PSA Screening: Why ERSPC and PLCO Point in Different Directions

The short answer

Two large randomized trials asked whether prostate-specific antigen (PSA) screening lowers the chance of dying from prostate cancer, and they returned opposite headlines. The European Randomized Study of Screening for Prostate Cancer (ERSPC) reported roughly a one-fifth relative reduction in prostate-cancer death among men invited to screen, while the U.S. Prostate, Lung, Colorectal, and Ovarian (PLCO) trial found no significant difference. The most-cited reason for the split is not that PSA behaves differently on two continents; it is that most men in PLCO's control group were screened anyway. The 2018 U.S. Preventive Services Task Force (USPSTF) statement read both trials together and landed on a grade C recommendation: for men aged 55 to 69, whether to screen is an individual decision.

What each trial actually measured

ERSPC enrolled more than 160,000 men across several European countries and randomized them to a screening invitation or to usual care. Screening intervals varied by site, most commonly a PSA test every four years. In its published mortality analyses, the trial found that men in the screening group were less likely to die of prostate cancer, with the relative reduction settling near 20 percent over long-term follow-up.

PLCO enrolled roughly 76,000 American men and assigned them to annual PSA testing for six years or to usual care. After nearly 15 years of follow-up, the two groups showed no meaningful difference in prostate-cancer death. On its face, that is a direct contradiction of ERSPC.

The direction of each result matters less than why they diverged, because the guideline that governs practice today rests on reconciling them rather than picking a winner.

Contamination: the control group that wasn't a control group

A randomized trial works by comparing people who get an intervention against people who do not. PLCO's problem is that its "do not screen" arm largely did get screened. Analyses of the trial found that a majority of control-group men received PSA testing outside the study, and estimates of that contamination ran past half of the control arm. When both groups are screened at similar rates, a trial loses its ability to detect a difference between screening and no screening, even if a real difference exists. It is closer to comparing organized screening against opportunistic screening than against none at all.

ERSPC had far less of this problem. The gap in actual PSA testing between its invited and control groups was much wider, so the comparison it drew was closer to the question the trials were designed to ask. A 2017 analysis in Annals of Internal Medicine used modeling to align the two trials on a common footing and concluded that, once differences in screening intensity and lead time were accounted for, the trials were not as contradictory as their headlines suggested; both were compatible with a modest screening benefit. That reconciliation is the quiet center of this whole debate.

Interval and intensity are not footnotes

The trials also differed in how aggressively they screened. PLCO tested annually; ERSPC mostly tested every four years yet still detected a mortality signal. That combination cuts against a simple more-testing-saves-more-lives reading. It suggests that the frequency of testing, the biopsy thresholds, and how detected cancers were treated all shape whether screening translates into fewer deaths. Two screening programs with the same test can produce different benefit-to-harm balances depending on these design choices.

Framing the benefit honestly

Relative reductions sound larger than they feel in absolute terms. In ERSPC's long-term follow-up, the USPSTF summarized the benefit as preventing on the order of one to two prostate-cancer deaths and a few cases of metastatic disease for every 1,000 men screened over roughly 13 years. Against that sits a substantial burden of harm: false-positive PSA results, biopsies, and the downstream consequences of diagnosing and treating cancers that would never have caused symptoms. This overdiagnosis, and the incontinence and erectile dysfunction that can follow treatment, is why a real but small mortality benefit does not automatically settle the question. Neither trial demonstrated a reduction in all-cause mortality, which is part of why the net benefit is judged small rather than large.

How the USPSTF translated the conflict

The 2018 recommendation split by age. For men 55 to 69, the Task Force assigned grade C, meaning the net benefit is small and the choice should rest on an individual's own weighing of a possible mortality reduction against the risks of false positives, overdiagnosis, and treatment side effects, discussed with a clinician. For men 70 and older, it assigned grade D and recommended against PSA-based screening, judging that the harms outweigh the benefit at that age. A grade C is not a shrug; it is a statement that the evidence genuinely does not point every man in the same direction, so the decision belongs to the person, not the guideline.

This article is educational and is not medical advice. Anyone weighing PSA screening should discuss their own risk factors, values, and circumstances with a clinician who knows their history.

The through-line

ERSPC and PLCO look like a contradiction and function more like a lesson in how trial design shapes conclusions. When a control group is quietly screened, a null result cannot be read as proof that screening does nothing. When the two trials are placed on comparable terms, the evidence points to a small, real, and uncertain benefit, which is exactly the shape of a shared decision.