PSA Screening

Below is an excerpt from an email sent out a few days ago by the American College of Physicians. It tells doctors that they should discuss the implications of PSA screening before offering the test to their patients. Why they are so far behind the curve is hard to understand. Here’s what I said on this subject two years ago: “Patients should talk to their doctor about the risks and benefits of undergoing prostate cancer screening before being tested. If their doctor doesn’t bring it up and just includes it (PSA testing) as part of his clinical evaluation call him on it. Ask your doctor what tests he’s going to order before they’re done. If he won’t spend the time talking to you about the issue, get another doctor.” Put “Prostate Cancer” into the search box on the right and you’ll get all the information needed to put you up to speed on screening for this disease.

That the ACP needs to alert physicians about something so basic and elementary and that the Archives of Internal Medicine has to publish two editorials telling doctors to do something that they should have figured out years ago on their own suggests that the best minds are no longer going into medicine.

The ACP email follows in italics.

PSA screens need more discussion

Physicians need to involve men more in the decision to undergo prostate-specific antigen (PSA) tests, concluded two studies and two editorials in Archives of Internal Medicine.

Clinicians strongly influence men’s decisions to undergo PSA screening, but the conversations about screening fail to qualify as shared decision making because patients received more information about the pros than the cons, had limited knowledge of their importance, and were not routinely asked for their preferences.

In one study, researchers conducted a telephone survey of 375 men who had either undergone or discussed with clinicians PSA testing in the previous two years. Researchers assessed the character of the discussion, the patient’s knowledge of prostate cancer and the importance of decision factors.

Almost 70% of patients discussed screening beforehand. Clinicians most often raised the idea of screening (64.6%), and 73.4% recommended PSA testing. Clinicians emphasized the pros of testing in 71.4% of discussions but addressed the cons in 32% of talks.

Researchers then asked the patients three questions to test their knowledge:

* “Of every 100 men, about how many do you think will die of prostate cancer?”
* “Of 100 men, about how many will be diagnosed as having prostate cancer at some time in their lives?”
* “For every 100 times a PSA test result suggests the need for further testing, about how many times does it turn out to be cancer?”

Although 58% of patients reported they felt well-informed about PSA testing, 47.8% failed to correctly answer any of the three questions, and only 7.2% of respondents could correctly answer more than one.

Only 54.8% of subjects reported being asked for their screening preferences. The clinicians’ recommendations were the only discussion characteristic associated with testing (odds ratio, 2.67; 95% CI, 1.08-6.58). “Indeed, few subjects sought second opinions,” researchers wrote.

An editorial concluded the research was an important step in prompting discussion about the tradeoffs between overdiagnosis and treatment complications versus the benefits of reduced risk of prostate cancer-related mortality.

A second editorial chided, “Today’s practice environment presents few incentives or support tools for those clinicians and patients who prefer a discussion rather than simply marking a checkbox for PSA on a laboratory requisition form.” It also noted some physicians may not ask to screen beforehand, but instead piggyback PSA tests onto other bloodwork.

A second study aimed to support individual decision making by creating a model of the likely benefits and harms.

Researchers in Australia created a model for men aged 40, 50, 60, and 70 years at low, moderate and high risk for prostate cancer. A Markov model compared patients with and without annual PSA screening using a 20% relative risk in prostate cancer mortality as a best-case scenario. The model estimated numbers of biopsies, prostate cancers and deaths from prostate cancer per 1,000 men over 10 years and cumulated to age 85 years.

Benefits and harms vary substantially with age and familial risk, the model found. As an example, among 1,000 60-year-old men with low risk screened annually, 115 would undergo biopsy triggered by an abnormal PSA screen. Among screened men, 53 would be diagnosed with prostate cancer over 10 years, compared with 23 men diagnosed as having prostate cancer among 1,000 unscreened men.

Among screened men, 3.5 would die of prostate cancer over 10 years compared with 4.4 deaths in unscreened men. For every 1,000 men screened from 40 to 69 years of age, there would be 27.9 prostate cancer deaths and 639.5 deaths overall by age 85 years compared with 29.9 prostate cancer deaths and 640.4 deaths overall in unscreened men. Higher-risk men have more prostate cancer deaths but also incurred more prostate cancers diagnosed and related harms.