The effects of the diagnosis-intervention packet payment reform in China: evidence from Guangzhou

Background This study evaluated the impact of Guangzhou’s Diagnosis-Intervention Packet (DIP) payment reform, introduced as an alternative to the traditional fee-for-service (FFS) model, on inpatient costs, patients’ out-of-pocket (OOP) payments, hospital length of stay (LOS) and 30-day readmission rate. Method We conducted a single-group interrupted time series analysis using monthly data from the Guangzhou Urban Employee Basic Medical Insurance Scheme, spanning January 2017 to June 2020. Outcome indicators included inpatient cost per case, medication expenditures, medical consumables expenditures, diagnostic and therapeutic service expenditures, OOP payments, LOS and 30-day readmission rate. Results While the DIP reform was associated with a modest reduction in the rate of LOS decline, it did not significantly lower inpatient cost. Instead, total inpatient expenditures exhibited a sustained upward trend in the post-reform period. Moreover, OOP payments per case increased significantly following the reform, indicating a heavier financial burden on patients. Sub-item analysis revealed that medication expenditures declined and stabilized after a pre-policy drop, whereas diagnostic and therapeutic service expenditures continued to grow without evident signs of deceleration. The 30-day readmission rate remained largely stable throughout the study period. Conclusion The DIP reform in Guangzhou did not achieve its intended goals of reducing inpatient expenditures or alleviating patients’ financial burdens. These unintended effects were driven by the incentive structure of the DIP system—particularly its reliance on historical cost data for RW calculation and its reimbursement deduction method—which encouraged hospitals to shift costs and elevate OOP payments. Policymakers should reassess DIP’s algorithmic parameters and strengthen cost-accounting transparency to ensure more equitable and efficient medical insurance fund allocation.