Objective: The aim of the study was to establish the most cost-effective strategy for using integrated 18F-fluorodeoxyglucose PET and computed tomography (CT) for mediastinal lymph node staging (N-staging) of preoperative non-small-cell lung cancer (NSCLC).
Patients and methods: Four N-staging decision models for operative NSCLC, model A (CT only), model B (PET/CT for negative CT), model C (CT and PET/CT for all), and model D (PET/CT only), were constructed according to the Chinese edition of NCCN Clinical Practice Guidelines in Oncology Non-Small-Cell Lung Cancer and China NSCLC N-staging practices. Data including the parameters for decision models, life expectancy, and expenditures were retrieved from the literature, from websites of PET Reservation Centers, and on consultation with experts. On the basis of model A, incremental cost-effectiveness ratios were calculated for the other three models. Robustness of the result was evaluated by univariate sensitivity analysis.
Results: Life expectancy (years) was 2.60 for model A, 2.57 for model B, 2.67 for model C, and 2.66 for model D. The probabilities for improper therapy due to wrong staging were 36.78, 13.00, 8.91, and 12.95%, respectively. Surgical mortality rates were 1.87, 1.01, 1.13, and 1.24%, respectively. Model C, for which the incremental cost-effectiveness ratio (37 960 CNY/year) was lower than that of model D (65 175 CNY/year) and 2012 GDP per capita in China (38 459 CNY), was cost-effective and optimal, and this result was robust.
Conclusion: Model C was the most cost-effective strategy for NSCLC N-staging in the China mainland. Introduction of model C into Chinese N-staging protocols for NSCLC would be helpful for treatment selection, for reducing surgical mortality rates, and for extending life expectancy.