Skip to Main Content
An official website of the United States government

Evaluating the Impact of Differential LDCT Screening Uptake, Adherence, and Novel Treatment Access on Racial/Ethnic Disparities in Lung Cancer Outcomes: A Modeling Study

Principal Investigator

Name
Michael Leo

Degrees
PhD

Institution
Oregon Health & Science University

Position Title
Senior Staff Scientist

Email
leom@ohsu.edu

About this CDAS Project

Study
PLCO (Learn more about this study)

Project ID
PLCO-2072

Initial CDAS Request Approval
Sep 8, 2026

Title
Evaluating the Impact of Differential LDCT Screening Uptake, Adherence, and Novel Treatment Access on Racial/Ethnic Disparities in Lung Cancer Outcomes: A Modeling Study

Summary
Low-dose computed tomography (LDCT) screening and emerging novel therapeutics have transformed lung cancer care, yet racial and ethnic disparities in lung cancer incidence, late-stage diagnosis, and mortality persist across the United States. Differences in screening uptake, longitudinal protocol adherence, and timely access to targeted or novel treatments across racial/ethnic groups may perpetuate or widen these outcome gaps. Quantifying how implementation barriers drive these disparities is essential for guiding targeted health policy and resource allocation.

We will conduct a decision-analytic microsimulation study using the MCEDSim framework (specifically, the new package MCEDsimDisparities). The target population comprises adults aged 50-80 years eligible for lung cancer screening under American Cancer Society (ACS) guidelines (greater or equal 20 pack-year smoking history). Incidence and survival parameters will be derived from Surveillance, Epidemiology, and End Results (SEER) case-listing data, stratified by race/ethnicity and histological subtype (e.g., adenocarcinoma, squamous cell, small cell). To adapt population-level SEER incidence to a screening-eligible cohort, baseline incidence rates will be adjusted using risk-elevation factors from established trials. Specifically, we will compare lung cancer incidence in the control arm of the PLCO to incidence in the subset of the control arms individuals who are screen eligible. Survival models will be fitted separately by race and used to project mortality outcomes, potentially adjusting for novel treatment.

MCEDSim will be modified to include the following modules: initial screening uptake (receiving ≥ 1 screen), adherence to annual screening protocols, and access/efficacy parameters for novel therapeutics. Using a factorial simulation design, we will compare lifetime outcomes across multiple implementation scenarios: (1) Equal Implementation, reflecting universal uptake, perfect adherence, and equal novel treatment access across all racial/ethnic groups; (2) As-Observed Real-World Scenarios, using empirical rates of screening uptake, adherence, and novel treatment access stratified by race/ethnicity; and (3) Counterfactual Interventions, testing incremental parameter improvements (e.g., stepwise increases in adherence or treatment access) to evaluate the sensitivity and impact of each factor on projected disparities.

Aims

Aim 1: Characterize lung cancer incidence, and survival by race and ethnicity in the screening eligible population
• PLCO data used to identify inflation factor for incidence in the screening eligible relative to the overall population.
• Fit parametric survival models to SEER case listing data by race/ethnicity and subtype in the screeningeligible population.
• Derive stage- and subtype-specific novel treatment efficacy estimates using published aggregate results.

Aim 2: Extend the MCEDSim microsimulation platform to model disparities in lung cancer screening and treatment.
• Develop MCEDsimDisparities, a modified version of MCEDSim that incorporates race/ethnicityspecific model calibration, differential screening uptake and adherence, differential screening test sensitivity, and differential treatment effects.
• Calibrate the model separately by race/ethnicity and cancer subtype to adjusted incidence and survival targets, with internal validation against observed outcomes.

Aim 3: Simulate implementation scenarios to quantify and decompose disparities in lung cancer outcomes.
• Specify a factorial set of scenarios spanning Equal Implementation, As-Observed Real-World Scenarios, and Counterfactual Interventions.
• Simulate screening-eligible cohorts from ages 50-80 through end of life and compute outcomes including late-stage diagnoses, lung cancer-specific mortality, life-years, and interval versus screen-detected cancer rates.
• Calculate absolute and relative disparity measures across scenarios relative to the non-Hispanic White reference group.
• Decompose the relative contribution of screening uptake, diagnostic follow-up, and treatment access to projected disparities, including marginal impacts of incremental improvements, to identify high-impact targets for policy intervention.

Collaborators

Ishfaq Ahmad Oregon Health & Science University
Michael Leo Oregon Health & Science University
Shelley Tworoger Oregon Health & Science University
Jane Lange Oregon Health & Science University
Nora Pashayan University of Cambridge