Cost shifting or cost cutting by hospitals as a response to reimbursement reform? The case of diagnosis-related groups (DRG) scheme in China

Background: Diagnosis-related group (DRG) systems for healthcare reimbursement were recently introduced among hospitals in China, raising concerns about cost-shifting, where hospitals may increase charges for self-financing patients to offset reimbursement cuts by DRG. In 2018, both Nanchang and Ganzhou Cities in Jiangxi Province installed DRG information systems, but only Nanchang fully implemented the DRG system during the 2019–2020 pilot period.

Materials and methods: Drawing from a healthcare administrative dataset of 14,310 patients' records, this study investigates the hospitalization costs associated with Intraocular Lens (IOL) implantation procedures in Jiangxi Province, China, from 2017 to 2020. By applying the quantile difference-in-differences (DID) and difference-in-difference-in-difference (DDD) methodologies, the research examines the impacts of DRG implementation on hospitalization costs, with a particular focus on self-financing patients.

Results: Upon the implementation of DRG in hospitals in Nanchang, the IOL cost ratio for patients in Nanchang was lower on average by 0.047 (p<0.001), with the largest reduction observed at higher quantiles (e.g., Q75: −0.105, Q90: −0.091) (p < 0.001). The DDD model revealed a decrease of 0.226 in the IOL cost ratio for self-financing patients in DRG hospitals post-implementation. These results remained robust across various healthcare cost ratios.

Conclusion: The study indicates that implementing the DRG system in Chinese hospitals effectively constrained healthcare costs without causing unintended cost-shifting among self-financing patients. The system's heightened impacts on higher quantiles indicate its efficacy in addressing high-cost outliers. These outcomes are likely attributed to the centralized medical procurement and healthcare information technology integrated into the Chinese healthcare system.

1 Introduction

The rising financial burden of medical costs globally has prompted policymakers and the public to seek cost containment strategies (1). The prospective payment system (PPS) with diagnosis-related groups (DRG) has been widely adopted as an effective tool for controlling healthcare costs (2–8). Under this system, hospitals receive fixed payments based on patients' diagnostic classifications, eliminating the financial incentive for overtreatment (9–11). However, hospitals may engage in cost shifting, raising costs to privately insured patients to offset financial losses from reduced DRG reimbursements (12–17). The practice of cost shifting thus affects the financial wellbeing of patient groups (18–22). Empirical studies based on financial data from U.S. hospitals have found mixed results (23–26). Some studies argue against the prevalence of dynamic cost shifting (23, 27–30), while others indicate that hospitals may engage in cost shifting to mitigate financial pressures (26, 31–39).

From 2019 to 2021, China launched a national pilot project of a standardized DRG payment system among public hospitals in 30 selected cities. As the Chinese government has ordered the adoption of the DRG payment system nationwide, there has been a heated debate about the potential impacts on the financial status of hospitals and treatment outcomes of patients (11, 40–42), especially concerns regarding cost shifting to self-financing patients in China (37, 39, 43–48). However, few empirical studies have been performed testing cost shifting among Chinese hospitals, and mixed findings were reported (42, 49–52).

This study leverages a quasi-natural experiment arising from the implementation of the DRG system in Jiangxi Province, China. In 2018, both Nanchang City and Ganzhou City installed national-standard DRG information systems. However, during the pilot-experiment period from 2019 to 2021, only Nanchang City was designated to fully implement the DRG system for over 94% of disease groups in the city's public hospital system. This unique context provides a valuable research opportunity to examine the impact of DRG adoption.

Cataract intraocular lens (IOL) implantation is well-suited to studying hospital cost-shifting due to the financial incentives associated with it. Despite being a standardized procedure, significant price variation exists among IOL types (53), with patients increasingly opting for self-financed lenses for improved outcomes. These features make IOL implantation procedures a valuable context for examining cost-shifting practices.

As a traditional inland agricultural region in central-eastern China, Jiangxi Province has a high incidence of cataracts due to its lower living standard and less developed economy during the past decades. In 2009, it had ~250,000 cataract patients with 18,000 new cases annually, and blindness cases due to cataracts accounted for 0.58% of the population in this province (54). Thus, the administrative data on IOL procedure hospitalization costs from Jiangxi Province represents a valuable dataset for studying this prevalent disease in China.

By applying the difference-in-difference model to leverage the quasi-experiment of the DRG system in Jiangxi Province, this study aims to specifically examine the impact of the DRG system on the IOL cost ratio for implantation procedures, as well as the cost differences among self-financing patients.

2 Data and methodology

2.1 Data source and sampling

Data for this study were obtained from the DRG database administered by the National Healthcare Security Administration in Jiangxi Province. The selection criteria were as follows:

1) Procedure type: Cataract intraocular lens (IOL) procedures are categorized under Lens Surgery (CHS-DRG code: CS10A) and Ambulatory Lens Surgery (CHS-DRG code: CS10B).

2) Age range: this study focuses on patients aged 50 and 80. This age group is selected due to its highest prevalence of cataracts in China (55), where the incidence of cataracts among individuals over 60 years old reaches 80%, accounting for over 50% of visually disabled individuals (56). Most studies on cataract surgery ehave established age 50 as the baseline to ensure inclusivity. Patients older than 80 are excluded from this study, as they often present with more complex health conditions and experience significantly higher healthcare utilization (57).

3) Length of stay (LOS): limiting to hospital stays of one or 2 days to control for case mix and severity. Ambulatory lens surgery (CHS-DRG code: CS10B) indicates cases with 1-day hospitalization.

4) Insurance coverage: including patients covered under urban employees/resident health insurance or those self-financed, ensuring a clear and consistent sample.

5) Hospital accreditation: this study focused exclusively on cases from tertiary public hospitals with Class A accreditation, the highest accreditation level in China. Class A hospitals are recognized for their superior professional services, advanced equipment, and exceptional management skills (58). They are highly trusted by patients, regardless of the severity of their conditions (59), and are the preferred choice for most intraocular lens (IOL) procedures in the country (60). Thus, consistent and high-quality healthcare cost data can be obtained from these hospitals, ensuring reliability and comparability.

6) IOL procedure record: this study focuses exclusively on hospitalization records for patients who underwent the intraocular lens (IOL) procedure, identified by associated IOL costs and surgery charges. Follow-up visits were excluded from this analysis. As per standard practice, the IOL procedure is performed on one eye at a time. Therefore, for patients who received the IOL procedure for both eyes, there are two separate hospitalization records.

7) Procedure costs: total costs of each episode include IOL cost, surgical fees, medication expenses, operating room utilization, surgical materials, diagnostic testing, nursing services, medical equipment, and anesthesia services (61). Records exhibiting implausible cost components were excluded. Specifically, those with IOL costs falling below CNY 200 were excluded, as these may indicate inaccurate documentation or incomplete cost recording. Additionally, cases with total episode costs exceeding CNY 25,000 were excluded, since these abnormally high amounts likely resulted from complications or additional treatments beyond standard cataract surgery.

8) Study period: July 2017 to December 2020, covering 2 years before and after the experimental period.

The final dataset comprises 14,310 patients' records, encompassing demographics, clinical diagnosis, resource utilization metrics, cost data, DRG grouping factors, and policy-linked variables. All healthcare cost data were standardized to 2020 price levels using the Chinese Medical Inflation Index and are reported in Chinese Yuan (CNY) based on the original billing records. Approximate United States Dollar (USD) equivalents, provided for international reference, were calculated using the 2020 annual average exchange rate (1 USD = 6.9 CNY).

2.2 Statistical method

This study adopted the difference-in-difference (DID) model to evaluate the impact of the DRG payment system on the total cost of IOL implantation procedures. In 2018, both Nanchang City and Ganzhou City installed the national-standard DRG information systems and conducted comprehensive training. Meanwhile, during the experiment stage from January 2019 to January 2020, only Nanchang actually adopted the DRG system for the healthcare cost management of most diseases. Recognizing that DRG payment reforms exhibit time-lagged implementation effects, where initial policy impacts may be obscured by systemic adaptation periods. Thus, we excluded the data from August to December 2018 to avoid capturing transitional anomalies. The pre-implementation baseline was consequently defined as July 2017 to July 2018, ensuring uncontaminated measurement of pre-reform medical costs prior to DRG rollout. As a result, patients' records in Nanchang City constituted the treatment group during the post-2019 DRG implementation period. By introducing an interaction term between the post-DRG period and the treatment group indicator, we captured the DID effect.

To gain a more detailed understanding of how the DRG implementation affects medical costs across different points of the cost distribution, we further utilized the quantile difference-in-difference (QDID) model. Quantile regression allows us to estimate the treatment effect at various quantiles, especially 10th, 25th, 50th, 75th, and 90th percentiles of the cost distribution, rather than focusing solely on the mean. This approach provides several advantages. First, the quantile regression method is robust to outliers in the outcome variable, which is very common in healthcare costs. Secondly, it offers a complete view of the relationship between variables by capturing the effects at different points in the outcome distribution (62, 63). This approach allowed for better identification of heterogeneous treatment effects, offering a more nuanced understanding than focusing solely on the mean (64).

To further examine how the treatment costs of self-financing patients varied upon the implementation of the DRG policy, we adopted the Difference-in-Difference-in-Differences (DDD) method. The DDD approach extends the DID framework by incorporating an additional layer of comparison, enabling us to isolate the effect of the DRG policy on self-financing patients specifically, who are particularly susceptible to cost shifts. This method involves a triple interaction term between the post-DRG implementation period, the treatment group indicator, and a binary variable indicating self-financing status. By doing so, we can disentangle the unique impact of the DRG policy on self-financing patients from other potential confounding effects. The DDD model was similarly adjusted for patient-level and hospital-level covariates to ensure robustness. This model specifically isolates the policy impact on self-financing patients, who are particularly susceptible to cost shifts.

2.3 Empirical models

2.3.1 DID model

The basic empirical equation has the specification as follows:

$\begin{array}{l}{Y}_{i,h,t}={\beta }_0+{\beta }_{1}Interventio{n}_h\times Pos{t}_t+{\beta }_{2}Pos{t}_t+{\beta }_{3}Interventio{n}_h\\ +{\beta }_{4}Self\_financin{g}_{i,t}+\delta X_{i,h,t}+{\epsilon }_{i,h,t}& (1)\end{array}$

where i represents a patient, h for a hospital, and t for a year, ε_{i, h, t} is a random error term.

Y_{i, h, t} represents the outcome variables: IOL cost ratio, log of IOL costs, and log of total episode costs, respectively. The IOL cost ratio indicates the proportion of the total episode costs attributed to the IOL costs. The relative ratio of intraocular lens (IOL) cost standardizes costs across diverse patient scenarios, allowing for meaningful comparisons. This approach mitigates the influence of outliers, offers a clearer understanding of typical cost distributions, and strengthens the robustness of our economic evaluations. The logarithm of IOL costs and total episode costs are adopted, because the distribution of healthcare costs is typically skewed.

The binary variable Intervention_h indicates an IOL case in the treatment group of Nanchang City (Intervention_{i, t} = 1), with Ganzhou (Intervention_{i, t} = 0) as the reference group. Post_t indicates observations from the period after DRG implementation (Post_{i, t} = 1: January 2019 to December 2020), with the period of July 2017 to July 2018 (Post_{i, t} = 0) as reference. Accordingly, the interaction term Intervention_h×Post_t of DID model captures the effect of the DRG payment reform.

For a better comparison, this study placed Self_financing_{i, t} as a separate control variable in the equation. Self_financing_{i, t} is a dummy variable that Self_ financing_{i, t} = 0 represents patients covered by Urban Employee Basic Medical Insurance (UEBMI) and Urban Residents Basic Medical Insurance (URBMI), while Self_financing_{i, t} = 1 identifies patients without social health insurance.

X_{i, h, t} represents a sector that includes various patient characteristics, such as age, gender, and length of stay (LOS).

2.3.2 DDD model

The DDD model is established as Equation 2.

$\begin{array}{l}{Y}_{i,h,t}={\beta }_0+{\beta }_{1}Interventio{n}_h\times Pos{t}_t\times Self\_financin{g}_{i,t}\\ +{\beta }_{2}Pos{t}_t\times Self\_financin{g}_{i,t}+{\beta }_{3}Interventio{n}_h\\ \times Self\_financin{g}_{i,t}+{\beta }_{4}Interventio{n}_h\times Pos{t}_t+{\beta }_{5}Pos{t}_t\\ +{\beta }_{6}Interventio{n}_h+{\beta }_{7}Self\_financin{g}_{i,t}\\ +\delta X_{i,h,t}+{\epsilon }_{i,h,t}& (2)\end{array}$

The coefficient of the triple interaction term, β₁ is the DDD estimator of interest, representing the net effect of the DRG payment reform among the self-financing patients.

Standard errors of quantile regressions were calculated by bootstrapping with 1,000 repetitions. Stata version 15.0 (Stata Corp LP, College Station, TX) was adopted for performing all statistical analyses in this study.

3 Empirical analysis

3.1 Descriptive results

Table 1 presents descriptive statistics of the analytical cohort (N = 14,310). Geographically, 75.49% (n = 10,802) of patients were treated in Nanchang vs. 24.51% (n = 3,508) in Ganzhou. Patients exhibited a mean age of 68.7 years, concentrated in the 70–80-year cohort (48.65%), which was consistent with cataract epidemiology in aging populations. Females comprised the majority of cases (59.13%; male-to-female ratio = 1:1.45), while reimbursement sources indicated 72.07% used medical insurance vs. 27.93% self-pay arrangements.

Table 1

Table 1. Descriptive statistics.

In panel B, the IOL costs constituted, on average, 39% of total episode costs. Mean IOL costs and total episode costs per patient were CNY 3344 (USD 484.6) and CNY 8728 (USD 1264.9), respectively. Regional analysis revealed higher mean costs in Nanchang (IOL: CNY 4065, Total: CNY 9943) compared to Ganzhou (IOL: CNY 1122, Total: CNY 4987). Patients were discharged after a mean hospital length of stay of 2.58 days. All analyzed variables exhibited mean values and standard deviations within the expected range, with no significant statistical outliers identified.

3.2 Parallel trend test

The “parallel trend assumption” is an important prerequisite for using the DID model to judge policy effectiveness (65–68). We plotted the time trend of the mean value of Y, which was plotted by month for the intervention and control groups in a simple and intuitive way.

Figure 1 shows the results of the DID parallel trend test. All the dependent variables between the two cities maintained a consistent trend before July 2018, while after that, the trend of the two groups showed a significant difference. That means the changes in the outcome variables between the two groups before and after the DRG's payment reform have nothing to do with the control group. Thus, the DID method can be used to evaluate the pure policy effects.

Figure 1

Figure 1. DID parallel trend test (time trend graph of the mean of Y).

3.3 Regression analysis of the DID model

Table 2 reports the estimation results of the intervention's effects on three cost measures, respectively. Panel A indicates that the intervention significantly reduced the IOL cost ratio post-implementation. As reported in column 1, the estimated coefficients of the interaction term (Intervention × Post) are negative and statistically significant on average and across most quantiles. This indicates that post-DRG implementation, the average IOL cost ratio in Nanchang hospitals declined by 0.047 (4.7 percentage points) relative to Ganzhou. It shows DRG effectively contained IOL-related expenditures without increasing total hospitalization costs, demonstrating its efficacy in targeted cost control. The largest reduction is observed at higher quantiles (e.g., Q75: −0.105, Q90: −0.091), with maximal declines occurring among high-cost patients where financial reallocation was most probable. For the age group, the results in column 1 are not significant at the 60–69-year range; however, the IOL cost ratio is relatively slightly decreased at the 75th and 90th percentiles in older patients, which is significant at the 1% confidence level. The IOL cost ratio of self-financing patients slightly decreased by 3.1 percentage points after DRG reform at the 1% significance level, and the decline was even greater at higher quantiles (e.g., Q75: −0.038, Q90: −0.047).

Table 2

Table 2. Result of DID estimates (OLS and quantile regressions)

The findings reported in Panels B and C keep the key results. The results show that IOL costs and total episode costs have significantly decreased in the intervention group (Nanchang city) after DRG reform by ~4.25 and 3.44 percentage points at the 1% significance level, respectively. Notably, these two costs fell the most at the 75th percentile. The full results were shown in Appendix Tables A.1, A.2. These effects are visualized in Figure 2.

Figure 2

Figure 2. Quantile regression results of the DID model.

3.4 Regression analysis of the DDD model

As reported in Table 3, for the IOL cost ratio, the coefficient of the triple interaction term is −0.226 (p < 0.001), indicating a reduction in the cost ratio for self-financing patients in the treatment group post-policy period. Across all three dependent variables, the coefficient for the triple interaction term is negative and highly significant at the 1% level. While the IOL costs and total episode costs of self-financing patients were also significantly decreased, more than those of insured patients, after DRG's payment reform in Nanchang city. This resulted in the DRG policy curbing the growth of medical costs for self-financing patients.

Table 3

Table 3. Key results of DDD estimates (OLS).

Figure 3 shows the quantile regression results of the DDD model on triple interaction terms. As illustrated in Figure 3, the policy effects among self-financing patients varied across quantiles, with the largest reduction at the 50th percentile (refer to Appendix Table A.4).

Figure 3

Figure 3. Quantile regression results of the DDD model on triple interaction terms.

3.5 Robustness check

To assess whether significant differences exist across the quantiles, we conducted interquartile tests by comparing other percentiles with the 50th percentile using the iqreg command in STATA (Appendix Table A.5).

An additional test incorporating observations from August to December 2018, which corresponds to the introduction period of DRG, was excluded from the main analysis. The results remained consistent (Appendix Table A.6).

4 Discussion

4.1 The cost-control effect of DRG policy

Applying the DID models, this study consistently found that the implementation of the DRG system in Nanchang City significantly reduced healthcare costs as measured by the ratio of IOL in the total cost of hospitalization. The results of this study align with prior research (11, 48, 49, 69–73). This might be because the price for a particular disease, including healthcare services, drugs, and medical consumables, was predetermined after DRG. Thus, doctors will concentrate more on the principal diagnosis and principal treatment procedure (11). Unlike the previous fee-for-service system, the case-based policy could motivate doctors to standardize diagnosis and treatment, and physician-induced overtreatment and medical costs could also be reduced.

Furthermore, the quantile regression results revealed substantial heterogeneity in the intervention's impact, with stronger effects observed at higher quantiles, where overtreatment and unnecessary resource utilization are most likely to occur (11). It might be because doctors are more likely to avoid unnecessary high-value medical treatment than induce patients to use higher-priced lenses after DRG reform. A previous study showed that healthcare providers would reduce overtreatment behavior after DRGs' reform (11). These findings underscore the dual value of the DRG system, a perspective that has not yet been addressed in the existing literature.

4.2 Impact of DRG reform on insured and self-financing patients and cost-shifting dynamics

The findings in the DDD model suggested that after the implementation of the DRG system in Nanchang City, the IOL cost ratio of self-financing patients was significantly lower by about twenty-three percentage points when compared with those having social healthcare insurance, ceteris paribus. This could be considered evidence that the cost-control effect of the DRG policy was more focused on self-financing patients and that the effect of the DRG policy was more pronounced among patients without social security. Thus, this empirical evidence did not support the cost-shifting hypothesis.

This result was consistent with studies in China (49, 52). However, some studies in China found opposite results: cost shifting to self-financing patients, uninsured, older, or more complicated patients after reform of the medical system (37, 39, 44, 46–48, 74). This finding also contrasted with previous literature in developed countries discussing cost-shifting, where hospitals shift part of medical costs to self-financing patients to cover losses from government cutbacks (15, 35, 36).

The cost-shifting could happen in developed countries, probably because it is popular for self-financing patients to purchase commercial insurance, which is likely the target of healthcare cost-shifting. The medical costs shifted to self-financing patients will ultimately be borne by insurance companies. In China, it's a different story. China's health insurance participation rate remains above 95% (75), and fewer patients purchase commercial insurance. Thus, if hospitals shift medical costs to self-financing patients, these costs will probably be borne by the patients themselves. For those self-financing patients, some patients do not have medical insurance, some patients are high-net-worth individuals who use expensive imported lenses, which are not reimbursable by medical insurance, and some patients take interprovincial medical treatment, so they can't use their medical insurance card. Self-financing patients, who bear the full burden of medical expenses out-of-pocket, tend to exhibit greater price sensitivity (76). Moreover, China's doctor-patient relationship has become more tense in recent years, and doctors may be less likely to induce high-priced medical services to avoid disputes. At the same time, after the implementation of centralized procurement of lenses, the prices of medical consumables are more open and transparent, the medical costs are under supervision of big data technology, any excessive self-financing ratios might display abnormal status in the health insurance system which will attract official investigation, thus for those self-financing patients, physicians may be less likely to induce extravagant health services. For these reasons, medical costs are not easy to pass on, and cost-shifting is also more challenging to implement.

Moreover, these positive spill-over effects of the DRG system on self-financing patients may be owing to the concurrent centralized medical procurement and healthcare information technology in China, which have improved pricing transparency for medical consumables (77, 78).

4.3 Strengths and limitations

This research delivers several field-advancing contributions to the health policy literature. First, by integrating QDID with the DDD method, the study detected distributional effects across patient subgroups in policy evaluation. Second, the research validated the spill-over effect on self-financing patients, disproving universal cost-shifting assumptions and establishing DRG as an equity-enabling policy. Third, the cost-saving effect verified in the research enables global policymakers, especially developing countries and low- and middle-income countries, to codify cross-group equity into further medical reform.

The study is subject to several limitations, which offer opportunities for future research. First, the primary limitation of this study lies in the data, which is derived solely from public hospitals within a province with a less developed economy in China. Findings may not generalize to wealthy provinces or private hospitals, where financial incentives and patient demographics differ. Further research should test DRG's impact across diverse regions and hospital types to assess generalizability. Due to data limitations, this study was unable to account for patient comorbidities or socioeconomic status. However, this is expected to introduce minimal bias, as the sample was restricted to a homogeneous, low-risk group undergoing standardized cataract surgery. Thus, the cost changes identified are attributable to physician behaviors and policy impacts. Future studies may explore variabilities further when comorbidity and socioeconomic data become available. Second, the study only focused on IOL implantation, a high-volume but technically standardized procedure. The findings of this study may not apply to complex, variable-cost procedures. Future research should broaden its scope to high-cost, heterogeneous procedures to identify whether DRG's anti-cost-shifting effect holds universally. Finally, reimbursement cuts could theoretically impact care quality despite efforts to constrain costs. This study did not assess the long-term impacts on clinical outcomes or patient satisfaction. Future research should track longitudinal data to evaluate comprehensive health outcomes.

5 Conclusion

This study investigates the impact of diagnosis-related groups (DRG) reimbursement systems on hospitalization costs in China, using the cases of Intraocular Lens (IOL) implantation procedures across two major cities in Jiangxi Province of China from 2017 to 2020.

DRG significantly reduced the IOL cost ratio. Crucially, maximal reductions occurred precisely where cost-shifting risk was highest, disproving concerns that hospitals would inflate charges for complex or high-cost cases. While the DDD estimator revealed a reduction in self-financing patients' IOL cost ratio post-DRG, reflected DRG policy brought a spill-over effect to them, which means the unintended yet beneficial consequence wherein DRG's cost-containment mechanisms disproportionately protected self-financing patients instead of triggering financial burden transfers. DRG policy also compressed hospitalization costs and established DRG reimbursement as a cost containment instrument for the medical system. These empirical findings suggest that DRG policies, when coupled with strong regulatory oversight and information technology support, can achieve their cost-containment goals effectively without major cost-shifting issues.

Given these positive outcomes, the study supports ongoing policy advancements to strengthen DRG mechanisms further and maintain high levels of governance transparency in the healthcare industry. Future research should expand these findings across varied clinical and regional contexts.

Data availability statement

The data analyzed in this study is subject to the following licenses/restrictions: Relevant administrative permissions were received from the Ganzhou Municipal Provincial Health Committee of China, who provided dataset for the study. Any prospective user of the original dataset should apply for separate permission directly from the health administration authority. Requests to access these datasets should be directed to Jinyun Zhu, Z3pzd2p3amdkd0BnYW56aG91Lmdvdi5jbg==.

Author contributions

JW: Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Resources, Software, Validation, Visualization, Writing – original draft, Writing – review & editing. JZhu: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Resources, Writing – review & editing. KH: Formal analysis, Visualization, Writing – review & editing. YC: Investigation, Methodology, Software, Validation, Writing – review & editing. JZha: Conceptualization, Formal analysis, Funding acquisition, Methodology, Project administration, Software, Supervision, Writing – review & editing. XW: Data curation, Formal analysis, Writing – review & editing.

Funding

The author(s) declare that financial support was received for the research and/or publication of this article. This study was partially funded by 2023 Guangdong Basic and Applied Basic Research Foundation (Grant No. SL2022A04J01523) and partially funded by Jiabi Wang's Nanfang College Ph.D. Programs Foundation (Grant No. 2022BQ012); Jinghua Zhang was financially supported by the Macau University of Science and Technology Foundation, Faculty Research Grant (Grant No. FRG-25-077-MSB). The funders had no role in the design of the study, the collection, analysis, interpretation of the data, or the writing of the manuscript.

Acknowledgments

We would like to express our gratitude to the Ganzhou Municipal Health Commission in Jiangxi Province of China for granting access to hospital administrative data. We also extend our thanks to Dr. Xiaolei Tang, Dr. Yu Song, and Dr. Shun-Ho Chu at the School of Business, Macau University of Science and Technology, for their invaluable discussions and insights.

Conflict of interest

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could potentially conflict of interest.

Generative AI statement

The author(s) declare that no Gen AI was used in the creation of this manuscript.

Publisher's note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fpubh.2025.1582001/full#supplementary-material

References

1. OECD Health Statistics. Health expenditure and financing: Health expenditure indicators. (2024). (Accessed December 05, 2024). Available online at: https://www.oecd.org/en/data/indicators/health-spending.html

Google Scholar

2. Fetter R. Diagnosis related groups in Europe: uses and perspectives. Springer Science & Business Media (2012).

Google Scholar

3. Goldfield N. The evolution of diagnosis-related groups (DRGs): from its beginnings in case-mix and resource use theory, to its implementation for payment and now for its current utilization for quality within and outside the hospital. Qual Manag Healthcare. (2010) 19:3–16. doi: 10.1097/QMH.0b013e3181ccbcc3

PubMed Abstract | Crossref Full Text | Google Scholar

4. Mihailovic N, Kocic S, Jakovljevic M. Review of diagnosis-related group-based financing of hospital care. Health Serv Res Manag Epidemiol. (2016) 3:2333392816647892. doi: 10.1177/2333392816647892

PubMed Abstract | Crossref Full Text | Google Scholar

5. Lin A-L, Hou J-H. Diagnosis-Related Groups payment reform and hospital cost control. Tech Health Care. (2025) 33:17–24. doi: 10.3233/THC-240110

PubMed Abstract | Crossref Full Text | Google Scholar

6. Zheng X, Xiao X, Shen K, Pei T, Lin X, Liu W, et al. Impact of Diagnosis-Related Groups (DRG) reform on cost homogeneity of treatment for patients with malignant tumours. Sci Rep. (2024) 14:21212. doi: 10.1038/s41598-024-71917-y

PubMed Abstract | Crossref Full Text | Google Scholar

7. Wang S, Wu N, Wang H, Zhang X, Li F, Wang X, et al. Impacts of a new diagnosis-related group point payment system on children's medical services in China: Length of stay and costs. Int J Health Plann Manage. (2024) 39:432–46. doi: 10.1002/hpm.3739

PubMed Abstract | Crossref Full Text | Google Scholar

8. Li Q, Fan X, Jian W. Impact of Diagnosis-Related-Group (DRG) payment on variation in hospitalization expenditure: evidence from China. BMC Health Serv Res. (2023) 23:688. doi: 10.1186/s12913-023-09686-z

PubMed Abstract | Crossref Full Text | Google Scholar

9. Yip WC-M, Hsiao W, Meng Q, Chen W, Sun X. Realignment of incentives for health-care providers in China. Lancet. (2010) 375:1120–30. doi: 10.1016/S0140-6736(10)60063-3

PubMed Abstract | Crossref Full Text | Google Scholar

10. Busse R, Geissler A, Aaviksoo A, Cots F, Häkkinen U, Kobel C, et al. Diagnosis related groups in Europe: moving towards transparency, efficiency, and quality in hospitals? BMJ. (2013) 346. doi: 10.1136/bmj.f3197

PubMed Abstract | Crossref Full Text | Google Scholar

11. Zhang L. Lihua Sun. Impacts of diagnosis-related groups payment on the healthcare providers' behavior in China: a cross-sectional study among physicians. Risk Manag Healthcare Policy. (2021) 14:2263–76. doi: 10.2147/RMHP.S308183

PubMed Abstract | Crossref Full Text | Google Scholar

12. Ginsburg PB. Can hospitals and physicians shift the effects of cuts in medicare reimbursement to private payers? Cost shifting is more pronounced under certain conditions, but there definitely is an economic basis for its existence. Health Affairs. (2003) 22:W3-472-W3-479. doi: 10.1377/hlthaff.W3.472

Crossref Full Text | Google Scholar

13. Sloan FA, Becker ER. Cross-subsidies and payment for hospital care. J Health Polit Policy Law. (1984) 8:660–85. doi: 10.1215/03616878-8-4-660

PubMed Abstract | Crossref Full Text | Google Scholar

14. Cutler DM. Cost shifting or cost cutting?: the incidence of reductions in Medicare payments. Tax Policy Econ. (1998) 12:1–27. doi: 10.1086/tpe.12.20061853

Crossref Full Text | Google Scholar

15. Zwanziger J, Bamezai A. Evidence of cost shifting in California hospitals. Health Aff. (2006) 25:197–203. doi: 10.1377/hlthaff.25.1.197

PubMed Abstract | Crossref Full Text | Google Scholar

16. Robinson J. Hospitals respond to Medicare payment shortfalls by both shifting costs and cutting them, based on market concentration. Health Aff. (2011) 30:1265–71. doi: 10.1377/hlthaff.2011.0220

PubMed Abstract | Crossref Full Text | Google Scholar

17. Clement JP. Dynamic cost shifting in hospitals: Evidence from the 1980s and 1990s. Inquiry. (1997) 34:340–350.

PubMed Abstract | Google Scholar

18. Hadley J, Feder J. Hospital cost shifting and care for the uninsured. Health Aff. (1985) 4:67–80. doi: 10.1377/hlthaff.4.3.67

PubMed Abstract | Crossref Full Text | Google Scholar

19. Fuchs VR. Cost shifting does not reduce the cost of health care. Jama. (2009) 302:999–1000. doi: 10.1001/jama.2009.1278

PubMed Abstract | Crossref Full Text | Google Scholar

20. Dranove D. Pricing by non-profit institutions: the case of hospital cost-shifting. J Health Econ. (1988) 7:47–57. doi: 10.1016/0167-6296(88)90004-5

PubMed Abstract | Crossref Full Text | Google Scholar

21. Rice T, Stearns S, DesHarnais S, Brasure M, Pathman D, Tai-Seale M. Do physicians cost shift? Health Aff. (1996) 15:215–25. doi: 10.1377/hlthaff.15.3.215

PubMed Abstract | Crossref Full Text | Google Scholar

22. Hay JW. The impact of public health care financing policies on private-sector hospital costs. J Health Polit Policy Law. (1983) 7:945–52. doi: 10.1215/03616878-7-4-945

PubMed Abstract | Crossref Full Text | Google Scholar

23. Frakt AB. How much do hospitals cost shift? A review of the evidence. Milbank Q. (2011) 89:90–130. doi: 10.1111/j.1468-0009.2011.00621.x

PubMed Abstract | Crossref Full Text | Google Scholar

24. Morrisey MA. Cost shifting in health care: Separating evidence from rhetoric. American Enterprise Institute (1994).

Google Scholar

25. Cook A, Averett S. Do hospitals respond to changing incentive structures? Evidence from Medicare's 2007 DRG restructuring. J Health Econ. (2020) 73:102319. doi: 10.1016/j.jhealeco.2020.102319

PubMed Abstract | Crossref Full Text | Google Scholar

26. Mendoza RL. Cost-shifting and cost-cutting as joint and mutually reinforcing strategies in the financial management of hospitals and similar healthcare organizations. J Health Care Finance. (2020) 47:1–25.

Google Scholar

27. Dobson A, DaVanzo J, Sen N. The cost-shift payment ‘hydraulic': foundation, history, and implications. Health Aff. (2006) 25:22–33. doi: 10.1377/hlthaff.25.1.22

PubMed Abstract | Crossref Full Text | Google Scholar

28. Sloan F, Morrisey M, Valvona J. Case shifting and the Medicare Prospective Payment System. American J Public Health. (1988) 78:553–6. doi: 10.2105/AJPH.78.5.553

Crossref Full Text | Google Scholar

29. Dranove D, White WD. Medicaid-dependent hospitals and their patients: how have they fared? Health Serv Res. (1998) 33:163.

PubMed Abstract | Google Scholar

30. Zuckerman S. Commercial insurers and all-payer regulation: Evidence on hospitals' responses to financial need. J Health Econ. (1987) 6:165–87. doi: 10.1016/0167-6296(87)90007-5

PubMed Abstract | Crossref Full Text | Google Scholar

31. Neby S, Lægreid P, Mattei P, Feiler T. Bending the rules to play the game: Accountability, DRG and waiting list scandals in Norway and Germany. European Policy Analysis. (2015) 1:127–48. doi: 10.18278/epa.1.1.9

Crossref Full Text | Google Scholar

32. Batty M, Ippolito B. Financial incentives, hospital care, and health outcomes: evidence from fair pricing laws. American Economic Journal: Economic Policy. (2017) 9:28–56. doi: 10.1257/pol.20160060

Crossref Full Text | Google Scholar

33. Altman SH. The lessons of Medicare's prospective payment system show that the bundled payment program faces challenges. Health Aff. (2012) 31:1923–30. doi: 10.1377/hlthaff.2012.0323

PubMed Abstract | Crossref Full Text | Google Scholar

34. Saint S, Rose J, Lichter AS, Forrest S, McMahon Jr LF. Shifting costs from high-cost to low-cost diagnosis-related groups? Eval Health Prof. (2002) 25:259–69. doi: 10.1177/0163278702025003001

PubMed Abstract | Crossref Full Text | Google Scholar

35. Wu VY. Hospital cost shifting revisited: new evidence from the balanced budget act of 1997. Int J Health Care Finance Econ. (2010) 10:61–83. doi: 10.1007/s10754-009-9071-5

PubMed Abstract | Crossref Full Text | Google Scholar

36. Zwanziger J, Melnick GA, Bamezai A. Can cost shifting continue in a price competitive environment? Health Economics. (2000) 9:211–226.3. doi: 10.1002/(sici)1099-1050(200004)9:3<211::aid-hec508>3.0.co;2-k

Crossref Full Text | Google Scholar

37. Zhang M, Hu M. Application of DRGs in Beijing's medical insurance payment management. Chinese Health Insurance. (2015) 51–53, 57. doi: 10.3969/j.issn.1674-3830.2015.04.022 (in Chinese).

Crossref Full Text | Google Scholar

38. Milstein R, Schreyögg J. The end of an era? Activity-based funding based on diagnosis-related groups: a review of payment reforms in the inpatient sector in 10 high-income countries. Health Policy. (2024) 141:104990. doi: 10.1016/j.healthpol.2023.104990

PubMed Abstract | Crossref Full Text | Google Scholar

39. Liu F, Chen J, Li C, Xu F. Cost sharing and cost shifting mechanisms under a per diem payment system in a county of China. Int J Environ Res Public Health. (2023) 20:2522. doi: 10.3390/ijerph20032522

PubMed Abstract | Crossref Full Text | Google Scholar

40. Qian M, Zhang X, Chen Y, Xu S, Ying X. The pilot of a new patient classification-based payment system in China: the impact on costs, length of stay and quality. Soc Sci Med. (2021) 289:114415. doi: 10.1016/j.socscimed.2021.114415

PubMed Abstract | Crossref Full Text | Google Scholar

41. Zhang T, Lu B, Song Y, Chen M. Impacts of outpatient payment reforms on volume and expenditures in public hospitals: evidence from a quasi-experimental analysis in Zhejiang, China. Risk Manag Healthc Policy. (2023) 16:415–24. doi: 10.2147/RMHP.S400385

PubMed Abstract | Crossref Full Text | Google Scholar

42. Zou K, Li H-Y, Zhou D, Liao Z-J. The effects of diagnosis-related groups payment on hospital healthcare in China: a systematic review. BMC Health Serv Res. (2020) 20:1–11. doi: 10.1186/s12913-020-4957-5

PubMed Abstract | Crossref Full Text | Google Scholar

43. National Healthcare Security NAdministration. Does DRG lead to a decrease in medical income? Caused the hospital to lose money? This pot DRG can't be carried! (2024). Available online at: https://www.nhsa.gov.cn/art/2024/4/10/art_14_12336.html (Accessed October 08, 2024).

Google Scholar

44. Zhang J. The impact of a diagnosis-related group-based prospective payment experiment: the experience of Shanghai. Appl Econ Lett. (2010) 17:1797–803. doi: 10.1080/13504850903317347

Crossref Full Text | Google Scholar

45. Xu J, Jian W, Zhu K, Kwon S, Fang H. Reforming public hospital financing in China: progress and challenges. BMJ. (2019) 365:l4015. doi: 10.1136/bmj.l4015

PubMed Abstract | Crossref Full Text | Google Scholar

46. Barber SL, Borowitz M, Bekedam H, Ma J. The hospital of the future in China: China's reform of public hospitals and trends from industrialized countries. Health Policy Plan. (2014) 29:367–78. doi: 10.1093/heapol/czt023

PubMed Abstract | Crossref Full Text | Google Scholar

47. Tan H, Zhang X, Bi S, Chen Y, Guo D. How significant is cost-shifting behavior under the diagnosis intervention packet payment reform? Front Public Health. (2024) 12:1431991. doi: 10.3389/fpubh.2024.1431991

PubMed Abstract | Crossref Full Text | Google Scholar

48. Suwei Y, Wenwei L, Fengqing W, Zhang H, Suping W, Weijun Z, et al. Impacts of hospital payment based on diagnosis related groups (DRGs) with global budget on resource use and quality of care: a case study in China. Iran J Public Health. (2019) 48:238. doi: 10.18502/ijph.v48i2.818

PubMed Abstract | Crossref Full Text | Google Scholar

49. Jian W, Lu M, Chan KY, Poon AN, Han W, Hu M, et al. Payment reform pilot in Beijing hospitals reduced expenditures and out-of-pocket payments per admission. Health Aff. (2015) 34:1745–52. doi: 10.1377/hlthaff.2015.0074

PubMed Abstract | Crossref Full Text | Google Scholar

50. Yip W, Eggleston K. Provider payment reform in China: the case of hospital reimbursement in Hainan province. Health Econ. (2001) 10:325–39. doi: 10.1002/hec.602

PubMed Abstract | Crossref Full Text | Google Scholar

51. Meng Q, Rehnberg C, Zhuang N, Bian Y, Tomson G, Tang S. The impact of urban health insurance reform on hospital charges: a case study from two cities in China. Health Policy. (2004) 68:197–209. doi: 10.1016/j.healthpol.2003.09.010

PubMed Abstract | Crossref Full Text | Google Scholar

52. Xue M, Mao J, Chen Y. Chinese hospital bed occupancy analysis: 1988–1997. Chinese Hosp Manage. (1999) 19:313 (in Chinese). Available online at: https://kns.cnki.net/kcms2/article/abstract?v=si2fSHvxfzgrS2KFkskePY6n1IE4PtLtxWKenT73UEg0OGIWI_MY9J0Wt_R6au23kbaZtsVrPMLo3oTkKzA7Za1M4WOYZ-KYtmQMvyxNaAWb6-BnvbQKVzuHm0FWrn3rVmwQ-Fvg8_KnQuYxDSr50OALU-46h_eZLdKSnI1C7LxlHwk1jKoiBQ==&uniplatform=NZKPT&language=CHS

Google Scholar

53. Wong TY, Zheng Y, Jonas JB, Flaxman SR, Keeffe J, Leasher J, et al. Prevalence and causes of vision loss in East Asia: 1990–2010. Br J Ophthalmol. (2014) 98:599–604. doi: 10.1136/bjophthalmol-2013-304047

PubMed Abstract | Crossref Full Text | Google Scholar

54. The State Council The people's republic of China. Jiangxi Province: 250,000 cataract patients will be able to receive free sight restoration surgery. (2009). Available online at: https://www.gov.cn/jrzg/2009-04/29/content_1300383.htm (Accessed December 05, 2024).

Google Scholar

55. Xin J, Shuai G, Xiaoying Z. Cataract surgery incidence and influencing factors in middle-aged and elderly Chinese: a prospective cohort study. Chin J Dis Contr Prevent. (2024) 28:303–9. doi: 10.16462/j.cnki.zhjbkz.2024.03.009 (in Chinese)

Crossref Full Text | Google Scholar

56. Gao H, Chen X, Shi W. Analysis of the prevalence of blindness and major blinding diseases in China. Chin J Ophthalmol. (2019) 55:625–8. doi: 10.3760/cma.j.issn.0412-4081.2019.08.016 (in Chinese)

PubMed Abstract | Crossref Full Text | Google Scholar

57. Zou M, Guo D, Chen A, Young CA Li Y, Zheng D, Jin G. Prevalence of visual impairment among older Chinese population: A systematic review and meta-analysis. J Glob Health. (2021) 11:08004. doi: 10.7189/jogh.11.08004

PubMed Abstract | Crossref Full Text | Google Scholar

58. Ministry Ministry of Health, PRC. Implementation Rules for the Evaluation Standards of Tertiary General Hospitals, 2011 Edition (in Chinese). (2011). Available online at: https://www.nhc.gov.cn/cms-search/downFiles/02b5187a195c42e0855e4e27445fc2f0.pdf (Accessed May 26, 2025).

Google Scholar

59. Zhang JH, Peng X, Liu C, Chen Y, Zhang H, Iwaloye OO. Public satisfaction with the healthcare system in China during 2013–2015: a cross-sectional survey of the associated factors. BMJ Open. (2020) 10:e034414. doi: 10.1136/bmjopen-2019-034414

PubMed Abstract | Crossref Full Text | Google Scholar

60. National Health Commission of the PRC. Notice of the National Health Commission on the issuance of the “Three-level Hospital Evaluation Standards (2022 Edition)” and its implementation rules (in Chinese). (2022). Available online at: https://web.archive.org/web/20241104104622/https://www.gov.cn/zhengce/zhengceku/2022-12/18/content_5732583.htm (Accessed May 25, 2025).

Google Scholar

61. National Healthcare Security Administration. China Healthcare Security Diagnosis Related Groups (CHS-DRG) Grouping and Payment Technical Specifications. (2019). Available online at: http://www.nhsa.gov.cn/module/download/downfile.jsp?classid=0&filename=8bcc3d2f85434fb598cdc3b7e035ce39.pdf (Accessed June 20, 2025).

Google Scholar

62. Zhang JH, Chou S-Y, Deily ME, Lien H-M. Hospital ownership and drug utilization under a global budget: a quantile regression analysis. Int Health. (2014) 6:62–9. doi: 10.1093/inthealth/ihu001

PubMed Abstract | Crossref Full Text | Google Scholar

63. Callaway B, Li T. Quantile treatment effects in difference in differences models with panel data. Quant Econom. (2019) 10:1579–618. doi: 10.3982/QE935

Crossref Full Text | Google Scholar

64. Koenker R, Bassett G Jr. Regression quantiles. Economet J Economet Soc. (1978) 46:33–50. doi: 10.2307/1913643

Crossref Full Text | Google Scholar

65. Abadie A. Semiparametric difference-in-differences estimators. Rev Econ Stud. (2005) 72:1–19. doi: 10.1111/0034-6527.00321

Crossref Full Text | Google Scholar

66. Athey S, Imbens GW. Identification and inference in nonlinear difference-in-differences models. Econometrica. (2006) 74:431–97. doi: 10.1111/j.1468-0262.2006.00668.x

Crossref Full Text | Google Scholar

67. Chen Z, Song Z, Zhang C. Effect of separating treatment and drug sales: evidence form medical insurance claims data. J Finan Res. (2018) 10:7288 (in Chinese). Available online at: http://www.jryj.org.cn/CN/Y2018/V460/I10/72

Google Scholar

68. Meyer BD. Natural and quasi-experiments in economics. J Bus Econ Statist. (1995) 13:151–61. doi: 10.1080/07350015.1995.10524589

Crossref Full Text | Google Scholar

69. Ma Y, Wang W. The impact of diagnosis related group payment on the performance of public hospitals. Am J Transl Res. (2021) 13:6796–801.

Google Scholar

70. Ding Y, Yin J, Zheng C, Dixon S, Sun Q. The impacts of diagnosis-intervention packet payment on the providers' behavior of inpatient care—evidence from a national pilot city in China. Frontiers in Public Health. (2023) 11:1069131. doi: 10.3389/fpubh.2023.1069131

PubMed Abstract | Crossref Full Text | Google Scholar

71. Zhang Y, He G, Liu J. Comparison of medical costs and care of appendectomy patients between fee-for-service and set fee for diagnosis-related group systems in 20 Chinese hospitals. Southeast Asian J Trop Med Public Health. (2016) 47:1055–61.

PubMed Abstract | Google Scholar

72. Zhao C, Wang C, Shen C, Wang Q. Diagnosis-related group (DRG)-based case-mix funding system, a promising alternative for fee for service payment in China. Biosci Trends. (2018) 12:109–15. doi: 10.5582/bst.2017.01289

PubMed Abstract | Crossref Full Text | Google Scholar

73. Lai Y, Fu H, Li L, Yip W. Hospital response to a case-based payment scheme under regional global budget: the case of Guangzhou in China. Soc Sci Med. (2022) 292:114601. doi: 10.1016/j.socscimed.2021.114601

PubMed Abstract | Crossref Full Text | Google Scholar

74. Xiang X, Dong L, Qi M, Wang H. How does diagnosis-related group payment impact the health care received by rural residents? Lesson Learn China Public Health. (2024) 232:68–73. doi: 10.1016/j.puhe.2024.04.021

PubMed Abstract | Crossref Full Text | Google Scholar

75. National Healthcare Security Administration. Statistical Bulletin on the Development of the National Health Security Service, 2023. (2024). (Accessed January 05, 2025). Available online at: https://www.nhsa.gov.cn/art/2024/7/25/art_7_13340.html

Google Scholar

76. Qian W. Trends and explanations of medical technology progress in China. Sci Tech Progres Policy. (2020) 37:35–41. doi: 10.6049/kjjbydc.2019030395 (in Chinese).

Crossref Full Text | Google Scholar

77. Long H, Yang Y, Geng X, Mao Z, Mao Z. Changing characteristics of pharmaceutical prices in China under centralized procurement policy: a multi-intervention interrupted time series. Front Pharmacol. (2022) 13:944540. doi: 10.3389/fphar.2022.944540

PubMed Abstract | Crossref Full Text | Google Scholar

78. Wang Z, Wang K, Hua Y, Dong X, Zhang L. Impact and implications of national centralized drug procurement in China. Int J Clin Pharm. (2024) 46:1–6. doi: 10.1007/s11096-024-01767-1

PubMed Abstract | Crossref Full Text | Google Scholar