How much could be saved in Chinese hospitals in procurement of anti-hypertensives and anti-diabetics?

Abstract

Background: Efficient use of government funding has been increasingly relevant for the success and sustainability of ongoing health-system reform in China; however, as there is no generic substitution policy, patients and basic health-insurance programs pay more for public-preferred brand originators. Such phenomenon is especially typical in public hospitals. The objective of this study is to estimate the potential cost savings in procurement by Chinese public hospitals when switching from brand originators of anti-hypertensive and anti-diabetic medications to their generic equivalents between 2012–2014.

Method: IMS Health volume and value consumption data (IMS China Hospitals Audit system 2012–2014) were used, which covered all Chinese hospitals with 100 beds and above. The top 60% IMS volume consumption of respective anti-hypertensive and anti-diabetic medication with unique dosage form and strength were included. The potential cost savings were calculated from a switch of brand originators with their generic equivalents on the Chinese and international market. An independent sample t-test was conducted to compare the difference of proportion of cost savings in value between the Chinese and international market.

Results: An average of 44% (US$44 million) and 87% (US$90 million) and a total of US$1.4 and 2.8 billion (2014 US$) could be saved from a switch from originator brand anti-hypertensives and anti-diabetics to domestically and internationally available generic equivalents, respectively. The differences of cost savings (in proportion) between domestic and international market were statistically significant (α = 0.005, p = 0.003, p = 0.002, p = 0.000).

Conclusion: Expensive brand originators dominated the anti-hypertensive and anti-diabetic market in Chinese hospitals between 2012–2014. Preference of brand originators wastes a huge amount of health resources in China and these limited resources could have been used more efficiently. As one of the world’s key generic suppliers, if China wants to use its health resource more efficiently on medicines, comprehensive measures are needed to address both demand-side (consumers’ low trust in the quality of local generics) and supply-side barriers (health professionals’ preference of brand originators).

Keywords:

• Brand originator
• generic equivalent
• hypertensives
• diabetics
• volume of consumption
• value of consumption
• switch analysis
• cost savings

Introduction

Under the current health system reform, the Chinese government has been continuously increasing its investment in health, aiming to achieve universal coverage of a quality basic healthcare system. Like in many other countries, efficient use of government funding has been increasingly relevant for the success and sustainability of the reform. Increasingly, the government asks about the efficiency when additional funds for health are requested. Improving use of medicines plays an important role in this effort, as medicines are one of the major drivers of health expenditure.¹

Following the achievement of universal basic health insurance coverage, insurance programs have been the major payers in China. In 2011, the total expenditure of the basic health insurance programs was CNY 641.4 billion (US$ 102, exchange rate = 6.3), accounting for 40% of the total revenue of health facilities. In some areas, this share reached 80–90%.^2,3 With the increasing financial pressure of improving benefit package (including continuously increased payment by insurance, covering more and more high-cost medicines), the basic health insurance programs have been adopting more cost containment measures, including those applicable to medicines.

However, evidence show that patients and the basic health insurance programs in China pay more for originator brands^4,5. Such a phenomenon is especially typical in public hospital. This implies inefficient use of limited health insurance resources, and there are significant economic implications for individuals, insurance programs, and the governments.

This study aimed to produce evidence of the potential cost savings from switching brand originators with their generic equivalents in Chinese public hospitals. We intended to help the Chinese insurance programs to understand the degree and extent of the problem of inefficient use of medicines.

The unique characteristics of the Chinese health system were considered when the study was designed to focus on public hospitals. However, the public providers dominated Chinese service delivery system is a tired system, including tertiary, secondary and primary care. As the primary care is weak and still under development, people directly seek hospital care for common diseases including hypertension and diabetes without referral by lower levels of care. Public hospital care dominates the service delivery system, and most of them offer a full range of medical services, including ambulatory and inpatient care. Medicines are mainly supplied through the hospital pharmacy. In 2014, hospital medicines procurement cost reached CNY 614.2 billion (US$102, exchange rate = 6), accounting for 65% of the total medicines market share (CNY 1089.3 billion, US$182, exchange rate = 6, mark-up rate in hospitals is calculated as an average of 15%)^6,7.

We chose anti-hypertension and anti-diabetes due to the fact that they are expected to be the therapeutics with the top consumption in the Chinese pharmaceutical market in the coming years⁸. China has been experiencing an epidemiological transition from infectious diseases to chronic diseases. There are 177 million adults with hypertension and 114 million adults with diabetes⁹. Given the large numbers who require curative treatment, the emergence of chronic diseases presents special challenges to China. The burden of chronic diseases and associated heathcare costs has been increasing substantially.

Methods

We estimated the savings from a switch of originator brand anti-hypertensive and anti-diabetic medicines to domestically and internationally available generic equivalents in Chinese hospitals.

Data source

The annual volume (Standard Units as defined by IMS) and value (in CNY) consumption of originator brands and their domestically available generic equivalents were obtained from the IMS CHPA database (2012–2014). Data were extracted directly from the Hospital Purchase and Inventory Electronic Systems (sample hospitals across the country, ≥100 beds). The median international buyer unit price (US$per TAB/CAP for oral products, per ML for AMP/VIAL/CARTDGS and per VIAL for single strength injections) of the internationally available generic equivalents (2012–2013) were obtained from the Management Sciences for Health (MSH) International Drug Indicator Guide^10,11.

Targeted medicines

Specific medicines (either brand originators or generics, with unique dosage form and strength) representing 60% of the volume consumptions of three therapeutic groups (anti-hypertensive medicines, insulins and oral anti-diabetic medicines) were chosen within the following IMS ATC classes: anti-hypertensions (C02), diuretics (C03), ß-block agents (C07), calcium antagonists (C08), and agents acting on the rennin-angiotensin system (C09); insulin and analogs (A10A); and oral anti-diabetes (A10B). These medicines were ordered by volume consumption, among which only brand originators were targeted for the study. The volume and value consumption of each targeted brand originator is listed in Supplementary Appendices 1–3.

Among these brand originators, only those with at least one domestic generic equivalent were targeted for switch analysis, and only those having both domestic and international generic equivalents were targeted for cost saving comparison. Brand originator is the first product licensed in China. Generic equivalents are defined as the products that are intended to be interchangeable, manufactured without authorization of the patented manufacturer or exclusivity rights. For each of the targeted brand originators, the domestically and internationally available generic equivalents were identified for analysis.

Calculation of consumption proportion

We calculated the consumption (volume and value) of each brand originator (among the top 60% volume consumption of respective Therapeutic Group) as a proportion of total consumption to its respective Therapeutic Group by dividing the number of units (Standard Units as defined by IMS) and expenditures (US$) of each brand originator with the total number of units and expenditures of its respective Therapeutic Group in the corresponding year (Supplementary Appendices 1–3). We also calculated the consumption (volume and value) of each brand originator as a proportion of the total consumption of its corresponding medicines by dividing the units and expenditures of each brand originator with that of its corresponding medicines for that molecule (with unique dosage form and strength) in the corresponding year (Supplementary Appendices 4–6).

Calculation of unit price of genetic equivalents

For each of the targeted brand originators we identified the corresponding generic equivalents on the domestic market and calculated their annual national average procurement unit price (CNY per TAB/CAP for oral products, per ML for AMP/VIAL/CARTDGS, and per VIAL for VIAL DRY) by having their value consumption (in CNY) divided by their volume consumption (number of Standard Units, defined by IMS) in thecorresponding year. We did not choose the generic equivalent with the lowest price as a comparator because we thought that this may not always be available to each hospital, or may be perceived as low quality. Hence, we calculated the median of annual average procurement unit price of all the generic equivalents on the domestic market for each brand originator in the corresponding year. This is the same case for generic price on the international market. We identified the median international buyer unit price of the generic equivalents of each targeted brand originator in the corresponding year. The number of domestic generic procurement unit prices and international buyer unit prices for each brand originator in each year are presented in Supplementary Appendices 4–6.

Based on Supplementary Appendices 1–6, we calculated the annual average proportion of consumptions (volume and value) to Therapeutic Group and to corresponding medicines by adding the proportions of the targeted brand originators for each year and dividing them by three. We chose a 3-year period instead of an annual period to take into consideration price variation between years. The annual average number of median unit prices of domestically and internationally available generic equivalents were calculated by adding the numbers for each year and dividing by three as well.

Calculation of cost savings

Only brand originators which have domestic generic equivalents were targeted. The reasons of non-existence of generic equivalent are lack of local production of the medicines or its particular strength. Twenty-six brand originators were included for analysis.

To estimate the cost savings from switching with generic equivalents, we adopted the principle of WHO/HAI price study to compare the median local medicines price and the median international reference price¹².

To obtain the cost savings from switching with domestically available generic equivalents, we first multiplied the volume consumed by the originator brand with the median of national annual average procurement unit prices of all their domestically available generic equivalents included in the IMS database. We then took the difference between the total value of brand originators and that of their generic equivalents. Expenditures were converted into US$ using the annual average middle exchange rate issued by the Administration of Foreign Exchange, China, in respective years¹³. Cost savings (US$) of 2012 and 2013 were adjusted for 2014 by US Dollar Implicit Price Deflators for Gross Domestic Product 1929–2014 (2009 = 100)¹⁴. Finally, we obtained the cost savings in 2014 US$.

To estimate the cost savings from switching with internationally available generic equivalents, we excluded the brand originators which do not have an international generic product price in the MSH International Drug Price Guide 2012, 2013, and 2014. We multiplied the volume consumed by the brand originator with the median international buyer unit price of their generic equivalents in respective years. Then, we converted the value consumptions of the brand originators into US$using the annual average middle exchange rate issued by the Administration of Foreign Exchange, China, in respective years. Finally, we took the difference between the total value of brand originators and that of their generic equivalents in US$. Cost savings (US$) of 2012 and 2013 were adjusted for 2014 as well.

Comparison of cost savings between domestic and international switch

Supplementary Appendix 7 presents the comparison of cost savings and proportion of cost savings from domestic and international switch with generic equivalents for brand originators of each Therapeutic Group in each year and all in 3 years (2012–2014). SPSS 20 was used to conduct an independent sample t-test of the proportion of cost savings to compare the difference of cost savings in proportion between domestic and international markets.

Results

During 2012–2014, there were 49 products (with specific dosage form and strength, brand originator or generic) accounting for 60% of the total volume consumption of all anti-hypertensives purchased in Chinese hospitals. Among the top 10 products, seven were brand originators. Except in 2014, all top three were brand originators. For insulins, five-to-six products had 60% of total volume consumption of all insulins purchased in Chinese hospitals, four of which were brand originators. For oral anti-diabetics, 18–19 products accounted for 60% of the total volume consumption of all targeted oral anti-diabetics, seven-to-eight of which were originator brands. Summing up brand originators into three therapeutic groups, the total consumption proportions (volume and value) of all targeted brand originators to the corresponding therapeutic group were 27.2% and 45.9% for anti-hypertensives, 34% and 44% for insulins, and 39% and 57% for oral anti-diabetics per year.

The average proportion of volume and value consumption of corresponding medicines (specific dosage form and strength, including both brand originator and generic equivalents) to respective ATC class (level III) per therapeutic group were 44% and 58% for anti-hypertensives, 89% and 85% for insulins, and 64% and 53% for oral anti-diabetics (Table 1).

Table 1. Proportion of volume and value consumptions of targeted brand originators for switch analysis to respective Therapeutic Group (1–3 as defined by the study) and to corresponded medicines (both brand originators and generics with unique dosage form and strength) (2012–2014 average per year).

INN/Present/Dose of the targeted brand originators (the brand name and manufacturer of the targeted brand originators were not presented)	ATC III	% of volume consumption of brand originator to respective therapeutic group (Groups 1–3)^a	% of value consumption of brand originator to respective therapeutic group (Groups 1–3)^a	% of volume consumption of brand originator to medicines consumption (specific dosage form and strength brand originators and generics)^b	% of value consumption of brand originator to medicines consumption (specific dosage form and strength brand originators and generics)^b	No. of Domesticprices reported^b	No. of Intl prices reported^b
METOPROLTARTRAT TAB 25 MG 20	C07A	7%	1%	90%	92%	6	NA
BISOPROLOL FUMARAT TAB 5 MG 10		1%	2%	53%	70%	4	3
METOPROLOL TARTRAT TAB 50 MG 20		1%	0.2%	92%	91%	5	NA
Sub-total/average per year		9%	3.2%	78%	84%	5	3
AMLODIPINE BESYLAT TAB 5 MG 7	C08A	3%	8%	47%	75%	73	6
NIFEDIPINE TAB GIT 30 MG 7		3%	8%	85%	91%	5	1
FELODIPINE TAB RTD 5 MG 10		2%	7%	81%	92%	10	NA
Sub-total/average per year		8%	23%	71%	86%	8	4
BENAZEPRIL HYDROCH TAB FLM CTD 10 MG 14	C09A	2%	4%	82%	87%	6	NA
FOSINOPRIL TAB 10 MG 14		1%	1%	79%	83%	3	NA
Sub-total/average per year		3%	5%	81%	85%	5	NA
IRBESARTAN TAB FLM CTD 150 MG 7	C09C	2%	4%	54%	72%	13	2
VALSARTAN CAP 80 MG 7		2%	4%	61%	79%	23	NA
TELMISARTAN TAB 80 MG 7/14		1%	2%	70%	84%	9	NA
LOSARTAN TAB 50 MG 7/14		0.8%	1.7%	54%	65%	10	5
Sub-total/average per year		5.8%	11.7%	60%	75%	14	4
HYDROCHLOROTHIAZIDE + IRBESARTAN TAB 162 MG 7	C09D	1%	2%	50%	53%	8	NA
HYDROCHLOROTHIAZIDE + LOSARTAN TAB 62.5 MG 7		0.4%	1%	53%	62%	7	NA
Sub-total/average per year		1.4%	3%	52%	56%	8	NA
All targeted brand originator anti-hypertensives in sum^a or per product		27.2%	45.9%	53%	62%	7	3
70% NPH/isophane insulin crystals & 30% regular insulin PENFILL 100 IU/ML 3 ML	A10C	19%	18%	64%	67%	2	3
70% NPH/isophane insulin crystals and 30% regular insulin PENFILL 100 IU/ML 3 ML		5%	5%	17%	17%	2	3
50% NPH/isophane insulin crystals and 50% regular insulin PENFILL 100 IU/ML 3 ML		5%	5%	93%	93%	2	NA
INSULIN GLARGINE 300 IU 3 ML		5%	16%	62%	69%	1	NA
All targeted brand originator insulins in sum^a/inaverage^b or per product		34%	44%	59%	62%	2	3
ACARBOSE TAB 50 MG 30	A10L	17%	27%	68%	76%	5	NA
VOGLIBOSE TAB 0.2 MG 30		1%	1%	37%	41%	17	NA
Sub-total/average per year		17%	27%	53%	59%	11	NA
METFORMIN TAB 500 MG 20	A10J	9%	8%	86%	92%	11	5
REPAGLINIDE TAB 1 MG 30	A10M	5%	8%	91%	95%	2	NA
GLICLAZIDE TAB SR 30 MG 30/60	A10H	3%	4%	81%	88%	21	1
GLIMEPIRIDE TAB 2 MG 15/30		2%	6%	44%	76%	27	2
GLICLAZIDE TAB 80 MG 60		1%	1%	52%	68%	42	5
GLIPIZIDE TAB 5 MG 14		1%	2%	79%	88%	4	NA
Sub-total/average per year		7%	13%	64%	80%	24	3
All targeted brand originator oral anti-diabetics in sum or in average		38%	56%	67%	78%	16	3

^a % in sum;

^b in average.

An average of 45% (US$45 million) and a total of US$1.8 billion (2014 US$) could be saved for anti-hypertensives; 16% (US$23 million) and US$203 million (2014 US$) for insulins, and 57% (US$48 million) and US$947 million for oral anti-diabetics. Accounting for all three therapeutic groups together in 3 years, an average of 39% (US$38 million) and a total of US$2.9 billion (2014 US$) could be saved (Table 2). Supplementary Appendix 8 presents the cost savings for each Therapeutic Group in each year and all in 3 years.

Table 2. Potential savings from a switch of targeted brand originators to domestically available generic equivalents for anti-hypertensive and anti-diabetic medicines (2012–2014).

Therapeutic group	Total savings (2014 US$)	Average savings per product (2014 US$)	Average % costs saved per product
Anti-hypertensive medicines	1793706 351	44 807 703	45%
Insulins	203 214 038	22 579 338	16%
Oral anti-diabetics medicines	946 733 365	47 672 749	57%
All therapeutic groups	2 943 653 754	38 353 263	39%

When looking at the brand originators which have both domestic and international generic equivalent prices, we found that an annual average (per product) cost savings of five anti-hypertensive brand originators were 49% (US$64 million) domestically and 85% (US$92 million) internationally, and a total of US$881 million and 1.5 billion (2014 US$), respectively, in 3 years. That for two insulin originator brands were 15% (US$22 million) domestically and 86% (US$120 million) internationally in average years, and US$134 million and 720 billion (2014 US$), respectively, in 3 years; that for five originator brand oral anti-diabetics were 67% (US$44 million) domestically and 91% (US$59 million) internationally in average year, and US$410 and 550 million (2014 US$), respectively, in 3 years. Counting the 12 originator brands together, there was an average of 44% (US$44 million) and 87% (US$90 million) and a total of US$1.4 and 2.8 billion (2014 US$) cost savings from domestic and international switch, respectively (Table 3).

Table 3. Comparison of potential savings from a switch of brand originators which have both domestic and international generic equivalents for anti-hypertensive and anti-diabetic medicines (2012–2014).

			Domestic market		International market
			Savings(2014 US$)	% costssaved	Savings(2014 US$)	% costssaved
5 anti-hypertensives targeted for comparison	2012	All products	217,664,609		393,615,265
		Average product	72,554,869	48%	131,205,089	97%
	2013	All products	322,519,503		480,843,467
		Average product	64,503,901	47%	96,168,694	71%
	2014	All products	340,890,505		601,190,976
		Average product	56,057,956	52%	47,700,345	87%
	2012–2014 all products (2014 US$)		881,074,617		1,475,649,708
	2012–2014 average product (2014 US$)		64,372,242	49%	91,91 376	85%
2 insulins	2012	All products	46,558,468		233,757,085
		Average product	23,279,235	16%	116,878,543	84%
	2013	All products	42,543,679		244,207,287
		Average product	21,271,840	14%	122,103,644	86%
	2014	All products	44,695,804		242,216,626
		Average product	22,347,902	16%	121,108,313	88%
	2012–2014 all products (2014 US$)		133,797,951		720,180,998
	2012–2014 average product (2014 US$)		22,299,659	15%	120,030,167	86%
5 oral anti-diabetics	2012	All products	57,664,712		79,776,351
		Average product	28,832,356	72%	39,888,175	89%
	2013	All products	173,758,542		251,823,403
		Average product	43,439,636	58%	62,955,851	94%
	2014	All products	178,866,357		222,900,375
		Average product	59,622,119	70%	74,300,125	90%
	2012–2014 all products (2014 US$)		410,289,611		554,500,129
	2012–2014 average product (2014 US$)		43,964,704	67%	59,048,050	91%
All 12 products	2012–2014 all products (2014 US $)		1,425,162,178		2,750,330,835
	2012–2014 average product (2014 US$)		43,545,535	44%	90,256,531	87%
	Total national basic health insurance expenditure(2012–2014) (2014 US$, billion) (%)		475.70 (0.3)		475.70 (0.6)
	Total national pharmaceutical expenditure(2012–2014) (2014 US$, billion) (%)		655.07 (0.2)		655.07 (0.4)
Difference between domestic and international market (Independent sample t-test)			% cost saved (p-value)
	2012		0.003
	2013		0.002
	2014		0.000

The difference of cost savings in proportion was statistically significant (α = 0.005, p = 0.003, p = 0.002, p = 0.000) (Table 3). The total domestic and international cost savings account for 0.3% and 0.6% of the total national basic health insurance expenditures, and 0.2% and 0.4% of the total national pharmaceutical expenditures during 2012–2014^15–20 (Table 3).

Discussion

China announced the achievement of universal health coverage with basic health insurance in 2009. With continued efforts of enhancing the benefit package (including the medicines package), more and more medicines are covered by the basic health insurance program, which greatly improved the financial risk protection of individuals. However, China is a country with a large population and huge disparity in social and economic development within a vast territory. The per capita total health expenditure of China was only $578 (PPP int. $), which was lower than the average of upper middle income countries ($766) and far below the average of high income countries ($4516) in 2012²¹. As the second largest market in the world, China’s spend on pharmaceuticals is just 9% per capita of that in the US²². Using health resources efficiently is critical for China, with limited per capita health resources. Our study results show that there is a potential for efficiency gains in the area of medicines procurement in hospitals. Brand originators dominated the hospital medicines market, despite that China has being considered the “pharmacy to the world”, with numerous generic suppliers. The reasons behind the preferences of brand originators include:

(1) Financial incentives: Government funding only accounts for a small proportion of the total revenue for public hospitals. Medical service price has been controlled by the government at a very low level. Public hospitals are allowed to generate revenue through fixed mark-up on medicines (this is changing). Over prescription and preference of expensive medicines were driven by the mixed affections of the above policies, and, therefore, are a common phenomenon in public hospitals.

(2) Reimbursement policies: The health insurance programs reimburse medicines in a fixed proportion of expenditures through a fee for service provider payment method. There is no generic substitution policy in China. These policies create no incentives to either patients or service providers to use lower priced medicines^23,24.

(3) Quality of generics: More importantly, although the majority of the Chinese pharmaceutical companies mainly produce generics, the public confidence of the quality of local generics is low. A centralized and independent single regulatory system was established following a series of government restructures during 1998–2013, which eliminated the conflicting standards that prevailed among provincial government agencies before the centralization through upgrading quality standards and strengthening inspections, etc. However, pitfalls were already created, thousands of local generics were already marketed before these efforts. In 2013, the State Food and Drug Administration (SFDA, now named China Food and Drug Administration, CFDA) formally announced to re-evaluate the generics they approved before 2007 (the New Regulation on Drug Registration was implemented in 2007 when registration of generics was fully regulated), to secure that the clinical efficacy of generics is consistent with that of the brand originators²⁵. This is also the key content of the 12th National Five-Year Plan for Drug Safety, which was developed in 2011²⁶. The 2013 initiative indicates the potential problems of quality and efficacy of local generics; Chinese health professionals and patients have been concerned about the quality of generics. Some studies disclosed the quality and efficacy gaps between local generics and brand originators²⁷. Before 2007, the reference standards of many generics registration applications were not fully regulated, as some reference standards were standards of generics rather than the brand originators. This definitely led to the unavoidable quality and efficacy gaps between generics and brand originators.

In many high income countries, there are comprehensive national and local price and reimbursement policies promoting the uptake of generics²⁸. Despite the lack of evidence from low and middle income countries regarding which pro-generic policy is more effective, ensuring a functioning drug regulatory authority and creating appropriate incentive for physicians, consumers and drug sellers are still the top tasks and the pre-requisites for promoting generics in these countries including China²⁹. It is expected that CFDA’s initiative to improve the quality of generics, the new medicines price policy under development by the basic health insurance programs, and the prospective provider payment pilots can create positive incentives for using medicines more cost-effectively.

China is the world supplier of generics, there are many generic equivalents available on the domestic market, and there are huge potential cost savings if the consumed brand originators are switched with lower priced local generic products. This can also be achieved and saved even more through switching with foreign generic equivalents on the international market. However, as foreign medicines must be registered with the national drug regulatory authority before entering into the Chinese market, and the competition of generics in China is already very fierce, few foreign pharmaceutical companies import generics to China. Adding that, the registration fee for imported generics is the same as that for imported new moleculars, which is 30-times that for the domestic generics³⁰; international generics suppliers like India do not have the interests to register their generics in China. Even if there are international generics registered in China, according to Chinese Government Procurement Law they may not be eligible as a bidder for the current provincial pooled medicines procurement for public hospitals. Chinese public hospitals can only procure from the international market when there is no domestic competitor. However, it should be noted that medicines procurement is not government procurement in China; even if it is managed by the provincial government, the payers of the contract are hospitals but not government.

The cost calculations are based on both locally-registered generic alternatives and those on the international market. The reason for us to do the shift with generic equivalents on the international market is that, first, showing a lower international price and presenting the potential savings will give implications of inefficiency of medicines procurement; second, this also implies that, if the medicines procurement no longer follows the Government Procurement Law, and the regulatory authority enables equal competition between domestic and international generic suppliers, there will be more cheaper generic equivalents available for Chinese patients, and a significant amount of public funding as well as individual expenditures can be saved.

In addition to the chemical products, as in many high-income countries, the increasing interests from leading pharmaceutical companies, multinational generics manufacturers and biotechnology companies also drive a strong growth for the biosimilar market in China. There are also huge spaces for cost savings in this area.

Although such savings only account for a very small amount of the national medicines expenditure, it already saves more than half of the medicines procurement budget for two types of common non-communicable diseases. If other categories of medicines are also included in the shift, which address the diseases with rapidly increasing incidence, the total potential cost savings will be tremendously huge.

The medicines expenditure per capita in China (CNY 963.69 in 2013, US$160, exchange rate = 6)³¹ is still far below the level in high income countries (US$285–983 in 2010)³²; such shifts and savings will greatly help to improve the overall affordability of medicines of the country.

This study only presented the cost savings through shifting of generic equivalents available on the market. We do not have the intention to suggest that medicines price should be “the lower the better”. To secure a sustainable development and a viable generics industry, there should be a balance between reasonable margins for the industry and the affordability and efficiency for public funding.

Limitations

There are several limitations of this study. The IMS data is procurement data rather than actual use. We neglected the possible unused medicines, which might flat the volumes and savings. We also ignored the average 15% mark-ups on medicines in Chinese hospitals (although there has been an increasing number of hospitals removing the mark-up under the national health system reform), which might lead to under-estimation of the savings. In addition, our estimation was based on the brand originators accounting for 60% of the mostly used products, not all of the brand originators, which further under-estimated the savings. Second, the IMS CHPA database only collected data from hospitals (with 100 beds and above) but excluded retail drug stores and an increasing number of primary care facilities, in which there have been increasing efforts in retaining the patients with chronic long-term conditions. In general, our analysis excluded 35% of the total medicines market share in the retail pharmacies and the primary care facilities. As hypertension and diabetic patients are the most targeted patients of the above efforts, the excluded proportion of market share for anti-hypertensive and anti-diabetic medicines should be higher than 35%. Third, IMS CHPA data were collected from sample hospitals, and extrapolated to the national level, which will introduce some error margin. However, this is the best estimation of the hospital market consumption available. Fourth, we used the median price rather than the lowest price for alternative generics to calculate the potential savings. The lowest price may not be feasible to obtain which may result in under-estimation of total saving. Last, we did not have patient-level drug price or clinical data to evaluate the economic and clinical impacts of the changing trends in hypertension and diabetic medicines treatment at individual level, as this was not our objective. Because our analysis was based on pharmaceutical sales, we do not have necessary data to assess the appropriateness of medication use.

Conclusion

Expensive brand originators dominated the consumption of anti-hypertensives and anti-diabetics in Chinese hospitals during 2012–2014. Preference of brand originators wastes huge health resources, which could have been used more efficiently. As one of the key world suppliers of generics, if China wants to use its health resource more efficiently, comprehensive measures are needed to address both the demand side (consumers’ low trust on the quality of generics) and the supply side barriers (health professionals’ preference of brand originators) for generic uptake.

Transparency

Declaration of funding

No funding was declared for this manuscript.

Declaration of financial/other relationships

The authors report no conflicts of interest, financial or otherwise. JME peer reviewers on this manuscript have no relevant financial or other relationships to disclose.

Acknowledgments

The authors acknowledge Dr Peter Stephen who helped JS and VW, and the two commissioners of the Lancet Commission Essential Medicines and initiators of the conception of this analysis, in co-ordinating internally within IMS, and bridging the collaboration between the two commissioners with RL. Dr Stephen also made valuable comments to the revision of the manuscript. We also thank Mr Xing Hu from IMS Health China, who helped us in understanding the rules and definitions of the IMS CHPA database, the medicines classification system of IMS, and answered many other enquiries from the authors about the data.