Prostate screening cut deaths 13 percent over 23 years, and the trade-off keeps improving

In 1993, researchers in eight European countries began splitting men aged 55 to 69 at random into two groups. One group was offered repeated blood tests for prostate-specific antigen, or PSA, a protein the prostate makes that leaks into the blood in larger amounts when the gland is enlarged, inflamed or cancerous. The other group carried on with ordinary care. By the time the trial closed its books, 162,236 men had been followed for a median of 23 years, half for longer and half for less, and those offered screening were 13 percent less likely to have died of prostate cancer. The gap did not fade with time. It grew: at 16 years, 628 men had to be invited for screening to prevent one death; at 23 years, 456.

How we know

The figures come from the European Randomized Study of Screening for Prostate Cancer, known as ERSPC, whose 23-year analysis appeared in the New England Journal of Medicine on 29 October 2025. Because men were assigned to the two groups by chance, the design of a randomised controlled trial, the only systematic difference between them was the invitation to screening, so the difference in deaths can be laid at its door. Prostate cancer mortality in the screening group was 0.87 times that of the control group, a rate ratio whose 95 percent confidence interval ran from 0.80 to 0.95. The 13 percent is a relative reduction; in absolute terms it was 0.22 percentage points. The trial counted the cost as carefully as the gain. Men offered screening were diagnosed with prostate cancer 30 percent more often, a rate ratio of 1.30, and many of those extra cancers were slow-growing tumours that would never have caused symptoms: overdiagnosis, the detection of a disease that would not have harmed the person in their lifetime. That ratio of harm to benefit improved as the years passed. At 16 years, 18 men were diagnosed for every death averted. At 23 years, 12. In the Finnish arm, which supplied nearly half the men, 16 percent had a raised PSA at their first tests, and a biopsy, in which a sample of prostate tissue is removed and examined, confirmed cancer in 24 percent of those. Screening found more low-risk cancers and fewer advanced ones.

Why it matters

Prostate cancer is the most common cancer in men, and the ERSPC authors expect deaths from it to rise worldwide as populations grow and live longer. For three decades the case against screening every man has been that a PSA test finds too many harmless cancers for too few lives saved, and the 23-year figures do not dissolve that argument so much as rebalance it: fewer men invited per death prevented, fewer men diagnosed per death prevented, with the overdiagnosis still real. The protection is also not permanent. Anssi Auvinen, who led the Finnish arm at Tampere University, said the effect starts to wane once screening stops and largely disappears within nine years. The authors' answer is not blanket testing but risk-based screening, which concentrates testing on the men most likely to carry a dangerous cancer and uses MRI scans to sort harmless tumours from ones that need treating. Finland, which supplied nearly half the men in the trial, still has no national screening programme. Whether countries build one, and how carefully they aim it, is now the question these 23 years leave open.