Prostate Cancer Mortality in Men not Attending a Population-based Screening Program: The 'Good', the 'Bad', and the 'Ugly' After 20-yr Follow-up in the European Randomised study of Screening for Prostate Cancer.

BACKGROUND AND

OBJECTIVE

Nonattendance can substantially decrease the effectiveness of existing and potential future population-based screening programmes for (prostate) cancer. We evaluated the long-term association between nonattendance and prostate cancer-specific mortality (PCSM) in the European Randomised Study of Screening for Prostate Cancer (ERSPC), with up to 20 yr of follow-up.

METHODS

This secondary analysis included 161 380 men aged 55-69 yr randomised to receive invitations for Prostate-specific antigen-based (PSA)-based screening (n = 72 460, screening arm [SA]) or to the control arm (CA) (n = 88,920) across seven ERSPC centres. Within the SA, men were classified as screening never-attenders (no attendance at any screening round) or screening ever-attenders (attended at least one screening round). The primary endpoint was PCSM. Cumulative incidence of PCSM was calculated using competing risk analysis, accounting for death from other causes as a competing event. Poisson regression was used to estimate rate ratios (RR) of cumulative PCSM between three groups: screening never-attenders, ever-attenders, and CA.

KEY FINDINGS AND LIMITATIONS: Of the 72 460 men in the SA, 12,401 (17%) were defined as never-attenders. At 20 yr, the cumulative PCSM was 1.4% (95% confidence interval [CI], 1.2-1.6) among never-attenders, 1.2% (95% CI, 1.2-1.3) in the CA, and 1.0% (95% CI, 0.9-1.0) in ever-attenders. For PCSM, we observed an RR of 1.39 (95% CI, 1.2-1.6; p < 0.001) for never-attenders compared to the CA, while for ever-attenders compared to the CA, we observed an RR of 0.77 (95% CI, 0.69-0.85; p < 0.001). Data on determinants of attendance are lacking.

CONCLUSIONS AND CLINICAL IMPLICATIONS: Men randomised to the SA but who never attended had higher observed PCSM than men in the CA who were not offered screening. Conversely, men in the SA who attended screening had lower observed PCSM than men in the CA, with a larger observed difference than that reported in prior intention-to-screen analyses. Trial registration: ISRCTN49127736.