Impact of a National Volume-Based Procurement Policy in China on purchase of nucleoside/nucleotide analog drugs for the treatment of chronic hepatitis B virus infection

Abstract

Background

Hepatitis B virus (HBV) infection is a global public health issue, with approximately 1.5 million new infections and around 820,000 deaths annually around the globe. China carries the heaviest burden of HBV infection in the world, making it crucial to improve the accessibility of antiviral treatment for hepatitis B. In 2019, China implemented a National Volume-Based Procurement (NVBP) policy, which included two nucleoside/nucleotide analog drugs (NAs) -- Entecavir (ETV) and Tenofovir Disoproxil Fumarate (TDF) -- in the list for centralized procurement. The policy aimed to reduce drug prices and alleviate financial burden on patients. This study aimed to evaluate the impact of the NVBP policy on purchase volume and expenditures for antiviral drugs to treat HBV.

Methods

The study used interrupted time series analysis of national procurement data for six antiviral therapies from January 2019 to December 2020. It examined the changes in monthly drug procurement volume and expenditures at national level and in different levels of hospitals.

Results

After the policy implementation, the prices of ETV and TDF decreased in all provinces, with maximum price reductions of 98.50% and 98.61%, respectively. Both ETV and TDF showed significant increases in purchase volume (38.0 [P < 0.001] and 6.8 [P < 0.001] million DDDs, respectively) and significant reductions in expenditures (-143.7 [P < 0.001] and − 30.4 [P < 0.001] million CNY, respectively) immediately after policy implementation. The absolute change in purchase volume and expenditure was highest in tertiary hospitals, but the relative change in magnitude was greatest in primary hospitals.

Conclusion

The NVBP policy was effective in reducing the cost of ETVs and TDFs and significantly increased the prescribing volume. More patients were likely to have access to NA drugs after the policy was implemented. This policy successfully reduced the financial burden on patients and facilitated their access to hepatitis B treatment.

Introduction

Hepatitis B virus (HBV) infection is a significant global public health problem. According to the World Health Organization, in 2019 there were 296 million people infected with the HBV, with approximately 1.5 million new infections occurring each year, and about 820,000 people die annually due to HBV-related complications, mainly attributed to cirrhosis and hepatocellular carcinoma [1]. In May 2016, the World Health Assembly adopted the first “Global Health Sector Strategy on Viral Hepatitis 2016–2021.” This strategy proposed to eliminate viral hepatitis as a public health threat by 2030, including reducing the number of new chronic infections by 90% and decreasing mortality rate by 65% compared to 2015 [2]. In May 2022, during the 75th World Health Assembly, more specific targets were proposed, with quantifiable national indicators.

China carries the heaviest burden of HBV infections in the world, with an estimated 100 million people infected with chronic hepatitis B (CHB) virus, accounting for nearly 39% of the global total [3–5]. The prevalence of hepatitis B surface antigen (HbsAg) in the general population in China is 9.8% [6]. Therefore, China will need to be a major contributor to the global goal of eliminating hepatitis B by 2030 [7].

HBV infection imposes a significant clinical and economic burden on patients. CHB patients have a high risk of developing severe complications [8]. Antiviral treatment is effective in improving the lives of CHB patients. Currently, antiviral treatment for CHB mainly consists of oral nucleotide analogs (NA), including entecavir (ETV), tenofovir disoproxil fumarate (TDF), tenofovir alafenamide fumarate (TAF), lamivudine (LAM), telbivudine (LdT), and adefovir dipivoxil (ADV) (Table 1). LAM has reasonable antiviral efficacy, but minimal use in clinical practice due to drug resistance. ADV and LdT have relatively weaker antiviral effects, longer virological response time, and due to the risk of renal tubular injury with ADV, they are also rarely used. ETV, TDF, and TAF are the first-line antiviral drugs recommended by guidelines for treatment of HBV infection [9].

Table 1.

Drug description of NAs

Drug			Launch
	Manufacturer of originator	Specifications	US	China	Patent expiration	Listing in China	Clinical effects	Resistance	Indications
LAM	GlaxoSmithKline	0.3 g*30	1998	1999	2019	2017	Strong	High	First oral antiviral medication marketed
ADV	Gilead Sciences	10 mg*30	2002	2005	2018	2017	Moderate	Low	Contraindicated in patients with impaired renal function
LdT	Merck & Co Inc.	0.6 g*7	2006	2007	2023	2017	Moderate	HIgh	Preferred medication for use during pregnancy
ETV	Bristol Myers Squibb	0.5 mg*7	2005	2005	2010	2017	Strong	Low	Dosage adjustment based on renal function
TDF	GlaxoSmithKline	0.3 g*30	2008	2014	2017	2010	Strong	No	Medication requires monitoring for renal tubular injury
TAF	Gilead Sciences	25 mg*30	2016	2019	2032	2020	Strong	No	Suitable for patients with impaired renal function

Although proactive antiviral treatments can halt disease progression and reduce the risk of hepatocellular carcinoma, only 1% of 257 million HBV-infected individuals in 2017 received adequate treatment [10]. Due to the necessity of prolonged treatment with high cost antiviral medications, the cost of long-term therapy can be a significant burden for patients and their families [11]. The high cost of NA drugs prevented patients from receiving relevant treatment, which in turn led to the deterioration of hepatitis B towards severe disease [12]. Therefore, lowering the prices of antiviral drugs has the potential to expand access to CHB treatment and reduce the number of deaths related to HBV-associated cirrhosis and liver cancer [7].

Countries have explored several different procurement models as strategies to lower drug prices. In the UK, qualified drugs with the lowest prices are selected through bidding, ensuring both drug quality and affordability [13]. In India’s Delhi state, a “Two-Envelope System” is implemented, where technical and price bids are placed in separate envelopes, only products that meet the technical standards set by the procurement committee have their price bids opened [14]. In the United States, many private healthcare institutions purchase medications through third-party organizations such as pharmacy benefits managers or group purchasing organizations, leveraging larger population sizes to negotiate lower prices [15].

Volume-based procurement is a crucial component of China’s healthcare system reform, aiming to alleviate the burden of medical expenses for the population. In November 2018, the National Joint Procurement Office initiated a pilot program for volume-based procurement in the so-called “4 + 7” cities, including four municipalities and seven provincial capitals. The policy allowed pharmaceutical companies to obtain large-scale orders through centralized transactions, with purchase volume clearly specified during the bidding, thus linking volume with price. Winning companies produce the agreed-upon quantity and ensure quality. Health authorities assess the volume used by hospitals, ensure timely payments at the agreed procurement price, and establish mechanisms within hospitals to encourage doctors to prescribe drugs from the procurement list. Pharmaceutical companies obtain a large volume of orders through centralized procurement, thereby achieving cost savings in drug circulation marketing expenses and resulting in reduced final drug prices.Literature showed that after implementation of the “4 + 7” pilot program, the prices of 25 selected drugs decreased by an average of 52% [4].

In December 2019, the National Health Security Administration expanded the “4 + 7” pilot program to a national scale. Two NA antiviral drugs for hepatitis B, namely ETV and TDF, were among the 25 drugs covered by the National Volume-Based Procurement Policy (NVBP) .Previous studies have evaluated the impact of the NVBP in pilot cities. For instance, Wen et al. (2023) utilized data from 11 pilot cities to demonstrate reduced costs and increased generic drug use [16]. However, the impact of the policy’s nationwide expansion, particularly the heterogeneity of its effects across different tiers of healthcare facilities (primary vs. tertiary), remains underexplored. ETV and TDF involved in procurement are both generic drugs that have undergone consistency evaluation. These generics not only match the quality and efficacy of the original drugs but also come with substantial price reductions. This contributed to alleviating the financial burden of medication expenses for patients with CHB [17, 18].

This study aims to assess the impact of the NVBP policy on drugs used for CHB treatment. We examined trends in the procurement volume and expenditures for antiviral drugs for hepatitis B before (January 2019 to December 2019) and after (January 2020 to December 2020) the implementation of the national policy. We also described changes in the number of hepatitis B patients hospitalized before and after the NVBP implementation. This research can provide valuable insights about the impacts of volume-based procurement in order to shape the implementation of future policies.

Methods

Data

Data on procurement by health facilities of all NA antiviral drugs for CHB treatment for the period January 2019 to December 2020 were extracted from the NHSA database. Both ETV and TDF were subject to centralized procurement in 2019, while the alternative antiviral therapies (TAF, LAM, LdT, and ADV) were not. The dataset included procurement records of 1.115 million health facilities in 24 province-level divisions, covering a population exceeding 762.0 million. It is important to note that this dataset specifically represents the regions covered by the national expansion of the NVBP policy and excludes data from the original ‘4 + 7’ pilot cities (Beijing, Tianjin, Shanghai, Chongqing, Shenyang, Dalian, Xiamen, Guangzhou, Shenzhen, Chengdu, and Xi’an), ensuring a consistent policy implementation start date of December 2019 across all included samples.

All NA products were classified as brand vs. generic, and by whether or not they had passed equivalency evaluation. In addition, all ETV and TDF products subject to centralized procurement were classified as bid-winning or non-winning. We classified hospitals by whether they were tertiary (large academic medical centers), secondary (typically large, multispecialty provincial hospitals), or primary (typically smaller municipal or rural hospitals).

In addition, we obtained the annual number of hospitalizations for a variety of liver diseases, including hepatitis B, from the Chinese Health Statistics Yearbook (2023) [19] to compare the impact of NVBP on severe liver disease before and after its implementation.

The investigation and research process conforms to the principles outlined in the Helsinki Declaration. This study uses medical institution procurement records and does not involve patient or individual information.

Outcomes

Outcomes included monthly drug purchase volumes and monthly expenditures. Volumes were measured in Defined Daily Doses (DDDs), calculated as the total quantity (in mg) of a purchased drug divided by its specific Defined Daily Dose, as standardized by the World Health Organization) [20]. Drug expenditures were reported in million Chinese yuan renminbi (CNY). We determined the average monthly purchase price per DDD by dividing total monthly expenditures by monthly purchase volumes. Since NVBP affected drug prices and DDDs at the same time, we further used DDDc to measure the average daily cost of drugs, which were defied as the daily expenditure divided by DDDs. Affordability of treatment at the national level was defined as the monthly average treatment cost for an individual patient divided by average daily wage in the corresponding year obtained from statistical yearbook (see Eq. 1).

1

Statistical analysis

We first calculated monthly DDDs total expenditures, and DDDc for all NAs, for each type of NA, and for each level of health facility. We displayed the data as time series and modeled the impact of the policy using the interrupted time series (ITS) analysis.

2

This quasi-experimental method can explore the effect of external shocks on the trends of volume and expenditure of drugs. The regression equation was constructed as Eq. 2. was the DDDs, expenditure, or DDDc of one drug at month t. was the constant term. The coefficient of time after the start of data (), , indicated the slope before the implementation of the policy. equals one after December 2019, the first month of NVBP implementation, and zero otherwise. The coefficient of whether the policy had been implemented at month t (), , indicated the immediate counterfactual change of the volume/expenditure at the first month of implementation. equals zero before the NVBP implementation, and the number of months since the start of NVBP. The coefficient of time after the implementation (), , indicated the change in slope due to the implementation. was the error term.

We further calculated the estimated percent change as follows: we first calculated the estimate for the last period, that is, the sum of the estimate of, the estimate of, the estimate of 24 times, and the estimate of 12 times. Since depicts the estimated time trend (24 months) and estimates the change in trend after the policy is introduced (12 months), we multiplied these two coefficients by the corresponding number of months. Second, we computed a counterfactual estimate of the value of the last period assuming no strategic purchase, that is plus 24 times. Finally, we used the estimated last period value divided by the estimated counterfactual last period value to get the estimated percent change.

To ensure the validity of the ITS model, we performed pre-estimation diagnostics. First, we assessed seasonality using autocorrelation function (ACF) plots; no significant seasonal patterns were observed, consistent with the chronic nature of Hepatitis B treatment. Second, stationarity was examined using the Augmented Dickey-Fuller (ADF) test. The Durbin-Watson statistics was used to detect residual autocorrelations. To address potential serial correlation and heteroscedasticity identified by these diagnostics, we estimated the regression models using Newey-West standard errors. A p-value of less than 0.05 was considered statistically significant. All analyses were performed in Python 3.11 and Stata SE 16.0.

Results

Changes in NAs at the national level

Unit price and affordability

This study included a total of six NA drugs (by generic names). ETV and TDF were drugs related to the NVBP policy, while TAF, LAM, LdT, and ADV were considered as substitute drugs. To observe the impact of policy on NVBP-related and substitute drugs, the research analyzed the change in unit price and affordability of these six drugs before and after the policy (Table 2). Since the implementation in December 2019, the affordability of ETV and TDF in various provinces showed an increasing trend. The maximum reduction in price could reach up to 98.50% and 98.61% respectively. Dawnrays Pharmaceutical from Suzhou achieved the most significant price reduction for ETV, lowering it from 13.3 yuan per tablet to 0.2 yuan per tablet, a reduction of 98.50%. This translated to a monthly expenditure decrease from 397.9 yuan to 5.5 yuan, a reduction of 98.61%. For TDF, the largest reduction per unit was achieved by Qilu Pharmaceutical, lowering the price from 14.4 yuan per tablet to 0.3 yuan per tablet, a reduction of 97.91%. This resulted in a monthly expenditure decrease from 431.2 yuan to 8.7 yuan, a reduction of 97.98%. In addition, we also found that except for a certain degree of reduction in TAF, other alternatives like ADV did not exhibit significant changes. By examining the “unit comparable price”, it can be observed that TAF’s price decreased from 39.3 yuan per tablet to 18.0 yuan per tablet, resulting in a reduction of 54.20%.

Table 2.

Changes in Affordability of NAs Following NVBP Implementation

			Affordability
			Pre			Post			Unit Price			Monthly Expenditure
Drug	Manufacturer	Specifications	Min	Mean	Max	Min	Mean	Max	Pre	Post	Daily/mg	Pre	Post
ADV	GlaxoSmithKline	Per mg	0.323	0.575	0.77	0.232	0.414	0.555	0.7	0.5	10	216.6	156.1
LAM	GlaxoSmithKline	Per mg	0.332	0.591	0.792	0.293	0.521	0.698	0.1	0.1	100	222.9	196.5
LdT	Beijing Novartis	600 mg*7	0.183	0.327	0.438	0.183	0.327	0.438	17.6	17.6	600	123.2	123.2
TAF	Gilead	25 mg*30	1.757	3.13	4.192	0.803	1.431	1.916	39.3	18	25	1180	539.4
ETV	Suzhou Dawnrays	0.5 mg*21	0.761	1.467	2.064	0.011	0.021	0.029	13.3	0.2	0.5	397.9	5.5
	Beijing Bai’ao	0.5 mg*28	0.656	1.264	1.779	0.011	0.022	0.031	11.4	0.2	0.5	342.9	5.9
	Fujian Cosunter	0.5 mg*28	0.665	1.283	1.805	0.016	0.03	0.043	11.6	0.3	0.5	348	8.2
TDF	Qilu	300 mg*30	0.824	1.59	2.237	0.017	0.032	0.045	14.4	0.3	300	431.2	8.7
	Hangzhou Heze	300 mg*30	0.86	1.659	2.334	0.022	0.042	0.06	15	0.4	300	449.9	11.4
	ChengDu Brilliant	300 mg*30	0.759	1.462	2.058	0.027	0.052	0.073	13.2	0.5	300	396.7	14

Purchase volume

Before the implementation, the monthly average procurement volume of the six NA drugs showed an instant increase of 46.6 million DDDs (Fig. 1; Table 3). As of December 2020, ETV (77.4%) and TDF (17.5%) accounted for the highest volume of procurement. We estimated it was equivalent to a 100.8% increase compared to not implementing the NVBP.

Fig. 1.

Volume of all NAs and specific categories of ETV and TDF at the national level

Table 3.

Volume, expenditure, and pricing of NAs at the national level (ITS)

Volume (thousand DDDs)	Baseline level	Baseline trend	Level change	Trend change	DW	Estimated percent change
All NA Antivirals	31,542.3***	206.6	46,606.6***	-818.5	2.538	100.8%
ETV	22,605.6*	201.2	38,044.5***	-521.2	2.620	115.9%
Bid-winning Generic	46.3	125.4	51,623.4***	-255.9	2.692	1588.8%
Non-winning Generic	19,726.6***	83.9	-12,454.4***	-264.9	2.445	-71.9%
Non-winning Originaor	1,538.0***	4.8	-358.0*	4.0	2.613	-18.8%
Non-winning Nonequivalent	1,164.1***	-15.6	-607.3***	-2.6	1.989	-80.9%
TDF	3,901.7**	159.1	6,800.1***	-86.9	2.669	74.6%
Bid-winning Generic	127.7	31.2	8,551.5***	96.3	2.590	1107.5%
Non-winning Generic	1,585.4***	107.6**	-1,154.4***	-139.3**	2.417	-67.8%
Non-winning Originator	2,188.6***	20.3	-597.1*	-43.9	2.633	-42.0%
ADV	2472.8	95.8	1276.0	-195.7	0.518	-22.5%
TAF	-4.3	1.1	-155.0	56.5*	0.953	2366.5%
LAM	729.8***	-16.5	78.4	4.2	2.291	38.6%
LdT	365.4***	-7.6	7.1	4.3	2.173	32.1%
Expenditure (million CNY)
All NA Antivirals	373.8***	-0.9	-169.6**	-7.4	2.295	-73.4%
ETV	281.3***	-1.5	-143.7***	-4.0	1.922	-78.1%
Bid-winning Generic	6.2***	-0.1	6.9***	-0.1	2.426	150.0%
Non-winning Generic	210.0***	-0.2	-132.4***	-4.6	2.377	-91.4%
Non-winning Originator	47.3***	-0.6	-11.0	0.3	2.497	-22.5%
Non-winning Nonequivalent	12.2***	-0.2	-6.4***	0.0	1.845	-86.5%
TDF	52.0***	1.2	-30.4***	-2.6*	2.415	-76.2%
Bid-winning Generic	2.6***	0.0	0.3	-0.1	2.011	-34.6%
Non-winning Generic	19.4***	1.0**	-19.0***	-1.7***	2.409	-90.8%
Non-winning Originator	30.1***	0.2	-11.7**	-0.8	2.523	-61.0%
ADV	20.6***	-0.4	4.6*	-1.2***	0.986	-89.1%
TAF	-0.2	0.0	-2.8	1.0*	0.946	——
LAM	16.9***	-0.4	1.6	-0.0	2.244	21.9%
LdT	6.5***	-0.1	0.1	0.1	2.184	31.7%
DDDc (CNY/DDDs)
All NA Antivirals	11.8***	-0.1*	-8.2***	-0.0	2.412	-87.2%
ETV	12.4***	-0.2*	-8.6***	0.1	2.478	-97.4%
Bid-winning Generic	13.0***	-0.5**	-7.1***	0.5*	2.165	——
Non-winning Generic	-15.0	1.1	0.3	-1.0	1.637	——
Non-winning Originator	0.8	1.1	-0.6	-1.0	2.018	-46.3%
Non-winning Nonequivalent	10.6***	-0.1	-0.8**	-0.1	2.364	-24.4%
TDF	13.1***	-0.1*	-8.7***	0.0	2.337	-81.3%
Bid-winning Generic	12.8***	-0.4***	-7.6***	0.4**	2.029	-87.5%
Non-winning Generic	12.0***	-0.1	-4.3***	-0.2*	1.922	-69.8%
Non-winning Originator	13.7***	-0.0	-2.1***	-0.2*	1.450	-32.8%
ADV	6.8**	-0.2	-0.1	-0.2	0.313	——
TAF	39.3***	-0.0	-7.7	-1.4	0.451	-62.3%
LAM	23.0***	0.1	-0.3	-0.2**	2.167	-10.6%
LdT	17.9***	-0.0	-0.0	-0.0*	1.503	0.0%

a. Boldface text indicates the names of the NA inhibitor drugs. Non-boldface text represents the subcategories (e.g., bid-winning generic, non-winning originator) under each boldface drug.* p<0.05, ** p<0.01, *** p<0.001

For ETV, the implementation of the policy was followed by an immediate and sustained increase of 38.0 million DDDs (P < 0.001) per month, with no significant change in trend (Table 3). The volume of the bid-winning generic ETV brands increased by 51.6 million DDDs instantly (P < 0.001), with non-winning generics decreasing by 12.5 million DDDs (P < 0.001), the non-winning originator decreasing by 358.0 thousand DDDs (P < 0.05), and the nonequivalent decreasing by 607.3 thousand DDDs (P < 0.001). No significant trend was observed before or after the policy.

TDF experienced an immediate increase in procurement volume by 6.8 million DDDs (P < 0.001, Table 3). Specifically, bid-winning generics increased by 8.6 million DDDs, while non-winning generics and originator decreased by 1.2 million DDDs (P < 0.001) and 597.1 thousand DDDs (P < 0.05) respectively. The rising trend of non-winning generics (107.6 thousand DDDs per month, P < 0.01) was reversed by the policy (-139.3 thousand DDDs per month, P < 0.01).

Among the alternatives, ADV, LAM, and LdT showed no significant policy impacts, while an rising trend of TAF volume was observed (56.5 thousand DDDs per month, P < 0.05).

Expenditures

Although purchase volume of NAs doubled (100.8%) after the policy, expenditures decreased by 73.4% (Fig. 2; Table 3). The average monthly expenditure for all NA drugs (373.8 million CNY) declined to 99.43 million CNY after the policy.

Fig. 2.

Expenditures of all NAs and specific categories of ETV and TDF at the national level

For ETV, monthly expenditures (281.3 million CNY, P < 0.001) instantly declined by 143.7 million CNY (P < 0.001). Expenditures of bid-winning generics increased by 6.9 million CNY (P < 0.001), but non-winning generics and the nonequivalent decreased by 132.4 and 6.4 million CNY (P < 0.001) respectively. No significant trend was observed before or after the policy.

For TDF, monthly expenditures (52.0 million CNY, P < 0.001) instantly declined by 30.4 million CNY (P < 0.001). Expenditures of bid-winning generics remained stable, but non-winning generics and the non-winning originator decreased by 19.0 and 11.7 million CNY (P < 0.001) respectively. No significant trend was observed before or after the policy. The rising trend of non-winning generics (1.0 million CNY per month, P < 0.01) was reversed by the policy (-1.7 million CNY per month, P < 0.001).

Among the alternatives, an rising trend of TAF expenditures was observed (1.0 million CNY per month, P < 0.05). The expenditures of ADV increased instantly (4.6 million CNY, P < 0.05) along with a declining trend (-1.2 million CNY per month, P < 0.001).

DDDc

DDDc is a direct measure of patients’ daily burden of diseases. For NAs as a whole, DDDc decreased by 8.2 CNY per day (P < 0.001, Fig. 3; Table 3) immediately after the implementation, translating into an 87.2% decrease from baseline.

Fig. 3.

DDDc of all NAs and specific categories of ETV and TDF at the national level

For ETV (-8.6 CNY per day, P < 0.001), the immediate decline was mainly contributed by bid-winning generics (-7.1 CNY per day, P < 0.001) and non-winning nonequivalents (-0.8 CNY per day, P < 0.01).

For TDF (-8.7 CNY per day, P < 0.001), the immediate decline was contributed by all specific categories, including bid-winning generics (-7.6 CNY per day, P < 0.001), non-winning generics (-4.3 CNY per day, P < 0.001), and the non-winning originator (-2.1 CNY per day, P < 0.001).

Among the alternatives, LAM (0.2 CNY/day decrease per month, P < 0.01) and LdT (<-0.1 CNY/day decrease per month, P < 0.05) exhibited a downward trend. No other impact was observed.

Heterogeneous impacts of NVBP by hospital type

DDDs rose significantly at all levels of hospitals following NVBP implementation, and expenditures and DDDc declined (Table 4; Fig. 4). Tertiary hospitals saw the largest immediate increase in DDDs (23.4 million DDDs, P < 0.001) and the largest cut in expenditures (-107.0 million CNY, P < 0.001). Primary hospitals, however, witness the largest improvement in DDDs (2268.2%) and expenditures (-285.7%). DDDc showed a similar magnitude of change for all levels of hospitals.

Table 4.

Volume, expenditure, and pricing of NAs across health facilities (ITS)

Volume (thousand DDDs)	Baseline level	Baseline trend	Level Change	Trend change	DW	Estimated percent change
All NA Antivirals
Primary	1,946.0***	-73.8	1,826.4***	178.2**	2.029	2268.2%
Secondary	11,549.9**	103.9	21,546.3***	-775.5	1.903	87.2%
Tertiary	16,274.5***	381.1	23,393.3***	-529.6	2.093	67.0%
ETV
Primary	1,544.6***	-60.6	1,648.1***	152.7**	1.969	3858.6%
Secondary	8,429.0**	116.8	18,337.0***	-595.8	1.997	99.6%
Tertiary	11,494.7***	251.9	18,504.6***	-261.1	2.079	87.6%
TDF
Primary	118.0*	1.0	122.4	23.9*	2.355	288.2%
Secondary	994.9*	35.8	2,437.8***	-39.7	1.793	105.8%
Tertiary	2,583.8***	145.5	4,307.4***	-114.3	2.141	48.3%
Expenditure (million CNY)
All NA Antivirals
Primary	22.6***	-1.0**	-5.6*	0.8*	2.061	——
Secondary	128.9***	-0.2	-57.0**	-3.1	2.222	-75.9%
Tertiary	222.3***	0.2	-107.0***	-5.1	2.292	-74.1%
ETV
Primary	18.0***	-0.8**	-4.9*	0.7*	2.111	——
Secondary	97.8***	-0.0	-52.5***	-2.0	2.277	-78.2%
Tertiary	159.9***	-0.2	-85.5***	-3.1	2.320	-79.1%
TDF
Primary	1.5***	0.0	-0.8**	-0.0	1.996	-53.3%
Secondary	12.9***	0.3	-7.2***	-0.6*	2.253	-71.6%
Tertiary	37.6***	0.9	-22.4***	-2.0*	2.407	-78.4%
DDDc (CNY/DDDs)
All NA Antivirals
Primary	11.7***	-0.1	-8.3***	-0.0	2.265	-89.2%
Secondary	11.1***	-0.1*	-7.8***	0.0	2.485	-89.7%
Tertiary	12.3***	-0.1	-8.4***	-0.0	2.345	-84.8%
ETV
Primary	11.8***	-0.1*	-8.8***	0.0	2.477	-93.6%
Secondary	11.5***	-0.1**	-8.3***	0.1	2.603	-78.0%
Tertiary	12.6***	-0.1*	-8.9***	0.0	2.381	-87.3%
TDF
Primary	12.8***	-0.0	-9.2***	-0.2*	2.266	-90.6%
Secondary	12.8***	-0.1*	-9.1***	0.0	2.403	-87.5%
Tertiary	13.3***	-0.2**	-8.5***	0.0	2.326	-100.0%

a.Boldface text indicates the names of the NA inhibitor drugs. Non-boldface text represents thesubcategories (e.g., bid-winning generic, non-winning originator) under each boldface drug.

Fig. 4.

Volume, expenditure, and DDDc of ETV and TDF across health facilities

For ETV, the immediate increases were 18.5 million DDDs (P < 0.001) in tertiary hospitals, 18.3 million DDDs (P < 0.001) in secondary hospitals, and 1.6 million DDDs (P < 0.001) in primary hospitals. A rising trend was found in primary hospitals by (152.7 thousand DDDs per month, P < 0.01). Although tertiary and secondary hospitals benefited most in absolute terms, the increase was relatively larger in primary hospitals (87.6% & 99.6% versus 3838.1%, respectively). A similar pattern was observed in reduced expenditures for ETV.

For TDF, the purchase volume substantially increased in secondary hospitals (2.4 million DDDs, P < 0.001) and tertiary hospitals (4.3 million DDDs, P < 0.001). In primary hospitals, the immediate increase was insignificant (122.4 thousand DDDs, P > 0.05), but volume started to increase by 23.90 thousand DDDs per month (P < 0.05). The expenditure for TDF decreased by 0.8 (P < 0.01) million CNY in primary hospitals, 7.2 million CNY (P < 0.001) in secondary hospitals, and 22.4 million CNY (P < 0.001) in tertiary hospitals. A downward trend was observe in secondary (0.6 million CNY per month, P < 0.05) and tertiary (2.0 million CNY per month, P < 0.05) hospitals.

Current situation of severe hepatitis incidence and hospitalization

Table 5 presents the national inpatient numbers (in thousand) and annual decline rates for selected diseases from 2017 to 2021. Most hepatitis patients who progress to the hospitalization stage are severely infected. Before the implementation of the policy, the inpatient numbers for hepatitis gradually decreased (2018, -4.3%; 2019, -4.7%), while the inpatient numbers for other liver-related diseases and all diseases showed an increasing trend each year (with the smallest increase in 2018 at 10.1% for liver cancer; and the smallest increase in 2019 at 6.9% for alcoholic liver disease). In 2020, the inpatient number for severe hepatitis significantly decreased (206.8 thousand in 2019 versus 128.6 thousand in 2020, -37.81%). Compared to the decline rates of other diseases like alcoholic liver disease (-8.2%), cirrhosis (-8.7%), liver cancer (-10.6%), diabetes (-12.2%) and all diseases (14.1%), it is evident that despite the COVID-19 pandemic indeed had a significant impact on patients’ hospitalization behavior, it is worth noting that CHB patient may have also been influenced by the NVBP policy to a considerable extent.

Table 5.

National inpatient statistics of other diseases from 2017 to 2021

Year	2017		2018		2019		2020		2021
Hepatitis B	227	(ref)	217	-4.30%	207	-4.70%	129	-37.80%	106	-17.90%
Alcoholic liver disease	48	(ref)	54	13.40%	58	6.90%	53	-8.20%	42	-21.50%
Cirrhosis	352	(ref)	399	13.50%	437	9.60%	399	-8.70%	351	-12.20%
Liver cancer	238	(ref)	262	10.10%	297	13.20%	265	-10.60%	213	-19.60%
Diabetes	1,768	(ref)	2,065	16.80%	2,414	16.90%	2,120	-12.20%	2,095	-1.20%
All	78,658	(ref)	87,577	11.30%	100,169	14.40%	86,065	-14.10%	81,017	-5.90%

Discussion

The aim of NVBP was to adopt a volume-for-price strategy to reduce the procurement price of drugs, thereby reducing the burden of drug costs on patients. Our analysis of the NA class of drugs for the treatment of CHB found that following the implementation of the NVBP at a near-national level in December 2019, there was a significant reduction in the unit price of drugs within the NVBP (ETV and TDF) and a significant increase in affordability and accessibility for patients. Specifically, the prices of ETV and TDF have decreased from 10 to 15 CNY/day to 0.2–0.5 CNY/day, and from 300 to 500 CNY/month to 5–15 CNY/month.

We performed ITS analyses on DDDs, total expenditures, and DDDc. The results showed that the average DDDs for the six NA drugs increased significantly by 46,606.6 (P < 0.001) immediately after the implementation of the policy, which may indicate that more patients received treatment for CHB; the total expenditures decreased significantly by CNY 169.6 (P < 0.01) million despite the increase in the usage volume of the drugs; and the increase in the DDDs and decline in the total expenditure resulted in a significant decrease in the DDDc by 87.2%, indicating that the burden of drug costs on patients for the treatment of CHB has been significantly alleviated.

Further analyses of NVBP drugs found that the purchase volume of ETV has increased most significantly and the total expenditure also decreased most significantly, which is the preferred treatment for patients with chronic hepatitis B [9]. The policy has led to higher purchase volume of ETV and TDF, consistent with the findings of Wen et al. [4], who evaluated the application of anti-hepatitis B drugs in Shenzhen based on drug centralized procurement in pilot cities. Specifically, Wen et al. reported a 92.85% increase in nucleos(t)ide analogues (NAs) procurement in pilot cities following policy implementation[Ref-Wen]. Our study confirms that this positive trend not only persisted but expanded during the national rollout.

When the NVBP drugs were categorized into wining original, winning generic, non-winning original, non-winning generic, and non-equivalent drugs (unqualified to engage in the NVBP), the results showed that both ETV and TDF had non winning original drugs, but the use of winning generic drugs changed immediately after policy implementation, with an overall stable fluctuation trend. The NVBP policy has resulted in many lower-priced generic drugs replacing original drugs, leading to a decrease in drug expenditure. It not only solves the problem of high drug prices but also improves the overall quality of generic drugs and the development level of the pharmaceutical industry in China [18]. To some extent, it could also promote the innovation of new drug companies. Access to non-patented drugs can reduce the prices of original drugs, thus helping to save drug expenses [21].

While the reform had the most significant impact on drugs within the NVBP (i.e., ETV and TDF), there might also be spillover effects of this policy for NA drugs out-of-NVBP, due to substitution effects. For example, we found increases in the affordability of all four categories of NA drugs outside the scope of the NVBP. with TAF experiencing the largest decrease in unit price. It is worth noting that TAF had the largest decrease in unit price compared to the other three drugs. ETV, TDF, and TAF are the first-line antiviral drugs recommended by guidelines for treatment of HBV infection [9] .and are clinically substitutable. Yuan et al. demonstrated that after the NVBP dramatically lowered the prices of ETV and TDF, TAF manufacturers were compelled to reduce their list prices to forestall further attrition of market share [22].

Overall, the NVBP has effectively reduced the total expenditure and DDDc of NA drugs, implying that the strategy of volume-for-price at the national level is a proven policy to reduce drug prices and mitigate patient cost burdens. The results of this policy echo those of the United Kingdom’s drug tendering policy and India’s “two-envelope” system described earlier. Moreover, it provides reference and inspiration for other countries to curb rising drug prices. For example, the Biden administration in the United States has initiated a policy of negotiating Medicare drug prices in 2024 to reduce the cost of prescription drugs for Medicare patients [23]. This type of policy could refer to the gains and losses of China’s NVBP.

In clinical practice, the 2015 guidelines for the treatment of chronic hepatitis B patients using NAs require two standards to be met: HBV DNA > 2 × 105 IU/ml for HBeAg positive patients or HBV DNA > 104 IU/ml for HBeAg negative patients; and ALT greater than 2 times the upper limit of normal [24]. The 2019 guidelines are revised that antiviral therapy is required for patients with HBV DNA positive and abnormal transaminase, but only on the premise that abnormal liver function caused by other reasons should be excluded [25]. This positive prescription recommendation mainly based on two reasons. Firstly, based on basic research and clinical practice in the past few years, it has been found that antiviral therapy can reduce the risk of liver fibrosis, liver cirrhosis and liver cancer in patients with hepatitis B [12, 26]. At the same time, it is also benefited of the NVBP policy based on the substantial price reduction of nucleoside antiviral drugs, so that more HBV infected patients can be affordable for long-term antiviral therapy. Consistent to guideline recommendation, the doctors guide patients to choose bid-winning products. Bid-winning ETV and TDF replaced non-winning brands. It means that patients were treated with high quality and low-cost NAs.

Declining costs and increased affordability of medicines may have two positive impacts. Our analysis of heterogeneity across the three hospital tiers shows that DDDs rose most in tertiary hospitals and total spending fell most in tertiary hospitals. However, in percentage terms, both DDDs and total expenditures changed most in primary hospitals. These patterns align with evidence from a national interrupted-time-series study on NVBP for ETV and TDF, which demonstrated that, after procurement-driven price reductions, the relative increase in DDDs was significantly larger in primary hospitals than in tertiary hospitals [17]. Similarly, a Singapore ITS study found that once a subsidy policy reduced drug costs, the share of DDDs dispensed in polyclinics (equivalent to primary hospitals) rose from 0.2% to 63% [27]. Taken together, the heterogeneous growth in DDDs across hospital levels suggests an expansion of patient access, indicating that not only are more patients initiating therapy, but treatment is increasingly delivered through primary-care settings, reflecting improved affordability of NA drugs and broader accessibility for the general population.While Wen et al. identified a general increase in generic drug usage, our analysis provides additional granularity by hospital tier, showing that the expansion of access was disproportionately driven by increased uptake in primary-care settings [16].

This conclusion is supported by our data on hospitalization of patients with hepatitis. The literature suggested that the high cost of NA drugs prevents patients with hepatitis B from adequately accessing relevant treatments, which in turn develops into serious illnesses requiring hospitalization [28]. By comparing the rate of change in the number of hospitalizations for each type of liver disease before and after the implementation of the NVBP, we found that the rate of decline in the number of hospitalizations for patients with hepatitis has been much higher than for other diseases since 2020. This might imply that patients receive treatment in a timely manner after lowering the price of NA drugs.

Overall, the NVBP not only reduced patient costs on NA drugs, but might have actually improved patient welfare as well. Zhao, X. et al. estimate that the proportion of CHB patients receiving first-line antiviral therapy will increase by 11.56–15.41% [29]. The previous study also revealed that the implementation of medical policy had an important impact on liver-related death for patients with chronic hepatitis B [30].

Our study has the following strengths. The data used in this study were hospital NA drug prescribing data at the nationwide level, and such data accurately reflect the impact of NVBP on the actual volume and cost of drug prescriptions. Second, we used hospital-level heterogeneity tests and rate-of-change in hospitalization data to explore the impact of NVBP on actual patient welfare, rather than stopping at a discussion of price.

However, there are some limitations to our study. Firstly, there may be other potential factors influencing policy implementation since we used a nation-level dataset. Secondly, the results were based on drug procurement records rather than the actual drug treatments and effects. Although procurement data and purchase volume data are generally consistent under a series of policies, there is still a possible mismatch. Despite these limitations, our study’s evaluation of the national drug centralized procurement policy demonstrates significant improvements in the accessibility of treatment for chronic hepatitis B. It has also reduced the financial burden on patients. This provides valuable insights for the promotion and improvement of similar policies domestically and internationally in the future.

Conclusion

After the implementation of the NVBP, the purchase volume of ETV and TDF showed an increasing trend with generics being the main choice. Meanwhile, their expenditures and DDDc exhibited a decreasing trend. At the same time we found the largest increase in prescribing in primary hospitals and a significant decrease in hepatitis hospitalizations. The NVBP policy in China has shown some effectiveness in lowering prices and overall drug expenditures, as well as improving equity and accessibility. This is particularly crucial for patients in economically underdeveloped areas with poor medical conditions and limited transportation access who require antiviral treatment for hepatitis B. The policy has enhanced medication fairness and accessibility while reducing the financial burden on patients, fully reflecting the significance and original intention of its implementation.

Authors’ contributions

Zhao Yang and Xiao Han contributed to the conception and design of the article, data acquisition and processing; Yue Zhao contributed to article writing and data analysis; Tiantian Yao contributed to article guidance and professional analysis; Fanyu Liu, Binghui Wang, Xiangyi Wu, Yanan Ma, Qiyun Zhu for article retrieval and data organization; Ross-Degnan Dennis for article guidance and professional support; Guiqiang Wang and Bin Jiang for article guidance and revision, and resource support.

Funding

This work was supported by National Science Fundation(72274008/72304006);Beijing Natural Science Foundation (9232010).

Data availability

The data that support the findings of this study are available from National Healthcare Security Administration but restrictions apply to the availability of these data, which were used under license for the current study, and so are not publicly available. Data are however available from the authors upon reasonable request and with permission of National Healthcare Security Administration.

Declarations

Ethics approval and consent to participate

Not applicable. This study has been submitted to the Ethics Committee of Peking University for review and approval. The investigation and research process conforms to the principles outlined in the Helsinki Declaration. This study uses medical institution procurement records, which do not involve patient or personal information, and therefore are exempt from informed consent.

Consent for publication

Not applicable. This study uses medical institution procurement records, which do not involve patient or personal information, and therefore are exempt from informed consent.

Competing interests

The authors declare no competing interests.