Skip to Main Content

COVID-19 is an emerging, rapidly evolving situation.

What people with cancer should know:

Get the latest public health information from CDC:

Get the latest research information from NIH:

Principal Investigator
Sonya Cressman
British Columbia Cancer Research Institute
Position Title
Health Economist
About this CDAS Project
NLST (Learn more about this study)
Project ID
Initial CDAS Request Approval
Apr 1, 2015
The Pan-Canadian Early Detection of Lung Cancer Study:Cost-effectiveness of risk-stratification in lung cancer screening selection and nodule management in Canada
The study proposes to evaluate the cost-effectiveness of screening for lung cancer in a pre-selected, high-risk cohort. The PanCan is a single-arm study with a focus on improving the cost-effectiveness of lung cancer screening by using a risk prediction algorithm prior to enrollment. The comparison will be made between the CT-screening data from the Pan-Canadian Early Detection of Lung Cancer Study and data from the chest-radiography arm of the NLST.

Specific Aim 1. Use data in the NLST radiography (CXR) arm to determine the expected lung cancer mortality rates for the subgroup of screening participants who were at a high-risk of getting lung cancer and who did not receive CT-screening.
Specific Aim 2. Estimate what the cost of treating unscreened lung cancer would be in Canada.
Specific Aim 3. Use long-term follow-up from the lung cancer treatment data in the NLST to help estimate the expected probability of relapse and death from lung cancer or other causes.

Datasets Requested:

A. Participant dataset

This NLST dataset will be used to classify the non-CT screened control arm into a sub-group of screening candidates that did not receive or participate in CT screening according to the Pan-Can study protocol, assuming that CXR screening is equivalent to not screening. Specific fields to be used include: Lung cancer risk factors (age, sex, ethnicity, education level, BMI, personal history of cancer, family history of lung cancer, history of COPD, smoking status, duration, intensity, quit time, and CXR or CT screening exam in the last three years) and linkage fields (participant number, screening arm and centre, time from randomization, time to diagnosis, death due to lung cancer or other causes and status at the last day of NLST follow-up).

B. Chest X-ray screening, Diagnostic Procedures, Medical Complications, and Treatment datasets

This NLST dataset will be used for Aim 2. We will link these datasets with the risk-selection criteria and use the diagnostic procedure, complications codes, and treatment category and code fields in these datasets to estimate the cost to treat unscreened lung cancer in Canada in a population that is eligible for screening based on the age and smoking history criteria used for recruitment in the NLST and a sub-group that is considered high-risk according to the risk-prediction tool (i.e. resource utilization in the CXR arm of the NLST will be used to inform the comparative non-screened arm in the cost-effectiveness analysis). The cost analysis will be based on a published costing methodology for the Canadian study (Cressman et al JTO, 2014). A list of the participants screened with CXR, the result of CXR-screening and recommended follow-up from CXR- screening are the specific fields that will be linked to the resource fields to estimate costs in the control cohort.

C. Lung Cancer Progression and Cause of Death

Specific fields such as the days to progression, the progression status, and cause of death in theses datasets will be used to estimate the lung cancer and all-cause mortality rates that could be expected for individuals who had lung cancer, both in the screened arm (CT) and the (CXR) arms of the analysis. The frequency of relapse and lung cancer mortality rates after relapse will be used to 1) project outcomes in the non-screened comparative arm of the analysis and 2) validate the medium-term outcomes observed in the pan-Can study for the CT-screened arm of the analysis, which will reduce uncertainty in the cost-effectiveness analysis.


Stephen Lam, MD, FRCPC, British Columbia Cancer Agency
Stuart Peacock DPhil, University of British Columbia
Martin Tammemagi PhD, Brock University