Excessive PSA testing in general practice

The time for actively recommending the screening of asymptomatic men has passed

There is little doubt that Australian doctors order too many prostate‐specific antigen (PSA) screening tests. The analysis by Franco and colleagues of electronic data for 142 000 Victorian general practice patients, published in this issue of the MJA, found that 46% of men aged 70–74 years had had at least two PSA tests during the preceding two years,1 despite Australian guidelines recommending against PSA screening of asymptomatic men in this age group.2 Indeed, the Royal Australian College of General Practitioners (RACGP) does not recommend PSA screening of most asymptomatic men of any age.3 Some testing might be justified (for example, screening of men at high risk, or prostate disease monitoring), but when Australian general practitioner registrars were asked to record specific reasons for ordering PSA tests, “asymptomatic screening” accounted for three‐quarters of requests.4

Another possibility is that some asymptomatic men may specifically request PSA tests, a reasonable option after a “specific discussion to address the benefits and harms (from overdiagnosis and overtreatment) of prostate cancer screening.”3 However, a survey of Australians undergoing PSA testing found that about 70% of 70–74‐year‐old men reported that their doctor either said it was part of a routine check‐up, or offered no discussion at all.5

The totality of the evidence is stark: we substantially over‐assess PSA in asymptomatic men. Why is it so?

Overtesting is a remediable problem; after all, our British colleagues order PSA tests at about one‐quarter the rate that we do.6 A qualitative study found that GPs (and patients) in the United Kingdom rarely raised the question of PSA testing, whereas many Australian GPs consider it an everyday part of consultations: in one study, all 69 Australian interviewees said they often received PSA requests from patients, and many themselves raised the possibility of testing unprompted, and typically ordered several PSA tests for asymptomatic men each day.6

The many drivers of overtesting were recently (June 2021) discussed in a popular private GP online forum (the GPs Down Under group on Facebook). Some participants mentioned patient expectations and pressures, while others had attended specialist urology updates where it was suggested that Australian GP guidelines were out of date. The participants generally expected that more active screening policies would be promoted by updated guidelines, whereas the opposite has been true for the past two decades, and no new evidence has been published that is likely to reverse this trend.

In 2011, the charity organisation Movember published the statement, “We believe that every man has the right to know if he has prostate cancer”, making the ambitious assumption that a man receiving a PSA result would then have that knowledge.7 The 2009 policy of the Urological Society of Australia and New Zealand (USANZ) similarly recommended both PSA testing and digital rectal examination as routine screening.8 USANZ has since joined other relevant medical organisations in endorsing the less aggressive Australian guidelines.2

Perhaps counterintuitively (yet inevitably, as with all screening programs), progress in the treatment of advanced prostate cancer will not increase the benefits of screening, but further reduce them, as fewer life years will be lost to the disease.

As someone who talks to many GPs about excessive screening, the most frequent response I hear is rarely quantified in research: “I just really want to help my patient.” This intrinsically altruistic motivation imagines conservative guideline makers (including myself) as being out of touch with practice, driven by cost savings, and impeding GPs caring for the special individuals sitting in front of them.

But consider a thousand such “special cases” seen by the well meaning GP over the course of a decade. The available evidence indicates that their life expectancy will not be improved compared with that of the patients of the GP next door who rarely orders PSA tests, and that at most one will avoid death from prostate cancer.9 On the other hand, 87 screened men will learn after their subsequent biopsies that their PSA results were false positives and 28 will have biopsy‐related complications; seven to ten more men will experience impotence or incontinence, and an extra 0.5 will have a heart attack.10 Sometimes it is safer not to be special.

Crunching the numbers for broad screening test recommendations (from cervical smears to COVID‐19 testing) requires population‐level big data rather than relying on personal experience or on the opinions of specialists primarily concerned with patients who have already been diagnosed with the disease.

Of the thousands of population‐wide screening tests ever proposed, only a handful have made it across all hurdles to be actively recommended. PSA testing had its moment, but it has passed, and there is little reason to expect its return.