A comparison of surgical intervention and stereotactic body radiation therapy for stage I lung cancer in high-risk patients: A decision analysis

Abstract

We sought to compare the relative cost-effectiveness of surgical intervention and stereotactic body radiation therapy in high risk patients with clinical stage I lung cancer (non-small cell lung cancer). We compared patients chosen for surgical intervention or SBRT for clinical stage I non-small cell lung cancer. Propensity score matching was used to adjust estimated treatment hazard ratios for the confounding effects of age, comorbidity index, and clinical stage. We assumed that Medicare-allowable charges were $15,034 for surgical intervention and $13,964 for stereotactic body radiation therapy. The incremental cost-effectiveness ratio was estimated as the cost per life year gained over the patient's remaining lifetime by using a decision model. Fifty-seven patients in each arm were selected by means of propensity score matching. Median survival with surgical intervention was 4.1 years, and 4-year survival was 51.4%. With stereotactic body radiation therapy, median survival was 2.9 years, and 4-year survival was 30.1%. Cause-specific survival was identical between the 2 groups, and the difference in overall survival was not statistically significant. For decision modeling, stereotactic body radiation therapy was estimated to have a mean expected survival of 2.94 years at a cost of $14,153 and mean expected survival with surgical intervention was 3.39 years at a cost of $17,629, for an incremental cost-effectiveness ratio of $7753. In our analysis stereotactic body radiation therapy appears to be less costly than surgical intervention in high-risk patients with early stage non-small cell lung cancer. However, surgical intervention appears to meet the standards for cost-effectiveness because of a longer expected overall survival. Should this advantage not be confirmed in other studies, the cost-effectiveness decision would be likely to change. Prospective randomized studies are necessary to strengthen confidence in these results.