Decision aids, PSA-based screening, and male decision-making

According to a newly published study in JAMA Internal Medicine, web-based and print-based decision aids do help men to resolve their own internal conflicts about whether or not to get PSA tests for risk of prostate cancer … but they don’t change men’s decisions about whether or not to get screened.

Unlike most studies of this type, which historically have involved relatively small numbers of men, this new study by Taylor et al. enrolled nearly 1,900 men who were randomized to one of three different arms of the trial, giving the results of this study a good deal more credibility than some of the earlier work on this topic.

Despite the recommendation of the U.S. Preventive Services Task Force that men should not be screened for risk of prostate cancer, many physicians and almost all prostate cancer advocates now take the position that the decision whether or not to have PSA tests should be based on patient risk, patient preference, and  the provision of information to patients that clearly spells out the potential risks and the potential benefits of regular PSA testing.

Taylor and her colleagues set out to compare the use of two different types of decision aid to “usual care” (a brief discussion between doctor and patient) in helping men to come to a decision about whether they did or did not want to have a PSA test to assess their risk for prostate cancer. The men were all between 45 and 70 years of age; they were racially diverse; they were recruited at three different clinical sites between 2007 and 2011; and they were randomized to one of the following forms of pre-test “education”:

  • Men in Group A were provided with a print-based decision aid.
  • Men in Group B were given a web-based, interactive decision aid.
  • Men in  Group C were given “usual care”.

The research team conducted telephone interviews with the patients three times: at baseline; at 1 month after use of the decision aid or provision of “usual care”; and at  13 months after the use of the decision aid or the provision of “usual care”. Patients were evaluated with regard to specific outcomes: prostate cancer knowledge, decisional conflict, decisional satisfaction, and whether the participants actually underwent prostate cancer testing.

Here are the study findings:

  • 4,794 eligible men were approached and asked to participate in the study.
  • 1,893 eligible men agreed to be randomized.
    • 628/1,893 (33.2 percent) were randomized to Group A (the print aid).
    • 625/1,893 (33.0 percent) were randomized to Group B (the web-based aid).
    • 626/1,893 (33.0 percent) were randomized to Group C (“usual care”).
  • At each follow-up assessment, men in Groups A and B exhibited significantly better prostate cancer knowledge and reduced decisional conflict compared to men in Group C (all P  < 0.001).
  • At the 1-month follow-up interview,
    • The standardized mean difference (known as Cohen’s d) in knowledge for men in Group B compared to men in Group C was d = 0.74.
    • The standardized mean difference in knowledge for men in Group A compared to men in Group C was d = 0.73.
    • Decisional conflict was significantly lower for men in Group B compared to men in Group C (d = 0.33).
    • Decisional conflict was significantly lower for men in Group A compared to men in Group C (d  = 0.36).
  • At the 13-month follow-up interview, these differences were smaller but remained significant.
  • Also at the 1-month follow-up interview,
    • More men in Group A (60.4 percent) reported high satisfaction that men in Group B (52.2 percent) or men in Group C (45.5 percent).
  • By comparison, at the 13-month follow-up interview,
    • Differences in the proportions of men reporting high satisfaction in Group A (55.7 percent) compared with Group B (50.4 percent) and Group C (49.8 percent)  were smaller and not statistically significant.
  • Actual PSA screening rates at 13 months did not differ significantly among groups.

Taylor and her colleagues conclude that

Both decision aids improved participants’ informed decision making about [prostate cancer] screening up to 13 months later but did not affect actual screening rates.

They further note that

Dissemination of these decision aids may be a valuable public health tool.

As with things like decisions about buying cars, men appear to have a very strong tendency to stick with their initial decisions once those decisions have been made. They will often hold firm to those decisions regardless of evidence to the contrary, and such behavior appears to correlate well with the “satisfaction” reported by patients with their decisions about treatment for prostate cancer — even when outcomes (to neutral eyes) might appear to be far less than “satisfactory”. Perhaps all any of this shows us is that men have a hard time admitting that their initial decisions might have been poor! This would appear to be supported by the idea that any available evidence (the print and web-based decision aids) simply help men to justify their decisions to themselves.

If this is true, then the idea that dissemination of these types of decisions aids “may be a valuable public health tool” is utterly spurious. Why? Because all that these tools may actually be doing is offering men a crutch to support a decision they have already made before they ever use any decision aid at all!

7 Responses

  1. In my opinion, all men should be screened with a PSA test by 60 years old. What needs to be done is what this site is doing, educating them so they can make prudent decisions after the test. Not knowing one has a cancer is not a solution.

  2. I don’t think that the information in this study necessarily shows that the decision tools don’t help. Perhaps, even though the overall screening rate was unchanged, the decision tool caused some change in which men were screened, so that the screening accorded more with each man’s individual valuation of the relative risks and their objective circumstances.

  3. Who said that “the information in this study necessarily shows that the decision tools don’t help”?

    We outlined a possibility. You have outlined another one. All the authors said is that is that there was no overall change in behavior.

  4. There’s an underlying assumption that improved knowledge and greater awareness should lead to less testing, and therefore to less unnecessary treatment. I think the second part is true but not the first.
    Most of us know it’s a very simple, inexpensive test, and I, for one, see no reason why it can’t be performed as a routine “vital sign” in low-risk prostate cancer patients who already have a cancer diagnosis, no matter how insignificant.the cancer is presumed to be. We’re not talking MRIs or bone scans here — we’re talking about a $10 item, a test that can be administered and resulted by a nurse, or a medtech, in any physician’s office anywhere, and one that may well help a patient achieve the peace of mind necessary to keep a dread illness at bay.

  5. Dear Rob:

    I don’t think that anyone is suggesting that one shouldn’t be giving regular PSA tests to a man who has been diagnosed with prostate cancer. The frequency of those tests would presumably depend on the stability of the patient’s PSA level over time. Some men would be getting them every 3 months (especially in the early stages of an active surveillance protocol). Other men with a diagnosis of low-risk prostate cancer and a PSA that was extremely stable over time might only need a PSA test once a year.

    I think the much greater concern has to do with the frequency of follow-up biopsies in men on AS protocols. It is clear that frequent biopsies come with risk for infections and other complications. We want to be doing biopsies that are necessary to minimize risk of not spotting clinically significant progression as opposed to biopsies for the sake of following a protocol that says every man should have a follow-up biopsy every year (regardless of whether he really needs one or not).

  6. Thanks Jim.

    Clearly, I’ve got to think about finding a new urologist. After six biopsies and a PSA ranging from 7 to 17 to near castrate level (on Lupron) and now back to 7 over a 5-year period, my cancer has been pronounced “insignificant” and I’ve been put on bi-annual PSA surveillance. I would much prefer quarterly checks, but my physician doesn’t agree, for reasons that escape me. The last biopsy (12-core), done a few months ago, showed only inflammation in all cores. I’d really like to know the cause, but again, there’s little interest on the other side of the desk. I blame most of this on the current trend toward doing away with unnecessary testing/treatment. It has had a spillover effect on us “gray area” patients who should have closer monitoring, which is too bad for all concerned. Thanks for your abiding interest in us, and your excellent site!

  7. I can’t believe the utter nonsense being babbled about the PSA test. It simply is what it is: a relatively cheap, utterly safe measure of prostate activity. A series of readings over time provide a trend of prostate activity, no more no less. Single readings are of very little value; however, a long-term trend can be life-saving or comforting. It’s as useful as monitoring weight, blood-pressure or cholesterol, etc.

    Men should start with a test at 40, 45, and 50; and based on those schedule future tests perhaps annually if warranted. This would catch early-onset aggressive cancers or provide a solid base-line for future monitoring. Costs for PSA testing are “beer-money” cheap.

    PSA testing may not be “cost-effective for the industry as a whole” but it sure as hell is “cost-effective” for the guy with early-onset prostate cancer, infection or BPH. It is also a cost-effective, life-long tool for those with normal PSAs, just knowing that their status is still steady, rising, and relatively “normal.”

    Certainly, PSA tests can be abused and can be misinterpreted. (1) Individual PSAs may be high due to infection, BPH, “having sex,” “riding a bicycle,” or sample/lab error. That’s why a multi-test trend across years is valuable. (2) The only “risk” from a PSA test is infection at the blood-draw site. (3) PSA tests do not cause unnecessary surgery; sloppy, bias or mis-interpretation does. So what else is new?

    Ya gotta use some common sense — which is utterly lacking in this ridiculous debate. Duhhhhhh.

    “Over 40, know yer PSA” … Plot and monitor its trend from 40-80.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: