Does DRG-based payment lead to unintended effects on care quality? A case under global budget with price adjustment in China
Xinyue Dong, Jing Wu
Abstract
BACKGROUND: The DRG-based payment system has gained widespread adoption globally, yet evidence about its impact on care quality remains scarce. In 2020, China implemented its inaugural national DRG payment initiative, the China Healthcare Security-Diagnosis Related Group (CHS-DRG), driven by the pressing need to regulate healthcare provider behaviors. OBJECTIVE: To evaluate the changes in process and outcome quality associated with the CHS-DRG under a global budget with price adjustment in Tianjin, China. METHODS: A retrospective, population-based analysis was conducted on 932 354 inpatient cases from pilot hospitals and 608 940 from non-pilot hospitals between November 2021 and September 2022, utilizing social medical insurance claims data and DRG payment enrollment files. A difference-in-differences approach was employed to examine the impact on diagnostic and therapeutic technology utilization, patient care experience, and moderately to extremely severe outcomes. The corresponding primary outcomes were the proportion of laparoscopic surgeries, waiting time for surgery, 30-day readmission rates and all-cause mortality. Robustness checks and common trends tests were performed to validate the findings. RESULTS: The improved efficiency in diagnostic and therapeutic practices was evidenced by a 19.17%-point increased likelihood of performing laparoscopic surgery (0.83%, P < 0.001) and a 53.20%-point reduction in the use of low-value drugs (-1.90%, P < 0.001), potentially contributing to a substantial relative decline of 2.57%-point in 30-day readmissions (-0.22%, P = 0.033) and 5.66%-point emergency department visits (-0.36%, P = 0.320). However, the CHS-DRG system was associated with longer waiting time for surgery by 5.96%-point (0.18 days, P < 0.001) and a 9.32%-point increase in the proportion of outpatient visits within three days before admission and after discharge (3.33%, P < 0.001), possibly compromising the patient care experience. Additionally, these changes were accompanied by a 6.88%-point increase in 30-day all-cause mortality (0.28%, P = 0.024). Notably, effect heterogeneity with respect to hospital characteristics and patient clinical severity was observed. CONCLUSIONS: The observed positive and negative effects highlight the imperative for continued research into the mechanism design of DRG payment systems to safeguard process and outcome quality.