Assessing the Generalizability of Findings from the NLST
Principal Investigator
Name
Louise Henderson
Degrees
PhD
Institution
University of North Carolina at Chapel Hill
Position Title
Professor of Radiology
Email
About this CDAS Project
Study
NLST
(Learn more about this study)
Project ID
NLST-876
Initial CDAS Request Approval
Jan 26, 2022
Title
Assessing the Generalizability of Findings from the NLST
Summary
Based on results of the National Lung Screening Trial (NLST), the United States Preventive Services Task Force currently recommends lung cancer screening with LDCT for adults aged 50-80 years with a smoking history of 20 pack-years, resulting in an estimated screening-eligible population of 14.5 million. Yet, current evidence shows that participants enrolled in NLST were younger, less racially diverse, more highly educated, and had fewer comorbidities than individuals undergoing lung cancer screening in the United States. In addition, NLST participants had ~95% adherence to three rounds of annual lung cancer screening, whereas adherence among individuals in clinical practice is much lower, ranging from 12-93%. These differences in the delivery of lung cancer screening in the NLST and clinical practice raise concerns about the external validity and utility of the NLST findings for informing broad clinical guidelines and public health recommendations.
To address these concerns, researchers have traditionally evaluated screening effects outside of trials by conducting observational studies comparing outcomes of screened and unscreened individuals. We propose a causal inference methodology that uses NLST data, Medicare data, and lung cancer screening registry data to improve the estimation of the effects of screening in clinical practice settings. The objective of this project is to quantify realistic population-level and subgroup effects of lung cancer screening with LDCT in clinical practice under different screening scenarios.
To address these concerns, researchers have traditionally evaluated screening effects outside of trials by conducting observational studies comparing outcomes of screened and unscreened individuals. We propose a causal inference methodology that uses NLST data, Medicare data, and lung cancer screening registry data to improve the estimation of the effects of screening in clinical practice settings. The objective of this project is to quantify realistic population-level and subgroup effects of lung cancer screening with LDCT in clinical practice under different screening scenarios.
Aims
1. Estimate the real-world effects of LDCT versus radiography on lung cancer mortality and false-positive findings by transporting the NLST to screened populations.
2. Evaluate the effects of annual screening duration with LDCT (i.e., one year, two years, three years) on lung cancer mortality using trial emulation methodologies.
Collaborators
Jennifer Lund, PhD (UNC Chapel Hill)