Earlier syntheses of PSA screening trials were equivocal, largely because follow-up was too short for a slow-growing cancer. This systematic review searched to October 2025 and pooled 5 randomised trials analysing 721,607 participants aged 45 to 80 with 11 to 23 years of follow-up, using fixed-effect models and GRADE.
At longest follow-up, PSA-based screening reduced prostate cancer-specific mortality among those who engaged in screening, rated high certainty (P < .001). Secondary analyses showed the relative reduction growing with longer follow-up. The authors note that differing screening protocols and contamination of control arms — men in the control groups getting PSA tests anyway — would both bias toward the null, so the true benefit is probably larger than measured.
The authors are careful, and so should we be. High-certainty evidence of a mortality benefit is not an argument for population screening, and they say so explicitly. Screening also produces overdiagnosis, biopsy complications and treatment of cancers that would never have caused harm, and the balance of those against a mortality reduction is a value judgement each man makes for himself. What changes today is the honesty of the conversation: it is no longer accurate to tell a man that PSA screening has not been shown to save lives. It has. Whether it is worth it to him is a different question, and the one worth spending clinic time on.
- Stop telling patients that PSA screening has no proven mortality benefit — that is now out of date.
- Present overdiagnosis and treatment harms alongside the benefit; both are real and both are large.
- The benefit accrues over 11 years and more, so life expectancy is central to the decision.
- MRI-first pathways and risk-stratified protocols are the emerging way to keep benefit and reduce harm.
- This is evidence about individual decisions, not a mandate for a population programme.
The statistics, in plain English
Control-arm contamination means some unscreened men were screened anyway, which makes the two arms more similar and pushes the measured effect toward zero — so the real benefit is likely larger than the pooled estimate, not smaller. A fixed-effect model assumes the trials estimate one common effect, which is defensible with five large trials and arguable given their different protocols. High GRADE certainty refers to confidence in the effect estimate, not to whether the effect is worth the harms.
Read the rest in the app
You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.
Scan to keep reading on your phone. No account needed to start.
Tomorrow morning, before your first patient
One edition a day for urology, written by the desk, every claim tied to its paper. Six minutes.
Get the app — free