Impact of PSA-based Screening on Mortality Among Men Aged 75-79: Target Trial Emulation
Impact of PSA-based Screening on Mortality Among Men Aged 75-79: Target Trial Emulation
Opportunistic prostate-specific antigen (PSA) testing is widespread among men aged 75-79. Current American and European guidelines recommend PSA-based screening through shared decision-making if men are expected to live at least 10 more years-a threshold previously tied to ages 70-74. With rising life expectancy, Swedish men aged 75-79 now exceed this benchmark, averaging over 10 years.
Since no randomized controlled trials have examined screening in this age group, this study will use data from the Stockholm Prostate Cancer Diagnostics Register (2007-2023) to emulate a target trial. Men without prior prostate cancer will be assigned to screened or unscreened groups based on PSA test records and followed until death or December 2023. Analyses will use survival models, cumulative incidence functions, and sensitivity checks (e.g., varying follow-up duration, calendar year vs. age enrollment, and comparisons with men aged 65-69).
The goal is to determine whether PSA screening at ages 75-79 improves overall and prostate cancer-specific survival, and whether guidelines should expand to include this group while balancing the risk of overdiagnosis. Results will inform both clinical practice and the design of future randomized trials.
Inclusion Criteria:
Men aged 75-79 between 2007-2023, with no prostate cancer diagnosis at the time of inclusion (based on each trial emulation) will be considered eligible for the trials. To ensure proper control for potential confounding, and to be able to evaluate the eligibility criteria, only men with complete information on their screening status, outcomes, and confounders in the registers will be included.
Exclusion Criteria:
Men younger than 75 or older than 79 between 2007-2023. Have prostate cancer diagnosis at the time of inclusion (based on each trial emulation) Missing data on the exposure, outcomes, and confounders.