A Risk-Stratified Decision Support Framework for Prostate Cancer Screening Using the PLCO Trial Dataset

Principal Investigator

Name
Priya Dubey

Degrees
Ph.D.

Institution
Tokyo International University

Position Title
ASSISTANT PROFESSOR

Email
pdubey@tiu.ac.jp

About this CDAS Project

Study
PLCO (Learn more about this study)

Project ID
PLCO-2014

Initial CDAS Request Approval
Feb 23, 2026

Title
A Risk-Stratified Decision Support Framework for Prostate Cancer Screening Using the PLCO Trial Dataset

Summary
Prostate cancer screening remains a clinically debated topic due to the risk of overdiagnosis and overtreatment, particularly in men with low-risk disease. While prostate-specific antigen (PSA) testing has improved early detection, population-level screening strategies lack individualized risk stratification that integrates demographic, clinical, and screening history variables.

The PLCO trial provides a unique longitudinal dataset containing baseline demographics, PSA measurements, digital rectal examination (DRE) data, screening compliance, and prostate cancer outcomes, making it suitable for developing risk-based analytical frameworks rather than single-variable screening rules.

This project aims to explore how multi-factor information from PLCO can be used to identify patterns associated with elevated prostate cancer risk, supporting more informed screening strategies while remaining strictly observational and non-interventional.

Aims

• To describe associations between demographic and screening-related variables and prostate cancer outcomes in the PLCO cohort. This aim will examine the distribution of age, race/ethnicity, family history of prostate cancer, PSA measurements, and digital rectal examination (DRE) findings among male participants in the PLCO trial. Statistical analyses will be used to assess population-level associations between these variables and prostate cancer diagnosis and timing, accounting for screening exposure and follow-up duration.

• To assess the joint influence of multiple screening variables on observed prostate cancer outcomes. This aim will evaluate how combinations of PSA values, demographic characteristics, and screening history relate to prostate cancer outcomes within the PLCO dataset. Analyses will focus on understanding the relative contribution of individual variables and their combined patterns, without attempting individual-level classification or prediction.

• To explore a structured analytical framework for grouping participants based on observed characteristics. This aim will involve developing and examining an exploratory grouping approach that organizes participants into categories based on shared demographic and screening profiles. The purpose is to support population-level interpretation of screening data and to generate hypotheses for future research. All results will be reported in aggregated form, with no clinical interpretation or individual-level inference.

Collaborators

Priya Dubey TOKYO INTERNATIONAL UNIVERSITY