Developing evidence of the risks and benefits of PSA screening in black men to inform national guidelines and personalized decision-making

The National Cancer Institute awarded $460,134 to Veterans Medical Research Foundation of San Diego on July 20, 2026, under the Cancer Detection and Diagnosis Research program (CFDA 93.394) to conduct secondary analysis of Veterans Health Administration data examining prostate-specific antigen screening benefits and harms in Black men.

The research addresses a critical evidence gap: Black men comprised only 4.4 percent of participants in landmark PSA screening trials despite facing 64 percent higher prostate cancer incidence and more than double the mortality compared to non-Hispanic White men. Major guideline organizations—the U.S. Preventive Services Task Force, National Comprehensive Cancer Network, and American Cancer Society—have explicitly called for evidence-based PSA screening recommendations tailored to Black men.

The recipient will analyze longitudinal Veterans Health Administration data spanning 2005 through 2024 from over 5 million veterans across 130 VHA facilities, including more than 900,000 Black men—the largest cohort of Black men with detailed PSA screening data ever assembled. The analysis employs a triangulated methodology combining individual-level marginal structural models addressing time-dependent confounding, facility-level analyses exploiting natural practice variation, and population-level microsimulation modeling. The work leverages VHA's Veterans Health Information Systems and Technology Architecture (VINCI) infrastructure, integrating the Corporate Data Warehouse across 60 data domains with additional datasets and advanced natural language processing to detect metastatic outcomes.

The project period extends from July 20, 2026, through June 30, 2031, with place of performance in San Diego, California.

Abstract

Background: prostate cancer affects american men of all backgrounds, but its burden falls unevenly. Black men face 64% higher incidence (188.7 VS. 114.9 per 100,000) and more than double the mortality (36.9 VS. 18.4 per 100,000) compared to non-hispanic white men. Despite this elevated risk, PSA screening guidelines lack evidence-based recommendations for black men because black men comprised only 4.4% of participants in landmark trials (e.g., plco). Major guideline organizations (uspstf, nccn, ACS) explicitly call for research to address this critical evidence GAP. Objective: conduct comprehensive secondary analysis of veterans health administration (VHA) data to determine PSA screening benefits and harms specifically in black men, addressing the largest evidence GAP in PSA screening for high-risk populations. Methods: analyze longitudinal data from 5+ million veterans of varied backgrounds (including 900,000+ black men) across 130 VHA facilities, 2005-2024; the largest cohort of black men with detailed PSA screening data ever assembled. Use triangulated methodology: (1) individual-level marginal structural models addressing time-dependent confounding, (2) facility-level analyses exploiting natural practice variation, and (3) population-level microsimulation modeling. Analysis leverages VHA's informatics and computing infrastructure (VINCI), integrating the corporate data warehouse (60+ data domains) with additional datasets, enabling comprehensive multi-level analysis impossible with traditional single-dataset approaches. Innovation: pioneers marginal structural model application to PSA screening research, leverages unprecedented VINCI dataset integration, and employs multi-level triangulation maximizing analytical rigor. Advanced natural language processing automatically detects metastatic outcomes (97.9% specificity) while facility practice variation provides natural experiments for causal inference. Enhanced microsimulation incorporates real-world VHA data rather than survey-based estimates. Impact: results will inform evidence-based race-specific PSA screening recommendations for black men, the group at the highest risk. Team includes national guideline panel members (nccn, ACS) positioned for rapid policy translation. The resulting framework improves screening for the highest-burden group, while advancing real-world methods applicable to all men. Expected outcomes: deliver race-specific estimates of PSA screening benefits and harms, enabling tailored recommendations that optimize the benefit-harm ratio for black men. This project delivers solution-oriented research using integrated real-world data, advancing NIH priorities in real-world data science and training the next generation of physician-scientists.