PURPOSE: China has developed and widely implemented an innovative case-based payment method for inpatient services under a regional global budget, termed the "Diagnosis-Intervention Packet" (DIP). This study aims to examine cost-shifting behaviour in lung cancer inpatient care under the DIP reform. METHODS: This study examines the impact of the DIP reform in Zunyi, a national pilot city, using double machine learning (DML). Specifically, we analyze the effects on the total health expenditures (THS), individual payments excluding reimbursement (IPER), proportion of IPER, copayments for category-B, proportion of copayments for category-B, copayments for category-C and proportion of copayments for category-C per case for LC inpatients in tertiary hospitals. RESULTS: The results indicate a significant reduction in THS per case after the DIP reform (β = -0.0778, p <
0.001). Following the reform, there was a significant increase in IPER (β = 0.0689, p <
0.05), copayments for category-B (β = 0.1682, p <
0.01), and the proportion of copayments for category-B (β = 0.0039, p <
0.05). Conversely, the proportion of copayments for category-C significantly decreased (β = -0.0108, p <
0.001). Notably, significant heterogeneity in the cost-containment and cost-shifting effects was observed across different hospital categories, teaching types, and insured classifications. CONCLUSION: The DIP reform significantly reduced the THS per case for LC inpatients, while shifting in-policy expenditures to IPER. The cost-shifting primarily occurred through the redistribution of copayments from category-C to category-B. It is imperative for policymakers to establish differentiated regulatory policies tailored to various cost categories, hospital types, and insured classifications to optimize the effectiveness of the DIP reform.