Changes in Hospitalization Expenditures and Length of Stay for Patients Underwent Hysterectomy and Adnexectomy for Carcinoma in situ and Nonmalignant Lesions Diagnosis-Related Group: Impact of Beijing’s Medicine–Consumables Linkage Reforms

Juanjuan Liang; Yuanqing Cai; Rong Fan; Yuwei Zhang; Yangyi Hu; Wenjuan Wang

doi:10.2147/RMHP.S620716

. 2026 Aug 24;19:620716. doi: 10.2147/RMHP.S620716

Changes in Hospitalization Expenditures and Length of Stay for Patients Underwent Hysterectomy and Adnexectomy for Carcinoma in situ and Nonmalignant Lesions Diagnosis-Related Group: Impact of Beijing’s Medicine–Consumables Linkage Reforms

Juanjuan Liang ¹, Yuanqing Cai ^2,^✉, Rong Fan ³, Yuwei Zhang ⁴, Yangyi Hu ⁵, Wenjuan Wang ⁶

PMCID: PMC13526913 PMID: 42676678

Abstract

Purpose

Hysterectomy and adnexectomy for carcinoma in situ and nonmalignant lesions (HAC) is a subgroup of the diagnosis-related groups (DRG). A comprehensive set of reforms in Beijing, referred to as the medicine–consumables linkage (MCL) reforms, is intended to improve the quality of medical services. This study aims to analyze the impact of the MCL reforms on the changes in hospitalization expenditures, out-of-pocket (OOP) expenditures, and length of stay (LOS) for patients underwent HAC.

Methods

We used an interrupted time series (ITS) model along with relevant data on HAC from January 2017 to December 2020 to analyze changes in hospitalization expenditures, OOP expenditures, and LOS for patients before and after the MCL reforms were implemented.

Results

The results showed that after the reforms average surgical expenditures per hospitalization increased by 75.34%, whereas average examination and medical consumable expenditures per hospitalization decreased by 17.51% and 12.63%, respectively. The proportion of OOP expenditures decreased by 14.26%, and the average LOS per hospitalization decreased by 21.86%. The ITS analysis showed that the average total, laboratory test, drug, and OOP expenditures per hospitalization decreased monthly after the reforms. The impact of the reforms on the subgroup without comorbidity and complications was even more significant and reached the expected goals of the reforms.

Conclusion

The hospitalization expenditure structure was optimized after the reforms, and the price of surgical services increased after the MCL reforms. The patients’ cost burden was reduced, and the LOS for patients declined after the reforms. This study provides evidence and reference points for comparing hospitalization expenditures and the LOS for patients in optimizing DRG programs in China and other countries with similar hospitalization systems.

Keywords: hysterectomy, diagnosis-related groups, out-of-pocket expenditures, length of stay, hospitalization expenditures structure

Introduction

Policy Background

To improve the quality of medical services for patients and promote high-quality development of China’s public hospitals, the Beijing municipal government issued the “Beijing medicine–consumables linkage comprehensive reforms implementation plan” on December 26, 2018, and on June 15, 2019 it began to implement a comprehensive set of medicine–consumables linkage (MCL) reforms covering over 3000 medical institutions in the city. Key components of these reforms include the following three aspects that comprise five “ones”: The first aspect involves “one reduction” and “one enhancement”, intended to reduce the price of the medical examination and laboratory test items performed with medical equipment; optimize the prices of medical services, such as surgery and traditional Chinese medicine; and highlight the value of medical workers’ technical labor. The second aspect is “one abolition” and “one procurement”, abolishing the mark-up on medical consumables and reducing the price of medical consumables, implementing joint procurement of medical consumables and volume purchasing of drugs, and making room for standardizing and adjusting the prices of medical service items. The third aspect is “one improvement”, which focuses on improving the patient’s experience with medical services, enhancing the patient’s sense of gain, promoting medical insurance reform, adding qualified medical service items to basic medical insurance coverage, and increasing the efforts to enact payment reforms. These objectives of MCL reforms may have an impact on the changes in hospitalization expenditures and LOS for patients underwent HAC DRG, as shown in Table 1.

Table 1.

MCL Reforms Objectives and Possible Trends in Indicators for Patients Underwent HAC

MCL Objectives	Objectives Explanation	HAC Indicators	Possible Trends
One reduction	Reducing the price of the medical examination and laboratory test items performed with medical equipment	AEEPH	Decrease
One reduction		ALTEPH	Decrease
One enhancement	Optimizing the prices of medical services, such as surgery and traditional Chinese medicine; and highlighting the value of medical workers’ technical labor	ASEPH	Increase
One enhancement		ATREPH	Increase
One abolition	Abolishing the mark-up on medical consumables and reducing the price of medical consumables, implementing joint procurement of medical consumables, and making room for standardizing and adjusting the prices of medical service items.	AMCEPH	Decrease
One procurement	Implementing volume purchasing of drugs, and making room for standardizing and adjusting the prices of medical service items.	ADEPH	Decrease
One improvement	Improving the patient’s experience with medical services, enhancing the patient’s sense of gain, promoting medical insurance reform, adding qualified medical service items to basic medical insurance coverage, and increasing the efforts to enact payment reforms.	AOOPEH	Decrease
		POOOPE	Decrease
		ALOSPH	Decrease

Open in a new tab

Abbreviations: MCL, medicine–consumables linkage; HAC, Hysterectomy and adnexectomy for carcinoma in situ and nonmalignant lesion; AEEPH, average examination expenditure per hospitalization; ALTEPH, average laboratory test expenditure per hospitalization; ATEPH, average total expenditure per hospitalization; ASEPH, average surgery expenditure per hospitalization; ATREPH, average treatment expenditure per hospitalization; AMCEPH, average medical consumable expenditure per hospitalization; ADEPH, average drug expenditure per hospitalization; AOOPEH, average out-of-pocket expenditure per hospitalization; POOOPE, proportion of out-of-pocket expenditures; ALOSPH, average length of stay per hospitalization.

DRG

DRG is a globally recognized and advanced payment method that helps control expenditures and reduce induced medical demands.^1–3 Patients are classified into diagnosis-related groups (DRG) based on age, disease diagnosis, comorbidities and complications (CC), treatment modality, severity of disease, and resource consumption. Health insurers are paid according to these DRG. DRG represent cases with similar clinical processes, such as disease diagnosis or surgical operations, and similar resource consumption. In the China Healthcare Security Diagnosis Related Groups (CHS-DRG) program, cases are grouped along two dimensions: clinical processes and resource consumption, where resource consumption refers to medical expenditures and length of stay (LOS) for patients. Consequently, Hospitalization expenditures out-of-pocket (OOP) expenditures, and LOS are the primary aspects of DRG payment system performance evaluation. Medical cost control is an important aspect of modernizing the governance and management of public hospitals, which can reduce the burden of medical costs and improve patients’ experience.⁴^,⁵ The OOP expenditures and LOS affect patients’ cost burden and experience of medical treatment.

HAC

Hysterectomy and adnexectomy for carcinoma in situ and nonmalignant lesion (HAC) is a subgroup of a DRG in the BJ-DRG (the Beijing version of the DRG system) and is further categorized into “with CC” (hereinafter referred to by the DRG code “NB33”) and “without CC” (hereinafter referred to by the DRG code “NB35”). Hospitalization volume for HAC is consistently in the top three of all DRG subgroups in Beijing, accounting for 6.12% of hospitalizations across all DRG subgroups between 2017 and 2020. Therefore, an analysis of hospitalization expenditures, OOP expenditures and LOS for patients underwent HAC is representative of the broader DRG system.

The incidence of hysterectomies is relatively high. One study showed that one in nine women undergo a hysterectomy in their lifetime.⁶ More than 600,000 hysterectomies are performed in the United States each year.⁷ An Indian study reported that the overall incidence of hysterectomy among women in North India is 11.59 per 1000 women per year.⁸ A survey conducted in the German city of Mainz and the region of Mainz–Bingen has observed that women are more likely to have a hysterectomy as they get older and that most of these procedures are performed to address benign diseases.⁹ An epidemiological study of over 9 million women across 31 provinces in China found that the age-standardized prevalence of hysterectomy among urban women in China was 2.36%, with a peak prevalence of 7.61% among women aged 55–59.¹⁰ A similar epidemiological survey in rural areas of China also showed that the prevalence of hysterectomy among women aged 40 and older was 5.01%. In terms of indications for surgery, uterine fibroids account for more than 70% of cases and are the leading cause of hysterectomy.¹¹ Between 2013 and 2022, the hysterectomy rate among patients with uterine fibroids at a certain Grade A tertiary hospital in China remained between 33% and 42%. Although the proportion of myomectomies rose year by year during the same period, hysterectomy continued to play a significant role.¹² Owing to the high incidence of hysterectomies, particularly for benign disease, the associated medical costs and LOS for patients have become a growing concern for the public. A retrospective cohort study of a tertiary care center found that among hysterectomy for benign disease, the 5-year indirect costs of vaginal hysterectomy were higher than those of laparoscopic hysterectomy.¹³ The LOS for patients underwent laparoscopic hysterectomy in Brazil’s largest city declined over the period from 2008 to 2018.¹⁴ A European study shown that hospitalization expenditures were reduced after the introduction of robot-assisted hysterectomy in a European setting.¹⁵ The cost burden and LOS for patients underwent HAC in China also deserves attention.

Studies related to the MCL reforms mainly evaluate the impact of inpatient and outpatient expenditures, emergency room expenditures, and expenditure structures overall.¹⁶^,¹⁷ There are few studies on changes in hospitalization expenditures by DRG under the impact of the MCL reforms. These studies have the fairly short time spans (mostly 1–2 years) and insufficient sample time after the reforms were implemented (mostly in 2019). Using an interrupted time series (ITS) model and incorporating longer time dimensions, this study analyzed the impact of MCL reforms on hospitalization expenditures, OOP expenditures and LOS for patients underwent HAC subgroup of the DRG. We analyzed changes in the cost burden and medical experience for patients, explored the effects of the MCL reforms on the process of implementing DRG-based payment, and proposed a pathway to improve the DRG payment program to provide a reference for the governance and management of public hospitals.

Methods

Study Data

The data for this study were obtained from the Medical Insurance Expense Settlement Information System of the Beijing Municipal Medical Insurance Bureau. They cover monthly aggregate data from six DRG pilot hospitals in Beijing across 108 DRG groups. The original variables include the number of hospitalizations, total expenditures, surgery expenditures, treatment expenditures, examination expenditures, laboratory test expenditures, drug expenditures, Western drug expenditures, Chinese drug expenditures, medical consumable expenditures, OOP expenditures, and LOS for patients. Under the DRG payment system, outlier cases are reimbursed through a separate outlier payment mechanism and are therefore excluded from routine DRG settlement data. Beijing was the first city in China to implement a DRG pilot program. In 2011, the Beijing Municipal Human Resources and Social Security Bureau and other departments jointly issued the “Notice on the development of disease-related groups (DRG) payment pilot work” (Beijing Human Social Medical issue [2011] No. 207). At that time, only six tertiary hospitals, namely, Peking University People’s Hospital, Peking University Third Hospital, Youyi Hospital, Chaoyang Hospital, Xuanwu Hospital, and Tiantan Hospital, offering 108 subgroups of DRG hospitalization treatments, were selected for the pilot study. In 2022, the Beijing Municipal Medical Insurance Bureau issued its “Notice on issues related to the development of China Healthcare Security Diagnosis Related Groups (CHS-DRG) reform” (Beijing Medical Insurance issue [2022] No. 10), which was implemented on March 15, 2022. 66 designated medical institutions were selected to start the actual DRG payment, including the six hospitals named above, and 172 designated medical institutions for data simulation. Until March 2022 only these six hospitals were DRG pilot hospitals in Beijing. From August 1, 2011, to the present they have always been designated as DRG payment hospitals. DRG payment have been in effect for over 10 years at these hospitals. To assess changes in hospitalization expenditures, OOP expenditures and LOS for patients underwent HAC before and after the Beijing’s MCL reform, we used data on the number of hospitalizations, hospitalization expenditures, OOP expenditures and LOS for patients underwent HAC with CC (Group NB33) and HAC without CC (Group NB35) for these six DRG pilot hospitals. The inclusion criteria were diagnosis codes for “HAC with CC” (Group NB33) and “HAC without CC” (Group NB35). The exclusion criteria were cases where information contained logical errors or was missing and could not be supplemented. After data cleaning, our dataset included 13,174 hospitalizations over the period from 2017 to 2020.

Indicators Selection

We selected average total expenditures per hospitalization (ATEPH), average surgery expenditures per hospitalization (ASEPH), average treatment expenditures per hospitalization (ATREPH), average examination expenditures per hospitalization (AEEPH), average laboratory test expenditures per hospitalization (ALTEPH), average drug expenditures per hospitalization (ADEPH) (including average Western drug expenditures per hospitalization (AWDEPH) and average Chinese drug expenditures per hospitalization (ACDEPH), and average medical consumable expenditures per hospitalization (AMCEPH) to represent hospitalization expenditures. We used average OOP expenditures per hospitalization (AOOPEH) and the proportion of OOP expenditures (POOOPE) to represent patients’ cost burden. The average LOS per hospitalization (ALOSPH) was used to represent the medical experience for patients, and the number of hospitalizations per month (HPM) represented the utilization of medical services.

Statistical Analysis

We used an ITS model to examine the baseline level and trend of the outcome variables before the MCL reforms and the changes in the level and trend after the MCL reforms.¹⁸ We constructed linear regression equations segmentally with point-in-time as the independent variable and hospitalization expenditures, OOP expenditures and LOS for patients as the outcome variables. In this study, the intervention point was June 2019, when the MCL reforms in Beijing were officially implemented. The pre‑intervention period comprised 29 months (January 2017 to May 2019), and the post‑intervention period comprised 19 months (June 2019 to December 2020). The regression equation is as follows:

Where, Inline graphic is the mean value of each outcome variable at time t specified as a given month. is a continuous variable that denotes the specific month during the study period, assumed a value of 1–48. , a dummy variable that denotes whether or not t is in the reforms implementation period, was set equal to 0 before the MCL reforms (January 2017-May 2019) and 1 after the MCL reforms (June 2019 to December 2020). Inline graphic a continuous variable that denotes the number of months after the intervention, was set equal to 0 before the MCL reforms (January 2017-May 2019) and assumed a value of 1–19 after the MCL reforms (June 2019 to December 2020). Considering the significant changes in hospitalization expenditures, average LOS per hospitalization, and number of hospitalizations per month due to the COVID-19 pandemic, we included a dummy variable Inline graphic in the regression model, set equal to 0 for the pre-pandemic period (January 2017 to December 2019) and 1 for the pandemic period (January 2020 to December 2020).¹⁹

Inline graphic denotes the pre-reforms baseline level, which is the estimate of the outcome variable at t=0. is the pre-reforms trend, which is the estimate of the slope of the outcome variable in the pre-reforms period. is the change in the level of the outcome variable after the MCL reforms, which indicates a short-term policy effect. Inline graphic is the change in the slope of the outcome variable after the MCL reforms, which is the difference between the slope of the pre-reforms period and the slope of the post-reforms, and indicates the long-run policy effect. is the trend after the reforms, which denotes the slope of the post-reforms period.²⁰ Inline graphic denotes the effect of COVID-19, and is a random error term.

The Durbin–Watson (DW) test, which is used to test for autocorrelation in a time series, revealed there was autocorrelation in our dataset. We used the Prais–Winsten estimation to eliminate the effect of autocorrelation, and the DW values for all of the adjusted models were close to 2, which can be considered as having almost no autocorrelation. To assess the robustness of our findings, we conducted two sensitivity analyses: incorporating Fourier terms (sin and cos pairs with a 12‑month period) to control for potential seasonality, and re‑estimating the model with a 1‑month lag to examine potential delays in the intervention effect. We used stata17 to conduct the statistical analyses.

Results

Descriptive Analysis

Changes in the mean value of hospitalization expenditures, OOP expenditures, LOS for patients, and hospitalizations per month before and after the MCL reforms are shown in Table 2. After the MCL reforms, average total expenditures per hospitalization increased. Of the various subcategories of expenditures, the average surgical expenditure per hospitalization increased significantly by 75.34%. In contrast, average examination and medical consumable expenditures per hospitalization decreased. The proportion of OOP expenditures, and average LOS per hospitalization decreased post-reforms. However, the number of hospitalizations per month increased from 250.27 to 314.78, indicating an increase in the use of medical services.

Table 2.

Descriptive Statistics for Variables Before and After the Medicine–Consumables Linkage Reforms

Variables	Pre-Reform (Mean ± SD)	Post-Reform (Mean ± SD)	P value	Change Rate (%)
ATEPH (¥)	6144.22 ± 395.71	7213.82 ± 911.92	<0.001	17.41
ASEPH (¥)	1690.65 ± 91.93	2964.43 ± 143.03	<0.001	75.34
ATREPH (¥)	209.58 ± 42.52	171.58 ± 63.78	0.016	−18.13
AEEPH (¥)	330.05 ± 39.20	272.25 ± 90.45	0.003	−17.51
ALTEPH (¥)	652.70 ± 95.65	860.18 ± 184.59	<0.001	31.79
ADEPH (¥)	1031.02 ± 108.95	1072.10 ± 172.15	0.315	3.98
AWDEPH (¥)	978.30 ± 107.05	1041.63 ± 162.52	0.109	6.47
ACDEPH (¥)	52.72 ± 10.67	30.47 ± 11.88	<0.001	−42.20
AMCEPH (¥)	1383.55 ± 157.47	1208.76 ± 274.68	0.007	−12.63
AOOPEH (¥)	1968.28 ± 96.59	1973.76 ± 153.25	0.879	0.28
POOOPE (%)	32.11 ± 1.79	27.53 ± 1.71	<0.001	−14.26
ALOSPH (days)	3.98 ± 0.37	3.11 ± 0.40	<0.001	−21.86
HPM	250.27 ± 49.57	314.78 ± 135.89	0.022	25.78

Open in a new tab

Note: SD is Standard deviation.

Abbreviations: ATEPH, average total expenditure per hospitalization; ASEPH, average surgery expenditure per hospitalization; ATREPH, average treatment expenditure per hospitalization; AEEPH, average examination expenditure per hospitalization; ALTEPH, average laboratory test expenditure per hospitalization; ADEPH, average drug expenditure per hospitalization; AWDEPH, average western drug expenditure per hospitalization; ACDEPH, average Chinese drug expenditure per hospitalization; AMCEPH, average medical consumable expenditure per hospitalization; AOOPEH, average out-of-pocket expenditure per hospitalization; POOOPE, proportion of out-of-pocket (OOP) expenditures; ALOSPH, average length of stay per hospitalization; HPM, hospitalizations per month.

ITSA Analysis

The results of the ITSA model analysis are shown in Table 3. The NB33 category results are shown in Figure 1, and the NB35 category results are shown in Figure 2.

Table 3.

Interrupted Time Series Model Results

Variables	DRG Subgroup	Baseline Level Pre-Reform β₀	Slope Pre-Reform β₁	Change in Level Post-Reform β₂	Change in Slope Post-Reform β₃	Slope Post-Reform β₁ +β₃	COVID-19 Effect β₄
ATEPH (¥)	NB33	7354.77***	67.02*	1859.12**	−274.21**	−207.19*	2229.55*
	NB33	(6531.76, 8177.78)	(16.44, 117.61)	(509.52, 3208.71)	(−463.73, −84.68)	(−390.44, −23.93)	(148.41, 4310.69)
	NB35	5575.12***	12.03	1018.98***	−179.10***	−167.06***	1891.68***
	NB35	(5266.43, 5883.80)	(−6.92, 30.99)	(515.88, 1522.09)	(−248.83, −109.36)	(−234.31, −99.82)	(1127.46, 2655.90)
ASEPH (¥)	NB33	1607.58***	12.57*	1292.45***	12.85	25.41	−79.28
	NB33	(1425.79, 1789.37)	(1.42, 23.72)	(998.60, 1586.29)	(−27.36, 53.05)	(−13.23, 64.05)	(−518.17, 359.62)
	NB35	1609.08***	3.51	947.12***	2.87	6.39	186.05
	NB35	(1511.51, 1706.66)	(−2.47, 9.50)	(789.40, 1104.85)	(−18.70, 24.45)	(−14.35, 27.13)	(−49.54, 421.64)
ATREPH (¥)	NB33	284.50***	7.39**	−135.59*	−12.03	−4.64	91.06
	NB33	(206.96, 362.04)	(2.63, 12.15)	(−261.67, −9.52)	(−29.43, 5.37)	(−21.41, 12.12)	(−99.45, 281.58)
	NB35	141.82***	2.49**	−53.44**	−15.23***	−12.74***	147.24***
	NB35	(118.58, 165.06)	(1.06, 3.92)	(−91.46, −15.42)	(−20.54, −9.92)	(−17.87, −7.61)	(88.96, 205.52)
AEEPH (¥)	NB33	427.26***	8.36*	83.23	−47.46**	−39.10***	337.21*
	NB33	(305.23, 549.29)	(0.86, 15.86)	(−116.71, 283.17)	(−75.48, −19.44)	(−66.18, −12.02)	(29.58, 644.84)
	NB35	287.24***	0.32	−35.35	−11.08**	−10.75**	89.92*
	NB35	(257.02, 317.45)	(−1.53, 2.18)	(−83.93, 13.24)	(−17.68, −4.48)	(−17.09, −4.42)	(18.04, 161.79)
ALTEPH (¥)	NB33	860.83***	16.51**	310.46*	−62.76**	−46.25*	498.79*
	NB33	(703.86, 1017.80)	(6.86, 26.16)	(52.57, 568.34)	(−99.17, −26.35)	(−81.49, −11.01)	(98.87, 898.70)
	NB35	498.49***	4.74*	171.32**	−36.21***	−31.48***	357.12***
	NB35	(431.87, 565.11)	(0.65, 8.82)	(63.67, 278.97)	(−50.94, −21.49)	(−45.63, −17.33)	(196.42, 517.81)
ADEPH (¥)	NB33	1478.19***	2.18	418.82*	−43.77	−41.59	329.69
	NB33	(1226.68, 1729.71)	(−13.22, 17.57)	(17.58, 820.06)	(−97.81, 10.28)	(−93.29, 10.11)	(−256.14, 915.51)
	NB35	933.70***	0.72	72.48	−28.19**	−27.47**	2.74**
	NB35	(846.11, 1021.29)	(−4.64, 6.08)	(−67.67, 212.63)	(−47.13, −9.25)	(−45.60, −9.33)	(73.91, 485.18)
AWDEPH (¥)	NB33	1414.41 ***	2.78	426.93*	−43.41	−40.63	323.44
	NB33	(1170.24, 1658.59)	(−12.17, 17.73)	(36.94, 816.92)	(−96.00, 9.19)	(−90.96, 9.70)	(−247.06, 893.93)
	NB35	870.92***	1.45	72.68	−26.28**	−24.84**	259.57**
	NB35	(786.30, 955.54)	(−3.74, 6.63)	(−62.81, 208.18)	(−44.60, −7.96)	(−42.38, −7.29)	(60.55, 458.59)
ACDEPH (¥)	NB33	64.26***	−0.65	−6.79	−0.43	−1.07	7.27
	NB33	(53.34, 75.18)	(−1.31, 0.02)	(−24.39, 10.80)	(−2.82, 1.97)	(−3.37, 1.23)	(−18.86, 33.39)
	NB35	62.33 ***	−0.70*	−1.11	−1.53	−2.23*	1.53
	NB35	(52.83, 71.83)	(−1.27, −0.12)	(−15.46, 13.25)	(−3.40, 0.34)	(−3.98, −0.47)	(−4.75, 34.36)
AMCEPH (¥)	NB33	1813.89 ***	1.49	−44.14	−33.02	−31.52	372.02
	NB33	(1558.64, 2069.14)	(−14.18, 17.17)	(−460.50, 372.23)	(−90.82, 24.79)	(−87.28, 24.24)	(−261.64, 1005.67)
	NB35	1347.55***	−1.90	19.42	−62.60***	−64.50***	633.08***
	NB35	(1241.61, 1453.48)	(−8.41, 4.61)	(−154.71, 193.55)	(−87.23, −37.97)	(−88.35, −40.66)	(362.56, 903.61)
AOOPEH (¥)	NB33	2210.22***	−0.79	228.12	−73.56**	−74.35**	905.51**
	NB33	(1998.54, 2421.91)	(−13.80, 12.22)	(−118.65, 574.90)	(−122.13, −24.99)	(−121.30, −27.41)	(372.27, 1438.76)
	NB35	1943.06***	−1.27	87.88	−21.36**	−22.64***	216.45**
	NB35	(1889.25, 1996.88)	(−4.58, 2.03)	(0.09, 175.67)	(−33.56, −9.17)	(−34.40, −10.88)	(82.81, 350.10)
POOOPE (%)	NB33	30.03***	−0.23***	−1.68	0.06	−0.17	2.40
	NB33	(28.19, 31.87)	(−0.34, −0.11)	(−4.67, 1.31)	(−0.36, 0.47)	(−0.57, 0.23)	(−2.12, 6.92)
	NB35	34.94***	−0.10*	−3.04**	0.39*	0.29*	−3.72*
	NB35	(33.51, 36.36)	(−0.19, −0.01)	(−5.29, −0.79)	(0.09, 0.69)	(0.00, 0.58)	(−6.95, −0.48)
ALOSPH (days)	NB33	5.05***	0.02	−0.21	−0.06	−0.04	−0.23
	NB33	(4.59, 5.51)	(−0.01, 0.05)	(−0.96, 0.55)	(−0.17, 0.04)	(−0.14, 0.06)	(−1.38, 0.92)
	NB35	4.40***	−0.05***	0.19	−0.03	−0.08**	0.51*
	NB35	(4.11, 4.70)	(−0.06, −0.03)	(−0.21, 0.59)	(−0.09, 0.02)	(−0.13, −0.03)	(0.01, 1.01)

Open in a new tab

Note: ***P < 0.001; **P < 0.01; *P < 0.05; The 95% confidence interval (CI) is within in brackets.

12 graphs of NB33 monthly expenses & stay length per hospitalization around June 2019 intervention. Twelve time-series graphs labeled a through l display trends in monthly average expenditures and length of stay per hospitalization for NB33 in Beijing. Each graph shares a common x-axis labeled time (month) ranging from 0 to 50. A vertical dashed line marks the intervention start at June 2019, approximately at month 30. Dots represent actual observed values and a segmented line represents predicted values, with one segment before and one after the intervention, capturing both level and slope changes. Graph a, labeled NB33 ATEPH, has a y-axis in yuan ranging roughly 6000 to 16000. Pre-intervention the predicted line rises gradually; post-intervention it rises sharply then declines. Graph b, labeled NB35 ASEPH, y-axis in yuan from roughly 1500 to 3500, shows a steady upward predicted trend continuing post-intervention. Graph c, labeled NB33 ATREPH, y-axis in yuan from roughly 200 to 700, shows a rising pre-trend with a notable step up then decline after intervention. Graph d, labeled NB33 AEEPH, y-axis in yuan from roughly 0 to 2000, shows a relatively flat trend throughout. Graph e, labeled NB33 ALTEPH, y-axis in yuan from roughly 500 to 3000, shows a rising pre-trend then decline post-intervention. Graph f, labeled NB33 ADEPH, y-axis in yuan from roughly 1000 to 2500, shows a stable pre-trend with a rise then decline post-intervention. Graph g, labeled NB33 AWDEPH, y-axis in yuan from roughly 1000 to 2500, shows a gradual rise pre-intervention and a sharp rise then decline post-intervention. Graph h, labeled NB33 AODEPH, y-axis in yuan from roughly 20 to 120, shows a consistent downward trend throughout both periods. Graph i, labeled NB33 AMCEPH, y-axis in yuan from roughly 500 to 2500, shows a stable pre-trend with a slight decline post-intervention. Graph j, labeled NB33 AOOPEH, y-axis in yuan from roughly 2000 to 5000, shows a flat pre-trend with a rise then sharp decline post-intervention. Graph k, labeled NB33 PQOPE, y-axis in percent from roughly 15 to 35, shows a declining pre-trend continuing post-intervention. Graph l, labeled NB33 ALOSPH, y-axis in days from roughly 3 to 7, shows a relatively stable trend across both periods with a slight post-intervention decline. Most expenditure panels show upward shifts or peaks immediately after June 2019 followed by declining slopes, while the length of stay panel l and the percent panel k show declining or stable patterns. — Trend in monthly average expenditures and LOS per hospitalization for NB33 in Beijing from 2017 to 2020.

**Notes**:(a)ATEPH, average total expenditure per hospitalization; (b)ASEPH, average surgery expenditure per hospitalization; (c)ATREPH, average treatment expenditure per hospitalization; (d)AEEPH, average examination expenditure per hospitalization; (e)ALTEPH, average laboratory test expenditure per hospitalization; (f)ADEPH, average drug expenditure per hospitalization; (g)AWDEPH, average western drug expenditure per hospitalization; (h)ACDEPH, average Chinese drug expenditure per hospitalization; (i)AMCEPH, average medical consumable expenditure per hospitalization; (j)AOOPEH, average out-of-pocket expenditure per hospitalization; (k)POOOPE, proportion of out-of-pocket expenditures; (l)ALOSPH, average length of stay per hospitalization.

Line plots showing NB35 monthly average expenditures and length of stay trends with June 2019 intervention. Image A showing title text ′Intervention starts: June 2019′. X-axis label: time (month), range 0 to 50. Y-axis label: NB35 AETPM (Y), range 5000 to 9000. Marks: dots labeled ′Actual′ and a line labeled ′Predicted′, plus a vertical dashed reference line at about 30 months. Predicted line rises slightly before 30 months, then shows a sharp rise to a peak near 8000 to 8500 around 38 to 40 months, then declines toward about 6500 to 7000 by 50 months. Image B showing title text ′Intervention starts: June 2019′. X-axis label: time (month), range 0 to 50. Y-axis label: NB35 AETPM (Y), range 1500 to 3500. Actual dots cluster near 1600 to 1900 before 30 months, then shift upward after 30 months with many points around 2600 to 3200. Predicted line is nearly flat to slightly rising before 30 months, then steps up after 30 months and trends slightly upward. Image C showing title text ′Intervention starts: June 2019′. X-axis label: time (month), range 0 to 50. Y-axis label: NB35 AETPM (Y), range 100 to 300. Predicted line increases before 30 months from about 140 to about 200, then drops after 30 months to near 120 to 150, then spikes near 200 around 40 months and declines toward about 130 to 150 by 50 months. Image D showing title text ′Intervention starts: June 2019′. X-axis label: time (month), range 0 to 50. Y-axis label: NB35 AETPM (Y), range 150 to 400. Predicted line is roughly flat near 300 before 30 months, then drops after 30 months to near 220 to 250, briefly rises near 300 around 38 to 40 months, then declines toward about 170 to 200 by 50 months. Image E showing title text ′Intervention starts: June 2019′. X-axis label: time (month), range 0 to 50. Y-axis label: NB35 AETPM (Y), range 400 to 1200. Predicted line rises before 30 months from about 500 to about 700, then drops after 30 months to near 600, rises to near 900 to 1000 around 38 to 40 months, then declines toward about 600 to 700 by 50 months. Image F showing title text ′Intervention starts: June 2019′. X-axis label: time (month), range 0 to 50. Y-axis label: NB35 AETPM (Y), range 500 to 1400. Predicted line is roughly flat near 900 before 30 months, then drops after 30 months to near 700 to 800, rises to near 1100 to 1200 around 38 to 40 months, then declines toward about 700 to 800 by 50 months. Image G showing title text ′Intervention starts: June 2019′. X-axis label: time (month), range 0 to 50. Y-axis label: NB35 ACDTPM (Y), range 600 to 1400. Predicted line is roughly flat near 1000 before 30 months, then dips after 30 months to near 900, rises to near 1200 around 38 to 40 months, then declines toward about 900 to 1000 by 50 months. Image H showing title text ′Intervention starts: June 2019′. X-axis label: time (month), range 0 to 50. Y-axis label: NB35 ACDTPM (Y), range 20 to 80. Predicted line declines before 30 months from about 70 to about 50, then continues downward after 30 months with a small bump near 40 months, reaching near 20 to 30 by 50 months. Image I showing title text ′Intervention starts: June 2019′. X-axis label: time (month), range 0 to 50. Y-axis label: NB35 ACDTPM (Y), range 600 to 2000. Predicted line is roughly flat near 1300 to 1400 before 30 months, then drops after 30 months to near 1100 to 1200, rises to near 1500 around 38 to 40 months, then declines toward about 900 to 1100 by 50 months. Image J showing title text ′Intervention starts: June 2019′. X-axis label: time (month), range 0 to 50. Y-axis label: NB35 ACDTPM (Y), range 1000 to 2500. Predicted line is roughly flat near 1700 to 1800 before 30 months, then drops after 30 months to near 1400 to 1500, rises to near 2000 around 38 to 40 months, then declines toward about 1300 to 1500 by 50 months. Image K showing title text ′Intervention starts: June 2019′. X-axis label: time (month), range 0 to 50. Y-axis label: NB35 ACDTPM (Y), range 25 to 40. Predicted line declines before 30 months from about 36 to about 33, then drops after 30 months to near 30, dips near 27 to 28 around 38 to 40 months, then rises slightly toward about 30 to 31 by 50 months. Image L showing title text ′Intervention starts: June 2019′. X-axis label: time (month), range 0 to 50. Y-axis label: NB35 ACDTPM (Y), range 2 to 6. Predicted line declines before 30 months from about 5.5 to about 3.5, then continues downward after 30 months with a small rise near 40 months, reaching near 2.5 to 3.0 by 50 months. Overall across A to L, each plot compares Actual dots to a Predicted line over time with a shared intervention reference at about 30 months and many panels show a level change and a change in slope after the intervention point. — Trend in monthly average expenditures and LOS per hospitalization for NB35 in Beijing from 2017 to 2020.

**Notes**:(a)ATEPH, average total expenditure per hospitalization; (b)ASEPH, average surgery expenditure per hospitalization; (c)ATREPH, average treatment expenditure per hospitalization; (d)AEEPH, average examination expenditure per hospitalization; (e)ALTEPH, average laboratory test expenditure per hospitalization; (f)ADEPH, average drug expenditure per hospitalization; (g) AWDEPH, average western drug expenditure per hospitalization; (h)ACDEPH, average Chinese drug expenditure per hospitalization; (i)AMCEPH, average medical consumable expenditure per hospitalization; (j)AOOPEH, average out-of-pocket expenditure per hospitalization; (k)POOOPE, proportion of out-of-pocket expenditures; (l)ALOSPH, average length of stay per hospitalization.

Total Expenditures and Structure

Total Expenditures

Average total expenditures per hospitalization for patient in the NB33 category increased by ¥67.02 (P = 0.011) per month before the MCL reforms, increased by ¥1,859.12 (P = 0.008) in the month of the MCL reforms, then decreased by ¥207.19 (P = 0.028) per month after the reforms, for an average decrease of ¥274.21 (P = 0.006) per month after the reforms were enacted compared with those before the reforms. Average total expenditures per hospitalization for patients in the NB35 category increased by ¥1,018.98 (P < 0.001) in the month of the reforms, then decreased by ¥167.06 (P < 0.001) per month after the reforms, a decline of ¥179.10 (P < 0.001) per month after the reforms compared with the period before the reforms (Table 3).

Surgery and Treatment Expenditures

Average surgery expenditures per hospitalization for patient in the NB33 category increased by 12.57 (P = 0.028) per month before the MCL reforms. Average surgery expenditures per hospitalization for patients in the NB33 and NB35 category rose by ¥1,292.45 (P < 0.001) and ¥947.12 (P < 0.001) respectively in the month of the MCL reforms (Table 3).

Average treatment expenditures per hospitalization for patients in the NB33 category increased by ¥7.39 (P = 0.003) per month before the reforms. Average treatment expenditures per hospitalization for patients in the NB35 category increased by ¥2.49 (P = 0.001) per month before the reforms. Average treatment expenditures per hospitalization for patients in the NB33 category decreased by ¥135.59 (P=0.036) in the month of the reforms. Average treatment expenditures per hospitalization for patients in the NB35 category decreased by ¥53.44 (P = 0.007) in the month of the reforms, and decreased by ¥12.74 (P < 0.001) per month after the reforms, a decrease of ¥15.23 (P < 0.001) per month after the reforms compared with the period before the reforms (Table 3).

Examination and Laboratory Test Expenditures

Average examination expenditures per hospitalization for patients in the NB33 category increased by ¥8.36 (P = 0.03) per month before the reforms. Average examination expenditures per hospitalization for patients in the NB33 and NB35 category decreased by ¥39.10 (P = 0.006) and ¥10.75 (P = 0.001) respectively per month after the reforms, a decline of ¥47.46 (P=0.001) and ¥11.08 (P = 0.002) respectively per month for the period after the reforms versus before they were enacted (Table 3).

Average laboratory test expenditures per hospitalization for patients in the NB33 category increased by ¥16.51 (P = 0.001) per month before the reforms. Average laboratory test expenditures per hospitalization for patients in the NB35 category increased by ¥4.74 (P = 0.024) per month before the reforms. Average laboratory test expenditures per hospitalization for patients in the NB33 and NB35 category rose by ¥310.46 (P= 0.019) and ¥171.32 (P = 0.003) respectively in the month the reforms were enacted, but decreased by ¥46.25 (P = 0.011) and ¥31.48 (P < 0.001) respectively per month thereafter, a decline of ¥62.76 (P = 0.001) and ¥36.21 (P < 0.001) respectively per month after the reforms compared with that before the reforms (Table 3).

Drug and Medical Consumables Expenditures

Average drug expenditures per hospitalization for patients in the NB33 category increased by ¥418.82 (P = 0.041) in the month the reforms were enacted. The average drug expenditure per hospitalization for patients in the NB35 category decreased by ¥27.47 (P = 0.004) per month after the reforms, a decline of ¥28.19 (P = 0.004) per month after the reforms versus that before the reforms (Table 3).

Average expenditures on Western drugs per hospitalization for patients in the NB33 category increased by ¥426.93 (P = 0.033) in the month the MCL reforms were implemented. Average Western drug expenditures per hospitalization for patients in the NB35 category decreased by ¥24.84 (P = 0.007) per month after the reforms, a decrease of ¥26.28 (P = 0.006) per month after the reforms compared with that before the reforms (Table 3).

The change in average expenditures for Chinese drugs per hospitalization for patients in the NB33 category before and after the MCL reforms was not statistically significant. For patients in the NB35 category, average expenditures for Chinese drugs per hospitalization decreased by ¥0.70 (P = 0.019) per month before the reforms, and decreased by ¥2.23 (P = 0.014) per month after the reforms (Table 3).

Changes in average expenditures for medical consumables per hospitalization for patients in the NB33 category before and after the reforms were not statistically significant. Average medical consumable expenditures per hospitalization for patients in the NB35 category decreased by ¥64.50 (P < 0.001) per month after the reforms, a decrease of ¥62.60 (P < 0.001) per month after the reforms were enacted (Table 3).

OOP Expenditures and LOS for Patients

Average OOP expenditures per hospitalization for patients in the NB33 and NB35 category decreased by ¥74.35 (P = 0.003) and ¥22.64 (P < 0.001) respectively per month after the reforms, a decline of ¥73.56 (P = 0.004) and ¥21.36 (P = 0.001) respectively per month after the reforms versus before the reforms (Table 3).

The proportion of OOP expenditures for patients in the NB33 category increased by 0.23% (P < 0.001) per month before the reforms, whereas the proportion of OOP expenditures for patients in the NB35 category decreased by 0.10% (P = 0.029) per month before the reforms. The proportion of OOP expenditures for patients in the NB35 category decreased by 3.04% (P = 0.001) in the month of the MCL reforms, and then increased by 0.29% (P = 0.048) per month after the reforms, an increase of 0.39% (P = 0.001) per month after the reforms compared with that before the reforms (Table 3).

The change in the average LOS per hospitalization for patients in the NB33 category was not statistically significant. The average days per hospitalization for patients in the NB35 category decreased by 0.05 days (P < 0.001) per month before the reforms and decreased by 0.08 days (P = 0.001) per month after the reforms (Table 3).

Robustness Check

The results showed that the direction, magnitude, and statistical significance of the key effect estimates were generally consistent with those from the main analysis (see Table 4 for detailed results).

Table 4.

Robustness Test Results

Variables	DRG Subgroup	Seasonal Adjustment			lag (1)
Variables	DRG Subgroup	Change in Level Post-Reform β₂	Change in Slope Post-Reform β₃	Slope Post-Reform β₁ +β₃	Change in Level Post-Reform β₂	Change in Slope Post-Reform β₃	Slope Post-Reform β₁ +β₃
ATEPH (¥)	NB33	1993.73**	−233.64	−165.49	1806.15**	−272.64**	−203.74*
	NB33	(105.78, 3881.68)	(−618.03, 150.75)	(−550.22, 219.24)	(426.36, 3185.94)	(−487.73, −57.54)	(−413.05, 5.57)
	NB35	1075.15***	−173.46**	−162.05**	1007.10***	−169.59***	−157.65***
	NB35	(514.87, 1635.42)	(−313.61, −33.32)	(−302.45, −21.66)	(510.82, 1503.38)	(−261.70, −77.47)	(−247.24, −68.06)
ASEPH (¥)	NB33	1303.92***	10.44	22.94	1287.88***	12.78	25.43
	NB33	(829.53, 1778.31)	(−41.42, 62.30)	(−28.86, 74.74)	(882.76, 1693.01)	(−32.11, 57.67)	(−17.57, 68.43)
	NB35	943.49***	2.94	6.53	942.45***	3.03	6.63
	NB35	(668.64, 1218.35)	(−20.38, 26.25)	(−16.53, 29.59)	(675.96, 1208.95)	(−25.16, 31.22)	(−21.72, 34.99)
ATREPH (¥)	NB33	−112.59	−11.9	−4.69	−131.53**	−12.45	−5.11
	NB33	(−254.36, 29.18)	(−37.07, 13.26)	(−29.63, 20.24)	(−262.29, −0.77)	(−30.48, 5.59)	(−22.56, 12.34)
	NB35	−43.08*	−14.86**	−12.54*	−51.11**	−13.96***	−11.57***
	NB35	(−92.69, 6.52)	(−28.10, −1.61)	(−25.87, 0.80)	(−97.71, −4.51)	(−22.71, −5.21)	(−20.12, −3.02)
AEEPH (¥)	NB33	102.34	−44.06	−35.74	75.72	−43.79*	−35.28
	NB33	(−161.07, 365.75)	(−112.64, 24.52)	(−104.61, 33.14)	(−133.38, 284.82)	(−91.22, 3.63)	(−82.24, 11.69)
	NB35	−30.4	−12.37**	−12.09*	−35.38*	−11.25***	−10.92***
	NB35	(−73.19, 12.39)	(−24.52, −0.22)	(−24.21, 0.04)	(−73.03, 2.26)	(−19.27, −3.22)	(−18.83, −3.00)
ALTEPH (¥)	NB33	323.30*	−52.51	−35.86	301.56**	−61.49***	−44.82**
	NB33	(−27.73, 674.33)	(−131.08, 26.06)	(−114.52, 42.79)	(71.05, 532.07)	(−99.82, −23.16)	(−82.19, −7.45)
	NB35	182.26***	−32.07**	−27.35**	170.79***	−35.85***	−31.11***
	NB35	(77.97, 286.54)	(−58.27, −5.87)	(−53.51, −1.20)	(78.79, 262.78)	(−55.99, −15.70)	(−50.92, −11.30)
ADEPH (¥)	NB33	486.39**	−24.41	−22.63	430.60**	−44.86*	−42.99*
	NB33	(86.01, 886.76)	(−91.64, 42.82)	(−88.72, 43.45)	(78.65, 782.56)	(−94.94, 5.21)	(−90.97, 4.99)
	NB35	97.4	−27.71*	−27.04*	70.78	−29.20***	−28.36***
	NB35	(−26.05, 220.85)	(−57.23, 1.81)	(−56.41, 2.34)	(−47.89, 189.44)	(−48.96, −9.43)	(−47.10, −9.62)
AWDEPH (¥)	NB33	491.44**	−24.38	−21.95	437.42**	−44.43*	−41.91*
	NB33	(99.80, 883.09)	(−90.92, 42.16)	(−87.48, 43.58)	(92.22, 782.62)	(−93.87, 5.01)	(−89.48, 5.66)
	NB35	96.25	−26.24*	−24.84*	71	−27.13***	−25.57***
	NB35	(−22.90, 215.39)	(−54.39, 1.91)	(−52.81, 3.12)	(−43.97, 185.97)	(−45.38, −8.88)	(−42.74, −8.40)
ACDEPH (¥)	NB33	−5.04	−0.03	−0.68	−6.79	−0.42	−1.07**
	NB33	(−21.53, 11.44)	(−1.88, 1.82)	(−2.08, 0.72)	(−20.59, 7.01)	(−1.78, 0.94)	(−1.95, −0.18)
	NB35	0.99	−1.45	−2.17**	−0.39	−2.03**	−2.75***
	NB35	(−9.49, 11.47)	(−3.36, 0.46)	(−4.08, −0.26)	(−10.94, 10.17)	(−3.98, −0.09)	(−4.66, −0.85)
AMCEPH (¥)	NB33	13.92	−29.24	−27.83	−36.75	−39.27	−37.65
	NB33	(−395.08, 422.92)	(−104.03, 45.55)	(−100.70, 45.03)	(−345.55, 272.05)	(−98.58, 20.03)	(−94.47, 19.16)
	NB35	59.58	−61.28**	−64.09**	28.83	−55.81***	−58.32***
	NB35	(−156.57, 275.73)	(−120.65, −1.91)	(−123.25, −4.92)	(−143.53, 201.19)	(−90.12, −21.49)	(−91.80, −24.83)
AOOPEH (¥)	NB33	281.56	−53.02	−53.89	222.05	−70.54*	−71.25*
	NB33	(−131.97, 695.09)	(−160.86, 54.82)	(−162.01, 54.23)	(−95.79, 539.88)	(−143.24, 2.16)	(−143.13, 0.64)
	NB35	104.02**	−16.42	−17.90*	92.59**	−20.25***	−21.70***
	NB35	(15.04, 193.01)	(−37.24, 4.39)	(−38.67, 2.87)	(2.46, 182.72)	(−34.23, −6.27)	(−35.39, −8.01)
POOOPE (%)	NB33	−1.55	0.17	−0.06	−1.66	0.06	−0.17
	NB33	(−4.90, 1.80)	(−0.24, 0.57)	(−0.46, 0.34)	(−4.28, 0.97)	(−0.27, 0.39)	(−0.47, 0.13)
	NB35	−3.33***	0.53***	0.44***	−3.21***	0.42***	0.33**
	NB35	(−5.35, −1.31)	(0.19, 0.88)	(0.11, 0.78)	(−5.42, −1.01)	(0.15, 0.70)	(0.06, 0.59)
ALOSPH (days)	NB33	−0.14	0	0.02	−0.2	−0.06*	−0.04
	NB33	(−1.04, 0.77)	(−0.14, 0.13)	(−0.11, 0.15)	(−0.92, 0.53)	(−0.13, 0.00)	(−0.11, 0.02)
	NB35	0.25**	−0.05*	−0.09***	0.2	−0.06**	−0.10***
	NB35	(0.02, 0.49)	(−0.10, 0.00)	(−0.14, −0.04)	(−0.04, 0.44)	(−0.11, −0.01)	(−0.15, −0.06)

Open in a new tab

Note: ***P < 0.01; **P < 0.05; *P < 0.1; The 95% confidence interval (CI) is within in brackets.

Discussion

Total Expenditures and Structure

A descriptive statistical analysis showed that after the MCL reforms, average surgery expenditures per hospitalization increased significantly by 75.34%, while average examination expenditures per hospitalization declined significantly by 17.51%, average expenditures on Chinese drugs per hospitalization declined significantly by 42.2%, and average medical consumable expenditures per hospitalization declined significantly by 12.63%, indicating that the structure of hospitalization expenditures could be optimized and that the price of surgical services was enhanced by the reforms. These findings are similar to the results obtained in other studies.¹⁶^,¹⁷

Immediately after the reforms, average total expenditures per hospitalization increased immediately. As medical technology improves, patients live longer and suffer less, which can lead to an increase in medical expenditures.²¹ Therefore, controlling medical expenditure is only one objective of the MCL reforms; it is equally, if not more, important to optimize the structure of medical expenditures.

Surgery, Treatment, Examination and Laboratory Test Expenditures

Average surgery expenditures per hospitalization have increased significantly from ¥1,690.65 pre-reforms to ¥2,964.43 post-reforms and increased immediately after the MCL reforms. After the MCL reforms, the original purpose of “one enhancement” in the price of surgery achieved.

After the reforms, average treatment expenditures per hospitalization decreased immediately for both NB33 and NB35 patients and then decreased from ¥209.58 pre-reforms to ¥171.58 post-reforms. The MCL reforms focuses on increasing surgery expenditures in accordance with its requirements; however, treatment expenditures, which also reflect the value of medical technology, are not emphasized, and the reforms do not set out clear requirements for such expenditures. The next set of reforms should increase the proportion of treatment expenditures within the DRG payment scheme for patients undergoing HAC appropriately to further optimize the price of medical services.²²

After the MCL reforms, the “one reduction” requirement of the cost of medical examinations associated with medical equipment met. After the MCL reforms, average examination expenditures per hospitalization decreased from ¥330.05 pre-reforms to ¥272.25 post-reforms, and the monthly trend declined for both NB33 and NB35 patients compared with the pre-reforms trend. Average laboratory test expenditures per hospitalization increased immediately after the reforms, but the monthly trend decreased thereafter, with a downward change compared with the pre-reforms trend. This indicates that controlling laboratory test expenditures, while initially effective, is still problematic at the implementation pathway level. Compared with other expenditure-related reform measures, clear guidance on goals for laboratory test expenditures has not been provided, and lowering standard costs has not been clearly stated as a goal. More detailed reform measures should be proposed in the future; for example, specific price adjustments should be made with respect to laboratory tests.²³ Mutual recognition of test results among different medical institutions should be further promoted to minimize the duplication of examinations and laboratory tests.²⁴

Drug and Medical Consumables Expenditures

After the MCL reforms were implemented, through abolishing mark-up on medical consumables, prices of medical consumables reduced, reflecting the effect of the medical consumables “one abolition” aspect of the MCL reforms. After the MCL reforms were implemented, average medical consumables expenditures per hospitalization decreased from ¥1,383.55 pre-reforms to ¥1,208.76 post-reforms. The monthly trend for NB35 patients declined compared with the pre-reforms trend. Average expenditures on Chinese drugs per hospitalization for NB35 patients decreased from ¥52.72 pre-reforms to ¥30.47 post-reforms, and the monthly trend decreased. Through joint procurement of medical consumables and volume purchasing of drugs, prices of both medical consumables and Chinese drugs reduced, reflecting the effect of the “one procurement” aspect of the reforms, which supports standardizing and adjusting the prices of medical services. After the reforms were implemented, average expenditures for Western drugs per hospitalization increased immediately for NB33 patients, while the monthly trend for NB35 patients decreased versus the pre-reforms trend. The descriptive statistics in Table 2 shows no significant difference in average Western drug expenditures per hospitalization before and after the reforms. The MCL reforms had a limited effect in reducing Western drug expenditures. This may be related to the fact that drug mark-ups were abolished in the previous phase of the comprehensive medicine–drug separation reform in Beijing, and the decline in Western drug expenditures has already yielded some results. After that reform was enacted, average expenditures per hospitalization for Western and Chinese drugs were ¥1,041.63 and ¥30.47 respectively. Chinese drug expenditures accounted for a relatively small proportion of total drug expenditures for HAC; therefore, reducing expenditures on Chinese drugs cannot have a significant impact on lowering overall drug expenditures. Follow-up should focus on changes in Western drug expenditures and trying to contain these costs within a reasonable range.²⁵

The 2017 medicine–drug separation comprehensive reform in Beijing abolished all drug mark-ups, and the proportion of drugs in secondary and tertiary public hospitals was reduced to 30%, excluding Chinese drug tablets. Chinese drugs are in a special position in health care reforms. Therefore, a different perspective is needed to focus on the effects of reforms of Chinese drugs.²⁶

Cost Burden and Hospitalization LOS for Patients

The patients’ cost and time burden were reduced after the MCL reforms. The patients’ experience with medical services and patients’ sense of gain were improved after the MCL reforms, thus achieving the goal of “one improvement” in medical services.

After the MCL reforms, the proportion of OOP expenditures and the average OOP expenditures declined per hospitalization. After the MCL reforms, the monthly trend in average OOP expenditures per hospitalization for NB33 and NB35 patients decreased, showing a decline versus the pre-reforms trend. The proportion of OOP expenditures decreased from 32.11% pre-reforms to 27.53% post-reforms for NB33 and NB35 patients, and decreased immediately by 3.04% for NB35 patients. Health insurance now covers more costs for patients, and the reform of health insurance payment methods has been promoted further.

Since 2011, the pilot reform program for DRG-based payment has clearly stated that OOP expenditure should be strictly controlled, the proportion of OOP expenditures should not be higher than the level during the same period in the previous year in the pilot hospitals, and that excess expenditures would be deducted from the medical insurance fund at the time of settlement at the end of the year. Such measures have greatly reduced the cost burden on patients and minimized unnecessary OOP medical items and drugs. Driven by the MCL reforms, the proportion of OOP expenditures and the average OOP expenditures per hospitalization for patients underwent HAC have been further reduced. Many common clinical avenues and treatments result in OOP expenditures. How to limit the amount of auxiliary drugs included in OOP expenditure, and which drugs classified as rescue and nationally essential should be included in non-OOP expenditures must be explored and summarized by doctors in the process of developing DRG reforms.²⁷

After the MCL reforms, the average LOS per hospitalization declined. The average LOS per hospitalization for NB33 and NB35 patients decreased to from 3.98 days pre-reforms to 3.11 days post-reforms and the monthly trend of the average LOS per hospitalization for NB35 patients declined by 0.08 days. Thus, MCL reforms have helped to shorten patients’ wait time for hospital beds, improved their experience of medical treatment, increased hospital bed utilization rate, and helped to optimize the allocation of medical resources.²⁸^,²⁹ Some hospitals in China have begun to promote day surgeries, using the proportion of day surgeries to elective surgeries as the performance evaluation index for tertiary public hospitals. These supporting measures for reform have played an important role in reducing the average LOS per hospitalization.³⁰

The average LOS per hospitalization, an important indicator of resource consumption, has declined gradually owing to various types of healthcare reforms and continuous improvement of the efficiency of healthcare services. The average LOS per hospitalization for NB33 and NB35 patients in the 2011 DRG program were 10.96 and 8.25 days, respectively. Driven by the MCL reforms, the average LOS per hospitalization for patients underwent HAC have been further reduced.

Some studies show that the average LOS per hospitalization is a key factor influencing the increase in hospitalization expenditures.³¹ Reducing the average LOS per hospitalization can help to standardize treatment regimens and thus reduce total inpatient hospitalization expenditures.³² Medical institutions should try to improve the efficiency of medical services, reduce ineffective medical treatment, and shorten the average LOS per hospitalization. This will require major improvements in technology, experience, and doctors’ responsibility for reducing the average LOS per hospitalization.³³^,³⁴

With or without CC

The effect of reforms on patients with and without CC was slightly different, and we find that the effect of the MCL reforms on those without CC is more significant. After the reforms were enacted, the monthly trend of the average expenditures for treatment, drugs, Western drugs, and medical consumables per hospitalization for patients without CC decreased compared with the pre-reforms trend. After the MCL reforms, the monthly trends of average expenditures for treatment, drugs, Western drugs, Chinese drugs, medical consumable, and the LOS per hospitalization for patients without CC decreased. Furthermore, the proportion of OOP expenditures for patients without CC decreased immediately. The corresponding effects of the reforms for patients with CC were not as notable There are several possible reasons for this phenomenon. First, the baseline level for patients with CC was higher than that for patients without CC before the reforms (see Table 3); therefore, declines in hospitalization expenditures and days were more difficult to achieve. Second, the level of treatment complexity for patients with CC tends to be greater than that for patients without CC, and thus the impact of the reforms is less certain.³⁵ It has been shown that comorbidities and complications are risk factors that can increase total expenditures when treating a disease.³⁶ As a disease with CC has more complex conditions and doctors may revise their regimens during treatment, these cases are more likely to have side effects during and after treatment, resulting in longer treatment cycles, more treatment regimens and drugs, and higher treatment expenditures, thus requiring greater focus. “Comorbidity and complication” is important factors for DRG groupings.³⁷ We suggested that more rational DRG programs should be developed to full account for the impact of CC on hospitalization expenditures and length of stay.³⁸

Policy Implications

While DRG-based payments can incentivize efficiency,^39–41 their continued optimization will require collaborative efforts from health insurers, financial department, healthcare institutions and doctors.⁴² The MCL reforms appear to have facilitated this process by reducing cost burdens and LOS, but further refinement of DRG grouping and physician engagement mechanisms remains warranted. In the tripartite relationship among payers, suppliers, and consumers of healthcare services, doctors, as the main suppliers of healthcare services, play a central role; thus, all healthcare reforms rely on the motivation of hospitals and doctors.⁴³

Limitations

This study has certain limitations. First, there have been several rounds of medical reforms in Beijing; therefore, the potential impact of reforms other than the MCL reforms on the results presented here cannot be ruled out. Second, the COVID-19 pandemic occurred during the study period, and it is possible that its effect could not be completely eliminated despite the use of the control variable in the regression. Third, the results of this study are not fully applicable to other DRG subgroups although they can be used as a reference. Future research efforts should include multiple DRG subgroups and study the differences between them.

Conclusions

This study has analyzed the impact of Beijing’s MCL reforms on the changes in hospitalization expenditures, OOP expenditures and length of stay for patients underwent HAC. The results have shown that the performance of DRG for patients underwent HAC met the expected goals of the MCL reforms. The hospitalization expenditure structure of DRG for patients underwent HAC has been optimized, and the price of surgical services increased after the MCL reforms. The cost burden for patients is now lower than that before the reforms. The reforms have reduced the LOS for patients. This study provides a framework for reducing hospitalization expenditures, OOP expenditures and LOS for patients undergoing hysterectomy and for optimizing DRG programs in China and other countries with similar systems.

Acknowledgments

The authors thank the Beijing Municipal Medical Insurance Bureau for the data and Enago (https://orders.enago.cn) for English language editing.

Funding Statement

This study was supported by the National Social Science Foundation of China (Grant number:23BGL253, 22CGL052); and Chinese Academy of Social Sciences “Qingqi Program” Project (Grant number: 2024QQJH010).

Abbreviations

HAC, Hysterectomy and adnexectomy for carcinoma in situ and nonmalignant lesions; DRG, Diagnosis-related groups; MCL, Medicine–consumables linkage; OOP, Out-of-pocket; LOS, Length of stay; ITS, Interrupted time series model; CC, Comorbidities and complications; CHS-DRG, China Healthcare Security Diagnosis Related Groups; ATEPH, Average total expenditures per hospitalization; ASEPH, Average surgery expenditures per hospitalization; ATREPH, Average treatment expenditures per hospitalization; AEEPH, Average examination expenditures per hospitalization; ALTEPH, Average laboratory test expenditures per hospitalization; ADEPH, Average drug expenditures per hospitalization; AWDEPH, Average Western drug expenditures per hospitalization; ACDEPH, Average Chinese drug expenditures per hospitalization; AMCEPH, Average medical consumable expenditures per hospitalization; AOOPEH, Average out-of-pocket expenditures per hospitalization; POOOPE, Proportion of out-of-pocket expenditures; ALOSPH, Average length of stay per hospitalization; HPM, Hospitalizations per month.

Data Sharing Statement

The data that support the findings of this study were obtained from a third party and are not publicly available.

Ethics Approval and Consent to Participate

This study was reviewed by the Medical Ethics Committee of Peking University Cancer Hospital and granted exemption from ethical review. As per the “Measures for Ethical Review of Life Science and Medical Research Involving Human Subjects” (February 18, 2023), the research qualifies for exemption under item [1 or 2] of Article 32. Specifically, this study does not involve subject protection and ethic. Therefore, no additional ethical approval was required from our institution.

Author Contributions

All authors made a significant contribution to the work reported, whether that is in the conception, study design, execution, acquisition of data, analysis and interpretation, or in all these areas; took part in drafting, revising or critically reviewing the article; gave final approval of the version to be published; have agreed on the journal to which the article has been submitted; and agree to be accountable for all aspects of the work.

Disclosure

All authors declare no competing interests in this work.

References

1.Aragón MJ, Chalkley M, Kreif N. The long-run effects of diagnosis related group payment on hospital lengths of stay in a publicly funded health care system: evidence from 15 years of micro data. Health Econ. 2022;31(6):956–16. doi: 10.1002/hec.4479 [DOI] [PMC free article] [PubMed] [Google Scholar]
2.Kim G-D, Hun PJ. Effects of new diagnosis-related group-based payment on efficiency of public hospitals. Korean J Public Admin. 2018;56(4):33–57. doi: 10.24145/kjpa.56.4.2 [DOI] [Google Scholar]
3.Sorensen MLK, Burau V. Why we need to move beyond diagnosis-related groups and how we might do so. J Health Serv Res Policy. 2016;21(1):64–66. doi: 10.1177/1355819615586444 [DOI] [PubMed] [Google Scholar]
4.Li LL, Du TT, Hu YP. The effect of different classification of hospitals on medical expenditure from perspective of classification of hospitals framework: evidence from China. Cost Eff Resour Allocation. 2020;18(1):35. doi: 10.1186/s12962-020-00229-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
5.Shu Z, Liu YL, Li ML, Li J. The effects of health system reform on medical services utilization and expenditures in China in 2004–2015. International Health. 2021;13(6):640–647. doi: 10.1093/inthealth/ihab041 [DOI] [PMC free article] [PubMed] [Google Scholar]
6.Novetsky AP, Boyd LR, Curtin JP. Trends in bilateral oophorectomy at the time of hysterectomy for Benign disease. Obstet Gynecol. 2011;118(6):1280–1286. doi: 10.1097/AOG.0b013e318236fe61 [DOI] [PubMed] [Google Scholar]
7.Green GT. Hysterectomy and Self-Esteem Among African American Women. 2022.
8.Rajkumari S, Chaudhary V, Kasaudhan S, Saraswathy KN. Incidence and determinants of hysterectomy among North Indian women: an 8-year follow-up study. Front Public Health. 2022;101065081. doi: 10.3389/fpubh.2022.1065081 [DOI] [PMC free article] [PubMed] [Google Scholar]
9.Tanaka LF, Schoffer O, König J, Weyer-Elberich V, Blettner M, Klug SJ. Changes in the probability of hysterectomy in the city of Mainz and Mainz-Bingen region, Germany. BMC Public Health. 2023;23(1):84. doi: 10.1186/s12889-022-14916-w [DOI] [PMC free article] [PubMed] [Google Scholar]
10.Bi Y, Ma Y, Zhang J, et al. Prevalence of hysterectomy in urban China and associations between metabolic disorders and hysterectomy: a multicenter population-based study involving over 9 million women. BMC Med. 2026;24(1). doi: 10.1186/s12916-026-04737-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
11.Liu F, Pan Y, Liang Y, et al. The epidemiological profile of hysterectomy in rural Chinese women: a population-based study. BMJ Open. 2017;7(6):e015351. doi: 10.1136/bmjopen-2016-015351 [DOI] [PMC free article] [PubMed] [Google Scholar]
12.Ye S, Wang W, Li Y, Shen W. Surgical management of uterine fibroids: a large retrospective study. J Minim Invasive Gynecol. 2025;32(9):823–829. doi: 10.1016/j.jmig.2025.05.011 [DOI] [PubMed] [Google Scholar]
13.Martínez-Maestre MA, Jódar-Sánchez F, Calderón-Cabrera AM, González-Cejudo C, Silván-Alfaro JM, Melero-Cortés LM. Healthcare and indirect cost of the laparoscopic vs vaginal approach in benign hysterectomy. JSLS. 2022;26(3):e202200048. doi: 10.4293/jsls.2022.00048 [DOI] [PMC free article] [PubMed] [Google Scholar]
14.Augusto CF, Caraça DB, Podgaec S. Epidemiological analysis of hysterectomies performed at the public health system in the largest Brazilian city. Rev Assoc Med Bras. 2021;67(7):937–941. doi: 10.1590/1806-9282.20210093 [DOI] [PubMed] [Google Scholar]
15.Dumont S, Van Trappen P. The clinical and financial impact of introducing robotic-assisted hysterectomy in a tertiary referral centre: a direct cost analysis of consecutive hysterectomies during a decade. Int J Med Robot. 2022;18(1):e2343. doi: 10.1002/rcs.2343 [DOI] [PubMed] [Google Scholar]
16.Gao L, Shi L, Meng Q, Kong X, Guo M, Lu F. Effect of healthcare system reforms on public hospitals’ revenue structures: evidence from Beijing, China. Soc Sci Med. 2021;283:114210. doi: 10.1016/j.socscimed.2021.114210 [DOI] [PubMed] [Google Scholar]
17.Liu M, Jia M, Lin Q, Zhu J, Wang D. Effects of Chinese medical pricing reform on the structure of hospital revenue and healthcare expenditure in county hospital: an interrupted time series analysis. BMC Health Serv Res. 2021;21:1–9. doi: 10.1186/s12913-021-06388-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
18.Turner SL, Korevaar E, Cumpston MS, Kanukula R, Forbes AB, McKenzie JE. Effect estimates can be accurately calculated with data digitally extracted from interrupted time series graphs. Res Synth Meth. 2023;14(4):622–638. doi: 10.1002/jrsm.1646 [DOI] [PMC free article] [PubMed] [Google Scholar]
19.Lagarde M. How to do (or not to do) … Assessing the impact of a policy change with routine longitudinal data. Health Policy Plan. 2012;27(1):76–83. doi: 10.1093/heapol/czr004 [DOI] [PubMed] [Google Scholar]
20.Li L, Jiang JN, Xiang L, Wang XF, Zeng L, Zhong ZD. Impact of critical illness insurance on the burden of high-cost rural residents in central china: an interrupted time series study. Int J Environ Res Public Health. 2019;16(19):3528. doi: 10.3390/ijerph16193528 [DOI] [PMC free article] [PubMed] [Google Scholar]
21.Jakovljevic M, Chang H, Pan J, et al. Successes and challenges of China’s health care reform: a four-decade perspective spanning 1985—2023. Cost Eff Resour Allocation. 2023;21(1):59. doi: 10.1186/s12962-023-00461-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
22.Xiong WY, Deng YF, Yang YL, Zhang YM, Pan JY. Assessment of medical service pricing in China’s healthcare system: challenges, constraints, and policy recommendations. Front Public Health. 2021;9787865. doi: 10.3389/fpubh.2021.787865 [DOI] [PMC free article] [PubMed] [Google Scholar]
23.Mahomed O, Cassim N. Appropriateness of laboratory expenditure for primary health care facilities across South Africa. African J Primary Health Care Family Med. 2023;15(1):1–8. doi: 10.4102/phcfm.v15i1.3740 [DOI] [PMC free article] [PubMed] [Google Scholar]
24.Van Wilder A, Bruyneel L, De Ridder D, et al. Is a hospital quality policy based on a triad of accreditation, public reporting and inspection evidence-based? A narrative review. Int J Qual Health Care. 2021;33(2). doi: 10.1093/intqhc/mzab085 [DOI] [PubMed] [Google Scholar]
25.Li Y. Generic price regulation and drug expenditures: evidence from Canada. Value Health. 2023;26(6):833–840. doi: 10.1016/j.jval.2023.01.008 [DOI] [PubMed] [Google Scholar]
26.Jin Y, Kim Y. Dietary advice in chronic care: comparing traditional Chinese and western medicine practiced in mainland China. Soc Sci Med. 2022;292114621. doi: 10.1016/j.socscimed.2021.114621 [DOI] [PubMed] [Google Scholar]
27.Gujar A, Gulecha V, Zalte A. Drug utilization studies using WHO prescribing indicators from India: a systematic review. Health Policy Technol. 2021;10(3):100547. doi: 10.1016/j.hlpt.2021.100547 [DOI] [Google Scholar]
28.Chen YJ, Zhang XY, Yan JQ, Qian MC, Ying XH. Impact of diagnosis-related groups on inpatient quality of health care: a systematic review and meta-analysis. Inquiry-J Health Care Organ Prov Finan. 2023;6000469580231167011. doi: 10.1177/00469580231167011 [DOI] [PMC free article] [PubMed] [Google Scholar]
29.Khammarnia M, Bagherbarahouei F, Ansari-Moghaddam A. Length of hospital stay and its related factors in Iran: a systematic review and meta-analysis. Iran J Public Health. 2023;52(9):1877–1888. doi: 10.18502/ijph.v52i9.13570 [DOI] [PMC free article] [PubMed] [Google Scholar]
30.Oberhollenzer F, Salzburger S, Auer HM, et al. Average length of stay after a cardiology day hospital stay compared to a normal ward: an analysis by using propensity score matching. Wien Klin Wochenschr. 2022;134(Suppl 2):194–195. [Google Scholar]
31.Bala C, Rusu A, Ciobanu D, Roman G. Length of hospital stay, hospitalization costs, and their drivers in adults with diabetes in the romanian public hospital system. Int J Environ Res Public Health. 2022;19(16):10035. doi: 10.3390/ijerph191610035 [DOI] [PMC free article] [PubMed] [Google Scholar]
32.Allaudeen N, Vashi A, Breckenridge JS, et al. Using lean management to reduce emergency department length of stay for medicine admissions. Qual Manag Health Care. 2017;26(2):91–96. doi: 10.1097/qmh.0000000000000132 [DOI] [PubMed] [Google Scholar]
33.Zhai CX, Bai LL, Xu Y, et al. Temperature variability associated with respiratory disease hospitalisations, hospital stays and hospital expenses the warm temperate sub-humid monsoon climate. Public Health. 2023;225:206–217. doi: 10.1016/j.puhe.2023.10.013 [DOI] [PubMed] [Google Scholar]
34.Golinelli D, Sanmarchi F, Toscano F, Bucci A, Nante N. Analyzing the 20-year declining trend of hospital length-of-stay in European countries with different healthcare systems and reimbursement models. Int J Health Econom Manag. 2024. doi: 10.1007/s10754-024-09369-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
35.Kim S, Jung C, Yon J, et al. A review of the complexity adjustment in the Korean Diagnosis-Related Group (KDRG). Health Inform Manag J. 2020;49(1):62–68. doi: 10.1177/1833358318795804 [DOI] [PubMed] [Google Scholar]
36.Maréchal F, Thomas L. The impact of medical complications on optimal hospital payment. J Public Econ Theory. 2021;23(6):1144–1173. doi: 10.1111/jpet.12530 [DOI] [Google Scholar]
37.Gluckman TJ, Spinelli KJ, Wang MS, et al. Trends in diagnosis related groups for inpatient admissions and associated changes in payment from 2012 to 2016. JAMA Network Open. 2020;3(12):e2028470. doi: 10.1001/jamanetworkopen.2020.28470 [DOI] [PMC free article] [PubMed] [Google Scholar]
38.Cook A, Averett S. Do hospitals respond to changing incentive structures? Evidence from medicare’s 2007 DRG restructuring. J Health Econ. 2020;73102319. doi: 10.1016/j.jhealeco.2020.102319 [DOI] [PMC free article] [PubMed] [Google Scholar]
39.Meng ZL, Ma YN, Song SH, et al. Economic implications of chinese diagnosis-related group-based payment systems for critically ill patients in ICUs. Crit Care Med. 2020;48(7):E565–E573. doi: 10.1097/ccm.0000000000004355 [DOI] [PubMed] [Google Scholar]
40.Baker LC, Lamiraud K. Adoption of hospital diagnosis-related group financing in Switzerland and the availability of computed tomography scanners. Health Econ. 2022;31(12):2537–2557. doi: 10.1002/hec.4594 [DOI] [PubMed] [Google Scholar]
41.Ma YF, Wang W. Original The impact of diagnosis related group payment on the performance of public hospitals. Am J Transl Res. 2021;13(6):6796–6801. [PMC free article] [PubMed] [Google Scholar]
42.Yu LH, Lang JJ. Diagnosis-related Groups (DRG) pricing and payment policy in China: where are we? Hepatob Surg Nutr. 2020;9(6):771–773. doi: 10.21037/hbsn-2020-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
43.Aktas P. Physician perspectives on the implications of the diagnosis-related groups for medical practice in Turkey: a qualitative study. Int J Health Plann Manage. 2022;37(3):1769–1780. doi: 10.1002/hpm.3445 [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Data Availability Statement

The data that support the findings of this study were obtained from a third party and are not publicly available.

[cit0001] 1.Aragón MJ, Chalkley M, Kreif N. The long-run effects of diagnosis related group payment on hospital lengths of stay in a publicly funded health care system: evidence from 15 years of micro data. Health Econ. 2022;31(6):956–16. doi: 10.1002/hec.4479 [DOI] [PMC free article] [PubMed] [Google Scholar]

[cit0002] 2.Kim G-D, Hun PJ. Effects of new diagnosis-related group-based payment on efficiency of public hospitals. Korean J Public Admin. 2018;56(4):33–57. doi: 10.24145/kjpa.56.4.2 [DOI] [Google Scholar]

[cit0003] 3.Sorensen MLK, Burau V. Why we need to move beyond diagnosis-related groups and how we might do so. J Health Serv Res Policy. 2016;21(1):64–66. doi: 10.1177/1355819615586444 [DOI] [PubMed] [Google Scholar]

[cit0004] 4.Li LL, Du TT, Hu YP. The effect of different classification of hospitals on medical expenditure from perspective of classification of hospitals framework: evidence from China. Cost Eff Resour Allocation. 2020;18(1):35. doi: 10.1186/s12962-020-00229-5 [DOI] [PMC free article] [PubMed] [Google Scholar]

[cit0005] 5.Shu Z, Liu YL, Li ML, Li J. The effects of health system reform on medical services utilization and expenditures in China in 2004–2015. International Health. 2021;13(6):640–647. doi: 10.1093/inthealth/ihab041 [DOI] [PMC free article] [PubMed] [Google Scholar]

[cit0006] 6.Novetsky AP, Boyd LR, Curtin JP. Trends in bilateral oophorectomy at the time of hysterectomy for Benign disease. Obstet Gynecol. 2011;118(6):1280–1286. doi: 10.1097/AOG.0b013e318236fe61 [DOI] [PubMed] [Google Scholar]

[cit0007] 7.Green GT. Hysterectomy and Self-Esteem Among African American Women. 2022.

[cit0008] 8.Rajkumari S, Chaudhary V, Kasaudhan S, Saraswathy KN. Incidence and determinants of hysterectomy among North Indian women: an 8-year follow-up study. Front Public Health. 2022;101065081. doi: 10.3389/fpubh.2022.1065081 [DOI] [PMC free article] [PubMed] [Google Scholar]

[cit0009] 9.Tanaka LF, Schoffer O, König J, Weyer-Elberich V, Blettner M, Klug SJ. Changes in the probability of hysterectomy in the city of Mainz and Mainz-Bingen region, Germany. BMC Public Health. 2023;23(1):84. doi: 10.1186/s12889-022-14916-w [DOI] [PMC free article] [PubMed] [Google Scholar]

[cit0010] 10.Bi Y, Ma Y, Zhang J, et al. Prevalence of hysterectomy in urban China and associations between metabolic disorders and hysterectomy: a multicenter population-based study involving over 9 million women. BMC Med. 2026;24(1). doi: 10.1186/s12916-026-04737-8 [DOI] [PMC free article] [PubMed] [Google Scholar]

[cit0011] 11.Liu F, Pan Y, Liang Y, et al. The epidemiological profile of hysterectomy in rural Chinese women: a population-based study. BMJ Open. 2017;7(6):e015351. doi: 10.1136/bmjopen-2016-015351 [DOI] [PMC free article] [PubMed] [Google Scholar]

[cit0012] 12.Ye S, Wang W, Li Y, Shen W. Surgical management of uterine fibroids: a large retrospective study. J Minim Invasive Gynecol. 2025;32(9):823–829. doi: 10.1016/j.jmig.2025.05.011 [DOI] [PubMed] [Google Scholar]

[cit0013] 13.Martínez-Maestre MA, Jódar-Sánchez F, Calderón-Cabrera AM, González-Cejudo C, Silván-Alfaro JM, Melero-Cortés LM. Healthcare and indirect cost of the laparoscopic vs vaginal approach in benign hysterectomy. JSLS. 2022;26(3):e202200048. doi: 10.4293/jsls.2022.00048 [DOI] [PMC free article] [PubMed] [Google Scholar]

[cit0014] 14.Augusto CF, Caraça DB, Podgaec S. Epidemiological analysis of hysterectomies performed at the public health system in the largest Brazilian city. Rev Assoc Med Bras. 2021;67(7):937–941. doi: 10.1590/1806-9282.20210093 [DOI] [PubMed] [Google Scholar]

[cit0015] 15.Dumont S, Van Trappen P. The clinical and financial impact of introducing robotic-assisted hysterectomy in a tertiary referral centre: a direct cost analysis of consecutive hysterectomies during a decade. Int J Med Robot. 2022;18(1):e2343. doi: 10.1002/rcs.2343 [DOI] [PubMed] [Google Scholar]

[cit0016] 16.Gao L, Shi L, Meng Q, Kong X, Guo M, Lu F. Effect of healthcare system reforms on public hospitals’ revenue structures: evidence from Beijing, China. Soc Sci Med. 2021;283:114210. doi: 10.1016/j.socscimed.2021.114210 [DOI] [PubMed] [Google Scholar]

[cit0017] 17.Liu M, Jia M, Lin Q, Zhu J, Wang D. Effects of Chinese medical pricing reform on the structure of hospital revenue and healthcare expenditure in county hospital: an interrupted time series analysis. BMC Health Serv Res. 2021;21:1–9. doi: 10.1186/s12913-021-06388-2 [DOI] [PMC free article] [PubMed] [Google Scholar]

[cit0018] 18.Turner SL, Korevaar E, Cumpston MS, Kanukula R, Forbes AB, McKenzie JE. Effect estimates can be accurately calculated with data digitally extracted from interrupted time series graphs. Res Synth Meth. 2023;14(4):622–638. doi: 10.1002/jrsm.1646 [DOI] [PMC free article] [PubMed] [Google Scholar]

[cit0019] 19.Lagarde M. How to do (or not to do) … Assessing the impact of a policy change with routine longitudinal data. Health Policy Plan. 2012;27(1):76–83. doi: 10.1093/heapol/czr004 [DOI] [PubMed] [Google Scholar]

[cit0020] 20.Li L, Jiang JN, Xiang L, Wang XF, Zeng L, Zhong ZD. Impact of critical illness insurance on the burden of high-cost rural residents in central china: an interrupted time series study. Int J Environ Res Public Health. 2019;16(19):3528. doi: 10.3390/ijerph16193528 [DOI] [PMC free article] [PubMed] [Google Scholar]

[cit0021] 21.Jakovljevic M, Chang H, Pan J, et al. Successes and challenges of China’s health care reform: a four-decade perspective spanning 1985—2023. Cost Eff Resour Allocation. 2023;21(1):59. doi: 10.1186/s12962-023-00461-9 [DOI] [PMC free article] [PubMed] [Google Scholar]

[cit0022] 22.Xiong WY, Deng YF, Yang YL, Zhang YM, Pan JY. Assessment of medical service pricing in China’s healthcare system: challenges, constraints, and policy recommendations. Front Public Health. 2021;9787865. doi: 10.3389/fpubh.2021.787865 [DOI] [PMC free article] [PubMed] [Google Scholar]

[cit0023] 23.Mahomed O, Cassim N. Appropriateness of laboratory expenditure for primary health care facilities across South Africa. African J Primary Health Care Family Med. 2023;15(1):1–8. doi: 10.4102/phcfm.v15i1.3740 [DOI] [PMC free article] [PubMed] [Google Scholar]

[cit0024] 24.Van Wilder A, Bruyneel L, De Ridder D, et al. Is a hospital quality policy based on a triad of accreditation, public reporting and inspection evidence-based? A narrative review. Int J Qual Health Care. 2021;33(2). doi: 10.1093/intqhc/mzab085 [DOI] [PubMed] [Google Scholar]

[cit0025] 25.Li Y. Generic price regulation and drug expenditures: evidence from Canada. Value Health. 2023;26(6):833–840. doi: 10.1016/j.jval.2023.01.008 [DOI] [PubMed] [Google Scholar]

[cit0026] 26.Jin Y, Kim Y. Dietary advice in chronic care: comparing traditional Chinese and western medicine practiced in mainland China. Soc Sci Med. 2022;292114621. doi: 10.1016/j.socscimed.2021.114621 [DOI] [PubMed] [Google Scholar]

[cit0027] 27.Gujar A, Gulecha V, Zalte A. Drug utilization studies using WHO prescribing indicators from India: a systematic review. Health Policy Technol. 2021;10(3):100547. doi: 10.1016/j.hlpt.2021.100547 [DOI] [Google Scholar]

[cit0028] 28.Chen YJ, Zhang XY, Yan JQ, Qian MC, Ying XH. Impact of diagnosis-related groups on inpatient quality of health care: a systematic review and meta-analysis. Inquiry-J Health Care Organ Prov Finan. 2023;6000469580231167011. doi: 10.1177/00469580231167011 [DOI] [PMC free article] [PubMed] [Google Scholar]

[cit0029] 29.Khammarnia M, Bagherbarahouei F, Ansari-Moghaddam A. Length of hospital stay and its related factors in Iran: a systematic review and meta-analysis. Iran J Public Health. 2023;52(9):1877–1888. doi: 10.18502/ijph.v52i9.13570 [DOI] [PMC free article] [PubMed] [Google Scholar]

[cit0030] 30.Oberhollenzer F, Salzburger S, Auer HM, et al. Average length of stay after a cardiology day hospital stay compared to a normal ward: an analysis by using propensity score matching. Wien Klin Wochenschr. 2022;134(Suppl 2):194–195. [Google Scholar]

[cit0031] 31.Bala C, Rusu A, Ciobanu D, Roman G. Length of hospital stay, hospitalization costs, and their drivers in adults with diabetes in the romanian public hospital system. Int J Environ Res Public Health. 2022;19(16):10035. doi: 10.3390/ijerph191610035 [DOI] [PMC free article] [PubMed] [Google Scholar]

[cit0032] 32.Allaudeen N, Vashi A, Breckenridge JS, et al. Using lean management to reduce emergency department length of stay for medicine admissions. Qual Manag Health Care. 2017;26(2):91–96. doi: 10.1097/qmh.0000000000000132 [DOI] [PubMed] [Google Scholar]

[cit0033] 33.Zhai CX, Bai LL, Xu Y, et al. Temperature variability associated with respiratory disease hospitalisations, hospital stays and hospital expenses the warm temperate sub-humid monsoon climate. Public Health. 2023;225:206–217. doi: 10.1016/j.puhe.2023.10.013 [DOI] [PubMed] [Google Scholar]

[cit0034] 34.Golinelli D, Sanmarchi F, Toscano F, Bucci A, Nante N. Analyzing the 20-year declining trend of hospital length-of-stay in European countries with different healthcare systems and reimbursement models. Int J Health Econom Manag. 2024. doi: 10.1007/s10754-024-09369-0 [DOI] [PMC free article] [PubMed] [Google Scholar]

[cit0035] 35.Kim S, Jung C, Yon J, et al. A review of the complexity adjustment in the Korean Diagnosis-Related Group (KDRG). Health Inform Manag J. 2020;49(1):62–68. doi: 10.1177/1833358318795804 [DOI] [PubMed] [Google Scholar]

[cit0036] 36.Maréchal F, Thomas L. The impact of medical complications on optimal hospital payment. J Public Econ Theory. 2021;23(6):1144–1173. doi: 10.1111/jpet.12530 [DOI] [Google Scholar]

[cit0037] 37.Gluckman TJ, Spinelli KJ, Wang MS, et al. Trends in diagnosis related groups for inpatient admissions and associated changes in payment from 2012 to 2016. JAMA Network Open. 2020;3(12):e2028470. doi: 10.1001/jamanetworkopen.2020.28470 [DOI] [PMC free article] [PubMed] [Google Scholar]

[cit0038] 38.Cook A, Averett S. Do hospitals respond to changing incentive structures? Evidence from medicare’s 2007 DRG restructuring. J Health Econ. 2020;73102319. doi: 10.1016/j.jhealeco.2020.102319 [DOI] [PMC free article] [PubMed] [Google Scholar]

[cit0039] 39.Meng ZL, Ma YN, Song SH, et al. Economic implications of chinese diagnosis-related group-based payment systems for critically ill patients in ICUs. Crit Care Med. 2020;48(7):E565–E573. doi: 10.1097/ccm.0000000000004355 [DOI] [PubMed] [Google Scholar]

[cit0040] 40.Baker LC, Lamiraud K. Adoption of hospital diagnosis-related group financing in Switzerland and the availability of computed tomography scanners. Health Econ. 2022;31(12):2537–2557. doi: 10.1002/hec.4594 [DOI] [PubMed] [Google Scholar]

[cit0041] 41.Ma YF, Wang W. Original The impact of diagnosis related group payment on the performance of public hospitals. Am J Transl Res. 2021;13(6):6796–6801. [PMC free article] [PubMed] [Google Scholar]

[cit0042] 42.Yu LH, Lang JJ. Diagnosis-related Groups (DRG) pricing and payment policy in China: where are we? Hepatob Surg Nutr. 2020;9(6):771–773. doi: 10.21037/hbsn-2020-8 [DOI] [PMC free article] [PubMed] [Google Scholar]

[cit0043] 43.Aktas P. Physician perspectives on the implications of the diagnosis-related groups for medical practice in Turkey: a qualitative study. Int J Health Plann Manage. 2022;37(3):1769–1780. doi: 10.1002/hpm.3445 [DOI] [PMC free article] [PubMed] [Google Scholar]

PERMALINK

Changes in Hospitalization Expenditures and Length of Stay for Patients Underwent Hysterectomy and Adnexectomy for Carcinoma in situ and Nonmalignant Lesions Diagnosis-Related Group: Impact of Beijing’s Medicine–Consumables Linkage Reforms

Juanjuan Liang

Yuanqing Cai

Rong Fan

Yuwei Zhang

Yangyi Hu

Wenjuan Wang

Abstract

Purpose

Methods

Results

Conclusion

Introduction

Policy Background

Table 1.

DRG

HAC

Methods

Study Data

Indicators Selection

Statistical Analysis

Results

Descriptive Analysis

Table 2.

ITSA Analysis

Table 3.

Figure 1.

Figure 2.

Total Expenditures and Structure

Total Expenditures

Surgery and Treatment Expenditures

Examination and Laboratory Test Expenditures

Drug and Medical Consumables Expenditures

OOP Expenditures and LOS for Patients

Robustness Check

Table 4.

Discussion

Total Expenditures and Structure

Surgery, Treatment, Examination and Laboratory Test Expenditures

Drug and Medical Consumables Expenditures

Cost Burden and Hospitalization LOS for Patients

With or without CC

Policy Implications

Limitations

Conclusions

Acknowledgments

Funding Statement

Abbreviations

Data Sharing Statement

Ethics Approval and Consent to Participate

Author Contributions

Disclosure

References

Associated Data

Data Availability Statement

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases