Abstract
OBJECTIVE: To investigate medical expenses and factors influencing surgical choices among lung cancer patients in a tertiary hospital in Xiamen, China, and to identify key cost differences between surgical and non-surgical approaches. METHODS: In this retrospective cross-sectional study, we analyzed 3,806 lung cancer patients treated in 2023. Data analysis was performed using SPSS 27.0, with independent-sample t-tests for cost comparisons and binary logistic regression to identify factors influencing surgical intervention. RESULTS: The study analyzed 3,806 lung cancer patients (60.7% male, 51.1% aged 61-80 years), revealing significant demographic and clinical predictors of surgical intervention. Multivariate analysis identified female gender (adjusted OR = 1.989, 95%CI:1.624-2.436, p < 0.001), younger age (61-80 years: adjusted OR = 0.454, 95% CI:0.305-0.676, p = 0.001; >80 years: adjusted OR = 0.353, 95%CI: 0.161-0.774, p = 0.009), and prolonged hospitalization (adjusted OR = 106.729, 95%CI: 79.485-143.312, p < 0.001) as key determinants, while insurance type showed no association. Surgical patients incurred 8.5-fold higher median costs (¥48,610 vs. ¥5,676), with medical consumables exhibiting the most pronounced disparity (>2,000-fold difference). The predictive model demonstrated excellent discrimination (AUC = 0.904), calibration (Hosmer-Lemeshow p = 0.402), and specificity (81.6%) at a 46.75% probability threshold, with length of stay being the strongest individual predictor (univariate AIC = 2,874 vs. full model AIC = 2,802). Insurance type showed no significant association with surgical treatment in either univariate or multivariate analyses. CONCLUSION: Gender, age, and hospital stay length were major factors associated with surgical decisions in lung cancer patients, with surgery significantly increasing total medical costs. Policy improvements in the management of high-value medical consumables and insurance reimbursement are needed to reduce financial burdens and enhance surgical accessibility.