behind a closed door facing patient questions on a significant medical test, I can't truly understand the dilemma of shared decision-making.

    I do have a question, though:

    Did the USPSTF try to impact shared decision-making? For example, since the previous recommendation was overall inconclusive, and physicians reported difficulty counseling patients, did the USPSTF seek to clarify the issue with either a firm screen or no screen recommendation?

    Or was your blog a retrospective, general observation on the way physicians can now move forward?

  14. Tests like monograms are more dangerous then helpful because of the radiation. We seem to forget that the negative reaction to radiation is dependent on that person's health. I just feel a lot of times the medicine community has to many tests. We have all these people with tests and have no results from them. Not all but most.

  15. 59yo just got a 4.9. Had long discussion with the Dr. about why he even ordered the test as the preventative task force rates it a “D” for not recommended.

    With that reading, there is a 75% chance of actually having nothing and a 19% chance of having something so low-grade that some researchers assert should not even be called “cancer.” Yet the urologist would render me impotent and incontinent for that 19%, oh yes, and the $20K they would get for the surgery.

    I pointed out that a biopsy is not a trivial procedure and can have complications.

    I said I no intention of a follow up test in 3 months and he said okay just wait until next year. Said I might not even have it done then. He said that was okay, it was my decision.

    As for those who say it saved their life, and a recommendation against PSA testing is irresponsible consider this analogy:

    Groups recommends against going to the casino to increase your wealth, most people end up poorer.

    One person says, "You're irresponsible to recommend against the casino, I won the jackpot there and am set for life."