Skip to Main Content

COVID-19

What people with cancer should know: https://www.cancer.gov/coronavirus

Get the latest public health information from CDC: https://www.cdc.gov/coronavirus/

Get the latest research information from NIH: https://covid19.nih.gov/

Principal Investigator
Name
Joshua Roth
Degrees
PhD, MHA
Institution
Fred Hutchinson Cancer Research Center
Position Title
Assistant Member
Email
About this CDAS Project
Study
NLST (Learn more about this study)
Project ID
NLST-142
Initial CDAS Request Approval
Jul 7, 2015
Title
Assessing the value of future research in low-dose computed tomography lung cancer screening in the United States
Summary
Value of information (VoI) analysis is an emerging comparative effectiveness research method that uses decision modeling and Bayesian statistical approaches to project the return on investment for future research studies. Put simply, VoI calculation involves estimation of how often a non-optimal clinical strategy is recommended with current evidence and after additional evidence is generated (i.e. how uncertain the decision is), the clinical/economic consequences of those decisions, and how many patients are impacted. The result of VoI analysis is the expected monetary value of a given study design to inform decisions for a specific affected population (e.g. Americans, Age 55-74, with 30+ pack years of smoking). This type of information can be valuable in many contexts, not the least of which is to inform the decisions of research funders with scarce resources seeking to priority-rank portfolios of candidate studies.

Our research group is among the few in the U.S. with expertise in VoI methods, and we recently completed a project applying VoI analysis to prioritize cancer genomics trials for the SWOG clinical trials cooperative group. Herein, we propose to apply VoI methods to assess the value of future studies related to low-dose computed tomography (LDCT) lung cancer screening. Specifically, we propose to augment an existing lung cancer screening Markov state-transition model with inputs from the National Lung Screening Trial (NLST) database, and calculate the expected value of a variety of randomized trial, observational, and patient preference elicitation study designs.

Partnership with the NLST is important because VoI is largely based on the degree of parameter uncertainty, and so access to patient-level outcomes that allow calculation of point estimates and 95% confidence intervals is critical to conducting a rigorous analysis. Furthermore, we plan to stratify analyses to assess the value of studies focused on NLST sub-groups defined by age, gender, and smoking status, so patient-level information will be necessary to facilitate these custom calculations.

Broad categories of research questions/designs we will evaluate include:

-Assessing the comparative effectiveness of new screening protocols/technologies (e.g. Lung-RADs)
-Evaluating the potential harms of cumulative radiation exposure from long-term annual LDCT screening
-Elucidating the frequency, types, and downstream impacts of LDCT incidental findings
-Preference elicitation studies to evaluate the health-related quality of life impacts of screening

This project is important because health care systems are making substantial investments in building lung cancer screening processes, personnel, and infrastructure. Additional research will be needed to ensure that these investments yield optimal returns in the forms of earlier cancer detection, decreased mortality, and increased health-related quality of life. The findings of the proposed study will provide systematic and quantitative evidence to inform future investments in this translational process.
Aims

Specific Aim 1: To expand an existing low-dose computed tomography lung cancer screening Markov state-transition model to calculate value of information using data from the National Lung Screening Trial.

Specific Aim 2: Using the expanded lung cancer screening model, assess the expected value of information for a range of plausible future research studies and designs related to low-dose computed tomography lung cancer screening.

Collaborators

Lotte Steuten, PhD, MSc, Fred Hutchinson Cancer Research Center
Scott Ramsey, MD, PhD, Fred Hutchinson Cancer Research Center
Sander Brinkhof, MSc, Fred Hutchinson Cancer Research Center
Bernardo Goulart, MD, MS, Fred Hutchinson Cancer Research Center
Richard Kim, MD, PhD, Fred Hutchinson Cancer Research Center

Related Publications