• Department of Radiology, Brigham and Women's Hospital, Boston, Massachusetts.
• Institute for Technology Assessment, Massachusetts General Hospital, Boston, Massachusetts.
• Department of Radiology, University of Pennsylvania Perelman School of Medicine, Philadelphia, Pennsylvania.
• Institute for Technology Assessment, Massachusetts General Hospital, Boston, Massachusetts. Electronic address: joey@mgh-ita.org.

INTRODUCTION: To evaluate the cost-effectiveness of a number of follow-up guidelines and variants for subsolid pulmonary nodules.

METHODS: We used a simulation model informed by data from the literature and the National Lung Screening Trial to simulate patients with a ground glass nodule detected at baseline CT undergoing follow-up. Nodules were allowed to grow and develop solid components over time. We tested guidelines generated by varying follow-up recommendations for low-risk nodules, i.e. pure ground glass nodules or those stable over time. For each guideline, we computed average U.S. costs and quality-adjusted life-years (QALYs) gained per patient and identified the incremental cost-effectiveness ratios (ICERs) of those on the efficient frontier. In addition, we compared the costs and effects of the most recently released version of Lung-RADS, v1.1, with the previous version, v1.0. Finally, we performed sensitivity analyses of our results by varying several relevant parameters.

RESULTS: Relative to no follow-up, the follow-up guideline system that was cost-effective at a willingness-to-pay of $100,000/QALY and had the greatest QALY assigned low-risk nodules a 2-year follow-up interval and stopped follow-up after 2 years for ground glass nodules and after 5 years for part-solid nodules; this strategy yielded an ICER of $99,970. Lung-RADS v.1.1 was found to be less costly but no less effective than Lung-RADS v1.0. These findings were essentially stable under a range of sensitivity analyses.

CONCLUSION: Ceasing follow-up for low-risk subsolid nodules after 2-5 years of stability is more cost-effective than perpetual follow-up. Lung-RADS v1.1 was cheaper but similarly effective to v1.0.