Objectives: Clinical guideline recommendations are generally informed by population- based evidence. However, interventions that are (cost-)effective on average may not be (cost-)effective for many (even for most) patients meeting trial inclusion criteria. This study aims to investigate the value of risk-stratified recommendations for lung cancer screening among current or former smokers between the ages of 55 and 74 years compared to a screen-all policy. Methods: Using data from the National Lung Cancer Screening Trial (NLST), we calculated the costs and QALYs for low-dose computed tomography (CT) versus chest radiography (X-ray) from empirically observed health states and 6 years life expectancy. Based on Kovalchik's risk of lung cancer death prediction model, we stratified 53,454 NLST trial patients into quintiles. The expected value of individualized care (EVIC) was calculated to quantify the value of using stratified information over population-based information. Results: The incremental cost-effectiveness ratio (ICER) of CT versus X-ray was $31,942 per QALY, for the “average” trial patient, indicating that CT would be the preferred option at a cost-effectiveness threshold of $50,000 per QALY. However, when stratified into quintiles, CT is dominated for the lowest risk quintile (i.e., X-ray is the preferred option for quintile 1) and CT is preferred for higher risk groups (quintiles 2 to 5). The EVIC was calculated at around $180 per person for cost-effectiveness thresholds of $50,000 per QALY and higher. Conclusions: Tailoring screening strategies to avoid CT scan in the lowest risk quintile of patients appears to be a superior strategy compared to population-wide CT scan screening, although results were sensitive to the cost-effectiveness threshold and the level of granularity of the analysis. This study shows the value of considering the risk-based heterogeneity of cost-effectiveness in clinical guideline recommendations and policy decisions.
Soeteman, D. I., Cohen, J. T., Neumann, P. J., Wong, J. B., & Kent, D. M. (2014). The Value of Risk-Stratified Information in the National Lung Cancer Screening Trial. Value in Health, 17(7), A324–A325. https://doi.org/10.1016/j.jval.2014.08.576