Lung cancer screening for Veterans: measuring real-world benefit and harms
Funded Grant
Overview
Affiliation
View All
Overview
description
Background: Two randomized trials demonstrated that lung cancer screening (LCS) prevents lung cancer death: the National Lung Screening Trial (NLST) and the Nederlands–Leuvens Longkanker Screenings Onderzoek (NELSON) trial. LCS-eligible Veterans have three times as many comorbidities as trial participants. Risks/benefits of LCS outside of healthy trial participants are unclear. National guidelines including those of the VA consistently emphasize screening only those in good health yet disagree on how to identify them. My early research shows that, in current VA practice, Veterans in poor health are screened for lung cancer despite lack of proven benefit for them. This exposes Veterans to the harms of screening without likely benefit. Scientific Rationale: We lack an understanding of real-world benefits of LCS across the spectrum of health. Veterans, clinicians, and the VA need precision estimates of the risks/benefits of LCS, tailored to individual health. To address this critical knowledge gap, this project will create two cohorts to examine the effectiveness of LCS as currently used in the VA, and identify pre-screening health metrics that predict benefit from LCS. Specific Aims: Aim 1A. To examine the effectiveness of LCS as currently used in the VA in shifting lung cancer to an earlier stage at diagnosis. Aim 1B. To identify the threshold of health at which LCS is beneficial in terms of stage shift. Aim 2A. To examine the effectiveness of LCS as currently used in the VA in reducing lung cancer death. Aim 2B. To identify the threshold of health in which LCS is effective in preventing lung cancer death. Exploratory analyses will examine how the benefit of LCS varies by race/ethnicity and sex. Methods: In Aim 1A, I will create a national cohort of Veterans using MVP data and test whether screening favorably shifts the stage at lung cancer diagnosis. I will apply propensity methods, survival analysis, and novel target trial methods to determine the effectiveness of current use of LCS in the VA (i.e., without regard to baseline health) in reducing advanced lung cancer. In Aim 1B, I will examine whether NELSON eligibility criteria (good/excellent self-rated health and/or can climb stairs) identify Veterans who benefit from LCS in terms of stage shift. In Aim 2A, I will create a national cohort of Veterans using Corporate Data Warehouse (CDW) data. I will use the same methods as in Aim 1A to measure LCS’ effectiveness to reduce lung cancer death in currently VA practice. In Aim 2B, I will examine whether LCS reduces lung cancer death among those with Care Assessment Needs (CAN) score≤60 (i.e., in similar overall health as NLST trial participants) or with life expectancy>5 years (VA’s recommended threshold of health before LCS, based on expert opinion). Innovation: This proposal is conceptually innovative: I will (1) determine the real-world effectiveness of LCS, (2) test whether health metrics predict benefit of LCS, and (3) utilize health metrics relevant to Veterans and LCS. This proposal is methodologically innovative: I will apply novel target trial methods to estimate LCS’ effectiveness in those who would/would not have qualified for trial inclusion due to health and use MVP data. Significance/Impact: The current project directly responds to CSR&D priorities to obtain an in-depth understanding of disease processes that are highly relevant to Veterans and will generate clinically relevant deliverables to test in subsequent independent work. Lung cancer kills more Veterans than any other cancer. 1.8 million Veterans are eligible for LCS, but screening is reaching Veterans who are least likely to benefit due to poor health. This project’s central research objective is to generate precision estimates of benefits of LCS stratified by health. In the future, these estimates will be applied to design decision support tools that promote screening among those likely to benefit. My training objectives are to become experienced with the use of VA data for research, gain a rich skillset of rigorous quantitative methods, develop LCS content expertise, and form partnerships with VA leaders in the field. The skills, experience, and relationships I gain from this work will support my goal of being an independent VA investigator focused on preventive interventions for Veterans.
