Prostate-specific antigen (PSA) test for prostate cancer screening

Prostate cancer screening remains controversial. A 2013 update of this Cochrane review concluded that prostate cancer screening does not significantly reduce disease-specific or overall mortality. Since then, results from two new, large trials have become available, alongside extended follow-up data from previously included trials. Thus, an update was warranted.

To assess the effects of prostate cancer screening compared to no screening in men with no prior diagnosis of prostate cancer.

We searched the Cochrane Library, MEDLINE, Embase, five other databases, and two trial registers from their inception to November 2025. We did not apply language restrictions.

We included randomised controlled trials (RCTs) of screening versus no screening for prostate cancer.

Critical outcomes were prostate cancer-specific mortality, overall mortality, and adverse events. Important outcomes included quality of life, prostate cancer diagnosis, localised (stage I and stage II) prostate cancer diagnosis, advanced (stage III and IV) prostate cancer diagnosis, and metastatic prostate cancer diagnosis.

We used the Cochrane risk of bias tool to assess the risk of bias in the studies.

Two review authors independently selected trials, extracted data, and assessed risk of bias. We used a random-effects meta-analysis model to combine data. We report dichotomous data as incidence rate ratios when outcomes were reported as person-years; otherwise, we report risk ratios (RRs) with 95% confidence intervals (CIs). We assessed the certainty of the evidence using GRADE and applied thresholds for minimal clinically important differences (MCIDs).

We included six RCTs with a total of 789,086 randomised participants. The trials were conducted in Europe and North America, and involved men aged 45 to 80 years. Five trials compared prostate-specific antigen (PSA) testing, with or without digital rectal examination (DRE), to usual care without systematic PSA screening. The three largest studies, accounting for 85% of randomised participants, were the CAP trial (UK), PLCO trial (USA), and ERSPC (Europe). One trial with preliminary results compared PSA-based screening in combination with kallikrein and magnetic resonance imaging (MRI) with no screening. Follow-up ranged from 3.2 to 23 years.

Screening based on PSA (alone) versus no screening These results are based on the certainty of evidence for each outcome and are therefore mainly derived from a sensitivity analysis by risk of bias, using the ERSPC trial. Screening likely results in a reduction in prostate cancer-specific mortality (sensitivity analysis from ERSPC: Rate Ratio 0.87, 95% CI 0.80 to 0.95; 1 study, 162,236 participants; moderate-certainty evidence; MCID: 1 per 1000 participants). Assuming a baseline risk of 16 prostate-cancer-related deaths per 1000, this corresponds to 2 fewer (95% CI 3 fewer to 1 fewer) per 1000. Based on the analysis that included all eligible trials, screening may reduce overall mortality, although the confidence interval includes the possibility of little to no effect (Rate Ratio 0.99, 95% CI 0.97 to 1.00; I² = 22%; 4 studies, 675,121 participants; low-certainty evidence; MCID: 1 per 1000 participants). Assuming a baseline risk of 491 deaths from any cause per 1000, this corresponds to 5 fewer (95% CI 15 fewer to 0 fewer) per 1000. Screening may result in little to no difference in adverse events (RR 1.32, 95% CI 0.48 to 3.65; 1 study, 408,721 participants; low-certainty evidence; MCID: 1 per 1000 participants), as assessed by deaths related to the prostate biopsy or prostate cancer treatment. Assuming a baseline risk of 3 deaths from any cause per 100,000, this corresponds to 1 more (95% CI 2 fewer to 8 more) per 100,000. Screening may result in little to no difference in quality of life (SF-6D range 0 to 1; higher values indicate higher quality of life; MCID 0.1; screening 0.777 versus control 0.779; P = 0.88; 1 study, 969 participants; low-certainty evidence). Screening likely increases prostate cancer diagnoses (sensitivity analysis data: Rate Ratio 1.30, 95% CI 1.27 to 1.34; 1 study, 162,241 participants; moderate-certainty evidence; MCID of 10 per 1000 participants), including localised (stage I and II) disease (sensitivity analysis data: RR 1.53, 95% CI 1.48 to 1.59; 1 study, 162,236 participants; moderate-certainty evidence; MCID: 10 per 1000 participants). Screening likely has little to no effect on advanced (stage III and IV) prostate cancer diagnosis (sensitivity analysis data: RR 0.90, 95% CI 0.85 to 0.95; 1 study, 162,236 participants; moderate-certainty evidence; MCID: 10 per 1000 participants). It may reduce metastatic prostate cancer diagnosis (sensitivity analysis data: RR 0.65, 95% CI 0.59 to 0.71; 1 study, 162,236 participants; low-certainty evidence; MCID: 5 per 1000 participants). Screening based on PSA, kallikrein panel, and MRI versus control Prostate cancer-specific mortality, overall mortality, adverse events, quality of life, and metastatic prostate cancer diagnosis were not reported in early and preliminary results of the included study. Screening based on PSA, kallikrein panel, and MRI likely has little to no effect on prostate cancer diagnosis compared to control (RR 1.85, 95% CI 1.56 to 2.19; 1 study, 60,745 participants; moderate-certainty evidence; MCID: 10 per 1000 participants). It likely also has little to no effect on localised prostate cancer diagnoses (RR 1.88, 95% CI 1.55 to 2.27; 1 study, 60,745 participants; moderate-certainty evidence; MCID: 10 per 1000 participants) and advanced prostate cancer diagnoses (RR 1.78, 95% CI 1.19 to 2.66; 1 study, 60,745 participants; moderate-certainty evidence; MCID: 10 per 1000 participants).

Screening likely reduces prostate cancer-specific mortality and may reduce overall mortality. It may have little to no effect on adverse events (as measured by intervention-related mortality). Study limitations, inconsistency, and imprecision reduced our confidence in the effect estimates. Interpretation of these findings is highly sensitive to the choice of MCID. We found insufficient evidence on the potential harms of screening, such as biopsy- and treatment-related complications. Emerging alternatives, such as screening with a kallikrein panel and MRI, may have little to no effect on diagnoses of prostate cancer, but the results on mortality are not yet known.

This Cochrane review had no dedicated funding.

Protocol (2004) DOI: 10.1002/14651858.CD004720 Original review (2006) DOI: 10.1002/14651858.CD004720.pub2 Review update (2013) DOI: 10.1002/14651858.CD004720.pub3.