- Split View
-
Views
-
Cite
Cite
Bin Jiang, Ruo Jing Zhou, Xing Lin Feng, The impact of the reference pricing policy in China on drug procurement and cost, Health Policy and Planning, Volume 37, Issue 1, January 2022, Pages 73–99, https://doi.org/10.1093/heapol/czab012
- Share Icon Share
Abstract
High drug costs are putting pressure on healthcare budgets and posing an obstacle for China to achieve universal health coverage. Policies such as the direct price ceiling and the Essential Medicines Program—with the Zero Markup Drug Policy (ZMDP) one key component—have been implemented, but out with limited evidence of success. As a benchmark of China’s recent health reform, Sanming city initiated the ZMDP in January 2013; and further piloted the first reference pricing (RP) policy in China in September 2014, with the intention to disincentivize the use of costly original drugs. In this study, we used hospital-based drug procurement data of 14 drug substances that were subjected to RP, from four hospitals in Sanming and a neighbouring city Longyan, between 2012 and 2016. Adopting the difference-in-difference (DID) approach, we evaluated the impacts of RP together with the ZMDP. On the one hand, we found that the ZMDP had no impact on drugs’ procurement prices, volumes and costs. While on the other hand, we found that the introduction of RP was not associated with changes in unit prices for the 14 drugs in Sanming. However, the RP pilot was associated with a 25.9% (95% confidence interval (CI), 12.9–37.0%) decrease in monthly drug procurement volumes and a 47.7% (95% CI, 33.7–58.7%) decrease in the total drug costs. In particular, it reduced the procurement volumes of original drugs by 56.8% (95% CI, 47.0–64.7%). Subgroup analyses by hospital level and therapeutic class found similar results. We draw lessons for the Chinese government to experiment with RP on a larger scale, considering the development and effective regulation of the generic market. This is a first report on the effects of RP in China, Asia and middle-income countries.
Reference pricing (RP) is an effective policy lever for drug cost containment in high-income countries. But little is known about its implementation in low- and middle-income settings and Asia.
Using hospital-based drug procurement data and adopting the difference-in-difference approach, we found that the RP pilot in Sanming China contributed to a 25.9% decrease in total volumes, and a 47.7% decrease in total costs, but seemed to have no effect on prices.
This is a first report of the impact of RP in China. The RP pilot in Sanming demonstrates great potential in controlling drug costs. Lessons could be drawn for the government to experiment with RP on a larger scale.
Introduction
High drug costs are putting pressure on healthcare budgets and posing an obstacle for China to achieve universal health coverage (Yip and Hsiao, 2014). The costs of drugs accounted for 30–40% of the nation’s total health expenditures in 2018 (National Health Commission of China, 2018). This figure is much higher than the OECD countries’ average (16.4%), the USA (12.0%), and Japan (18.6%) and Korea (20.7%) from Asia (OECD, 2018). Causes were attributed to the absence of effective pricing strategies and distorted financial incentives by healthcare providers in overprescribing high-priced drugs (Yip et al., 2012; Yi et al., 2015; Hu and Mossialos, 2016).
In the past three decades, two tides of policies, namely, the direct price ceiling and the Essential Medicines Program, have been implemented in China, with limited evidence for their success (Yip and Hsiao, 2014; Hu and Mossialos, 2016). Since the early 1990s, the Chinese government has made over 30 rounds of efforts that directly set the price ceiling for each individual brand of drug (Wu et al., 2015). These efforts seemed not to be successful in achieving the policy objectives, which might be attributed to the government’s inability to measure real manufacturing costs, or to healthcare providers’ reliance on markups from drug sales (Wu et al., 2015). Consequently, drug costs continued to increase at an annual rate of 14% from 2000 to 2010 in China, and became unaffordable to the public (Hu and Mossialos, 2016). Since 2009, China has nationally promoted the Essential Medicines Program, which consists of three components (Yip et al., 2012; Ding and Wu, 2017). First, the central government made an essential medicine list referring to generic names. It is compulsory for public hospitals to provide these drugs, and for public health insurance schemes to reimburse the costs. Second, each province established its own procurement system to dynamically determine the suppliers and unit prices for the listed drugs. The procurements are annually conducted, and for each generic name, one original drug brand and several generic brands are often selected. Public hospitals must purchase from this procuring list for no more than the unit prices. Third, a Zero Markup Drug Policy (ZMDP) was gradually put forward to regulate the listed drugs’ retail prices, where all public hospitals were prohibited from adding additional retail markups to the procuring price, which used to be officially set as 15% (Yip et al., 2012; Yi et al., 2015).
Targeting those drugs included in the Essential Medicines Program, the ZMDP aims to encourage the use of low-priced drug brands under the same generic name, and to control unnecessary over-prescriptions, whose effects have been intensively investigated previously (Li et al., 2013; Yang et al., 2013; Chen et al., 2014; Ding and Wu, 2017; Wei et al., 2017; He et al., 2018; Li et al., 2018; Zeng et al., 2019). However, due to methodological inconsistencies, for example in the definition of indicators and the lack of retail data from hospitals’ pharmacies, the effectiveness of this policy remains unclear (Yi et al., 2015; Ding and Wu, 2017; Wei et al., 2017; Fu et al., 2018) (see Appendix 12 for a thorough literature review). Particularly, the drug costs were in fact constituted by drug prices and prescription volumes, but indicators were too broadly defined previously. In China, few studies have been conducted to investigate the association of the existing drug policies with these two cost components by focusing on specific drugs (Song et al., 2014; Tang et al., 2018), thereby limiting the explanatory power of evidence to inform policy.
The systemic health system reforms in Sanming City have been benchmarked by the Chinese government (Office of State Council of China, 2019). In addition to the ZMDP, Sanming has piloted a reference pricing (RP) policy since September 2014, which was the first in China. In this study, we used hospital-based drug procurement data from Sanming, and a neighbouring city Longyan, between 2012 and 2016 to evaluate the impacts of RP, together with the ZMDP. Adopting a natural experimental design, we provided evidence that the implementation of RP was associated with substantial decreases in drug costs, while the ZMDP seemed not to be. This is a first empirical report on the impacts of RP on drugs’ prices, volumes and costs from China and other Asian countries. Based on the difference-in-difference (DID) approach and using comparative data by all drug types that were subjected to RP, we add more consistent and stronger evidence to RP’s applicability from the low- and middle-income world as well.
Methods
Settings
Locating in Fujian Province, East Southern China, Sanming is a city at the prefectural level that administers nine counties. In 2014, Sanming had a GDP per capita of 9227 US dollars and accommodated a population of 2.51 million. In China, each prefectural city has one general hospital at the tertiary level that functions as the regional referral centre; and each county has one general hospital at the secondary level which is capable of providing most general diagnosis and treatment services (Office of State Council of China, 2015). Since there is only one tertiary hospital—that is, the Sanming Prefecture General Hospital—and Sanming is the only city that piloted RP policy, we included just one tertiary hospital in this study. In addition, we included in the intervention group the only one secondary hospital from the largest county of Sanming—that is Youxi, which accommodated a population of 0.35 million in 2014. As the control to Sanming, we selected Longyan, a neighbouring prefectural city, which had a per capita GDP of 8961 US dollars and a population of 2.59 million in 2014. Accordingly, we chose a control for Youxi County — the Shanghang County from Longyan, which is 220 km away from Youxi and had a population of 0.37 million in 2014.
RP was piloted in all hospitals’ inpatient sectors in Sanming, with an intention to control health insurance’s reimbursement and to disincentivize the use of costly original drugs (normally imported drugs). Fifteen drug substances, according to the generic names (International Non-proprietary Names, INN), were chosen because they had the highest historical prescription volumes. In implementation, the RP for each drug substance was officially set to the lowest procurement price for each generic name, and social health insurance schemes only reimbursed hospitals at this price for drugs under the same generic name, irrespective of the brands.
RP was introduced in Sanming in September 2014. And the ZMDP was prompted in both Sanming and Longyan in 2013, one year earlier than RP, but the timing of introduction varied across the two cities and across different levels of hospitals. Of note, the ZMDP initiated in all hospitals in Sanming in January 2013. While at the same time, in Longyan, the ZMDP was only introduced in county hospitals, i.e. secondary hospitals. And in June 2015, the ZMDP eventually came out to the prefectural (tertiary) hospitals in Longyan. Figure 1 shows the chronology of the relevant policies in the two cities, respectively. Our data covers the years 2012–2016, which thus provides three important time windows that shed light on this evaluation. First, during the years 2012–2013, there were no policies in either of the two cities. Second, from 2013 to September 2014, the ZMDP started in all hospitals in Sanming but only in the secondary hospitals in Longyan. Third, after September 2014, RP initiated in the inpatient sector of all hospitals in Sanming, which coexisted with the ZMDP thereafter.
Study design
We adopted a natural experimental design with the DID approach. Given that the ZMDP and RP were not simultaneously introduced in Sanming, and the city already had the ZMDP when RP was initiated, we conducted this evaluation in two steps. In the first step, we only used data from January 2012 to September 2014, when RP had not yet been implemented, to evaluate the independent effects of the ZMDP. To perform the DID, we set up the timing of the policy initiation to January 2013, when the ZMDP started. Because both prefectural and county hospitals in Sanming initiated the ZMDP in 2013, while only county hospitals in Longyan—the control city—did so, we only included the two prefectural tertiary hospitals in the intervention and control groups, respectively. This analysis came out with no impact of the ZMDP on any outcome variables. In the second step, we used data from all hospitals in the whole study period, i.e. January 2012 to December 2016, and set up the timing of the policy initiation to September 2014 when RP was prompted, to evaluate the impact of RP.
Data and indicators
We extracted procurement data of the 15 drug substances that were subjected to RP during January 2012 to December 2016 from the four sample hospitals’ Hospital Information System (HIS). Such databases provided information on each hospital’s procurement records for each drug substance under specific brand names. The procurement records were separated by each hospital’s inpatient and outpatient sector, because the two sectors’ pharmacies were separately operated. The procurement records contained information on each drug’s generic name; dosage form; specification; package; suppliers; whether it is an original or generic version; and its procurement date, price, volume and cost. We converged single procurement records into monthly data. We examined each drug’s procurement price by measuring the Defined Daily Dose (DDD) unit price, in order to standardize price comparisons across drug substances. DDD is an international standard that is used to eliminate diversity in drug packaging, usage and dosage (WHO, 1981b). We measured drug volumes in DDDs and drug costs in absolute monetary values.
In addition, previous studies that investigated the impact of ZMDP, for example Yi et al. (2015), found that such a policy may have unintended spill-over effects that incentivize more hospital visits (see Appendix 12 for a literature review). We therefore extracted hospital visit data from the same hospitals to ascertain RP’s potential effects on inpatient admissions and outpatient visits.
Empirical strategies
We performed three types of subgroup analyses. First, previous studies showed that China’s health providers may have distorted incentives to generate revenues (Yi et al., 2015; Ding and Wu, 2017). RP was only piloted in hospitals’ inpatient sectors in Sanming; however, because in Chinese hospitals both inpatient and outpatient sectors share the same economic incentives, we also evaluated the effects of RP in the outpatient sector, in order to examine possible spill-over effects. Second, we performed the analysis stratifying by prefectural and county hospital, because despite the varied timing of policy initiation, these hospitals serve different functions and may vary in patient type and present different patterns in drug prescriptions. Third, we stratified by therapeutic classifications of the drug substances, because effects of drug pricing/reimbursement policies may vary across drugs which have different costs and demands. These 15 drug substances were eventually classified into six therapeutic classes in accordance with the World Health Organization’s Anatomical Therapeutic Chemical (ATC) classifications (WHO, 1981a): namely, drugs used in diabetes, antibacterials for systemic use, drugs used for blood and blood-forming organs, drugs used for cardiovascular system diseases, drugs for acid-related disorders and drugs for respiratory system diseases.
To tease out the possible long-term effects of the ZMDP and other confounding factors, we conducted several sensitivity analyses. First, given the fact that RP was only piloted in hospitals’ inpatient sectors in Sanming, we pooled data from the sample hospitals’ inpatient and outpatient sectors and performed time–varying DID analysis. This approach was used to evaluate the impacts of RP and the ZMDP under a unified framework, but assumed that drug demands were comparable between the inpatient and outpatient sector, as well as between the tertiary and secondary hospitals (Appendix 6). Second, we excluded data before January 2013 when the ZMDP had not yet been implemented. We assumed that if the ZMDP confounded the effects of RP, the estimated impacts would be different in this subsample than in the full sample (Appendix 7). Third, we only included January 2012–January 2013 as the pre-intervention period, and June 2015–December 2016 as the post-intervention period, and compared the estimation to that from the full sample. We assumed that if the ZMDP has lagged effects, the estimation of RP should be different on this occasion to that with the whole timing sets, and the estimated impact should be different by different level of hospitals because the ZMDP was initiated in Longyan’s secondary hospitals much earlier than in its tertiary level ( Appendix 8). Fourth, we performed placebo tests by artificially changing the timing of RP’s implementation step-wised by each month, from six months before to six months after the real timing of the event (Appendix 9).
Results
Overall, we extracted 13 440 monthly data points on drug procurement records and 480 monthly data points on hospital visits from the four hospitals from January 2012 to December 2016 (60 months in total). We only included 14 drug substances that were subjected to the RP policy in Sanming because the original version of amoxicillin sodium and clavulanate potassium (injection) was not procured during the study periods by any of the four hospitals.
Trends in drug prices
In Sanming, for most drug substances subjected to RP, there was only one supplier for each original drug and 4–10 suppliers for the generic versions. As shown in Table 1 and Appendix 2, there were no obvious trends in the unit prices which were associated with the implementation of RP or the ZMDP for any of the 14 original drugs. For example, the price of the original version of acarbose was 12.4 (range, 12.4–12.4) Chinese Yuan (CNY) before and after the implementation of these policies. In addition, for the 14 drug substances in Sanming, the price differences between the original drugs and their generic versions varied from 0.1 CNY to 106.8 CNY in absolute forms and 5% to 2000% in relative forms.
. | . | . | Original drugs . | Generic drugs . | ||||||
---|---|---|---|---|---|---|---|---|---|---|
. | . | . | . | Before RP . | After RP . | . | . | . | ||
International nonproprietary names . | Pharmacology classification . | Therapeutic subgroup . | Number of producers . | Median . | Range . | Median . | Range . | Number of producers . | Median . | Range . |
Acarbose | Blood glucose–lowering drugs, excluding insulins | Drugs used in diabetes | 1 | 12.4 | (12.4–14.2) | 12.4 | (12.4–12.4) | 3 | 7.7 | (7.7–8.1) |
Cefoperazone and sulbactam | Other beta-lactam antibacterials | Antibacterials for systemic use | 1 | 183.7 | (183.7–211.2) | 183.7 | (183.7–183.7) | 1 | 11.4 | (11.4–11.4) |
Cefuroxime sodium (injection) | Other beta-lactam antibacterials | 2 | 103.2 | (103.2–103.2) | 103.2 | (103.2–103.2) | 7 | 4.8 | (4.6–19.3) | |
Aspirin | Antithrombotic agents | Drugs used for blood and blood–forming organs | 1 | 0.4 | (0.4–0.5) | 0.4 | (0.4–0.4) | 8 | 0.4 | (0.3–0.4) |
Peritoneal dialysis solution | Peritoneal dialytics | 1 | 3.1 | (3.1–3.1) | 3.1 | (3.1–3.1) | 3 | 2.5 | (2.5–2.9) | |
Human albumin (injection) | Blood substitutes and plasma protein fractions | 1 | 1132.5 | (1132.5–1132.5) | 1132.5 | (1132.5–1260.0) | 4 | 1080.0 | (1080.0–1260.0) | |
Mecobalamin (injection) | Vitamin B12 and folic acid | 1 | 6.0 | (6.0–7.2) | 6.0 | (5.7–6.0) | 3 | 1.3 | (1.1–6.0) | |
Clopidogrel | Antithrombotic agents | 1 | 16.5 | (16.5–19.0) | 15.5 | (15.5–16.5) | 2 | 6.6 | (6.6–8.6) | |
Atorvastatin calcium | Lipid–modifying agents, plain | Drugs used for cardiovascular system | 1 | 8.6 | (8.6–9.9) | 7.9 | (7.9–8.6) | 3 | 5.7 | (5.7–8.9) |
Amlodipine | Selective calcium channel blockers with mainly vascular effects (CCBs) | 1 | 4.7 | (4.7–5.4) | 4.7 | (4.3–4.7) | 10 | 0.2 | (0.2–2.2) | |
Nifedipine (controlled release) | Selective calcium channel blockers with mainly vascular effects (CCBs) | 1 | 4.3 | (4.3–5.1) | 3.8 | (3.8–4.3) | 5 | 2.8 | (2.3–2.8) | |
Omeprazole (injection) | Drugs for peptic ulcer and gastroesophageal reflux disease (PPIs) | Drugs for acid-related disorders | 1 | 43.5 | (43.5–61.0) | 41.0 | (41.0–44.0) | 9 | 0.7 | (0.7–70.2) |
Pantoprazole (injection) | Drugs for peptic ulcer and gastroesophageal reflux disease (PPIs) | 1 | 108.7 | (108.7–125.0) | 108.7 | (108.7–108.7) | 5 | 2.0 | (2.0–17.1) | |
Ambroxol hydrochloride (injection) | Expectorants, excluding combinations with cough suppressants | Drugs for respiratory system | 1 | 16.2 | (16.2–18.6) | 16.2 | (15.3–16.2) | 7 | 8.0 | (8.0–13.7) |
. | . | . | Original drugs . | Generic drugs . | ||||||
---|---|---|---|---|---|---|---|---|---|---|
. | . | . | . | Before RP . | After RP . | . | . | . | ||
International nonproprietary names . | Pharmacology classification . | Therapeutic subgroup . | Number of producers . | Median . | Range . | Median . | Range . | Number of producers . | Median . | Range . |
Acarbose | Blood glucose–lowering drugs, excluding insulins | Drugs used in diabetes | 1 | 12.4 | (12.4–14.2) | 12.4 | (12.4–12.4) | 3 | 7.7 | (7.7–8.1) |
Cefoperazone and sulbactam | Other beta-lactam antibacterials | Antibacterials for systemic use | 1 | 183.7 | (183.7–211.2) | 183.7 | (183.7–183.7) | 1 | 11.4 | (11.4–11.4) |
Cefuroxime sodium (injection) | Other beta-lactam antibacterials | 2 | 103.2 | (103.2–103.2) | 103.2 | (103.2–103.2) | 7 | 4.8 | (4.6–19.3) | |
Aspirin | Antithrombotic agents | Drugs used for blood and blood–forming organs | 1 | 0.4 | (0.4–0.5) | 0.4 | (0.4–0.4) | 8 | 0.4 | (0.3–0.4) |
Peritoneal dialysis solution | Peritoneal dialytics | 1 | 3.1 | (3.1–3.1) | 3.1 | (3.1–3.1) | 3 | 2.5 | (2.5–2.9) | |
Human albumin (injection) | Blood substitutes and plasma protein fractions | 1 | 1132.5 | (1132.5–1132.5) | 1132.5 | (1132.5–1260.0) | 4 | 1080.0 | (1080.0–1260.0) | |
Mecobalamin (injection) | Vitamin B12 and folic acid | 1 | 6.0 | (6.0–7.2) | 6.0 | (5.7–6.0) | 3 | 1.3 | (1.1–6.0) | |
Clopidogrel | Antithrombotic agents | 1 | 16.5 | (16.5–19.0) | 15.5 | (15.5–16.5) | 2 | 6.6 | (6.6–8.6) | |
Atorvastatin calcium | Lipid–modifying agents, plain | Drugs used for cardiovascular system | 1 | 8.6 | (8.6–9.9) | 7.9 | (7.9–8.6) | 3 | 5.7 | (5.7–8.9) |
Amlodipine | Selective calcium channel blockers with mainly vascular effects (CCBs) | 1 | 4.7 | (4.7–5.4) | 4.7 | (4.3–4.7) | 10 | 0.2 | (0.2–2.2) | |
Nifedipine (controlled release) | Selective calcium channel blockers with mainly vascular effects (CCBs) | 1 | 4.3 | (4.3–5.1) | 3.8 | (3.8–4.3) | 5 | 2.8 | (2.3–2.8) | |
Omeprazole (injection) | Drugs for peptic ulcer and gastroesophageal reflux disease (PPIs) | Drugs for acid-related disorders | 1 | 43.5 | (43.5–61.0) | 41.0 | (41.0–44.0) | 9 | 0.7 | (0.7–70.2) |
Pantoprazole (injection) | Drugs for peptic ulcer and gastroesophageal reflux disease (PPIs) | 1 | 108.7 | (108.7–125.0) | 108.7 | (108.7–108.7) | 5 | 2.0 | (2.0–17.1) | |
Ambroxol hydrochloride (injection) | Expectorants, excluding combinations with cough suppressants | Drugs for respiratory system | 1 | 16.2 | (16.2–18.6) | 16.2 | (15.3–16.2) | 7 | 8.0 | (8.0–13.7) |
Notes: Procurement price per DDD for each drug was reported in Chinese Yuan, CNY. In 2014: 1 United States dollar = 6.14 Chinese yuan. RP policy for 15 drug substances was piloted in Sanming, China in September 2014 to December 2016. Amoxicillin sodium and clavulanate potassium (injection) are excluded because of no procurement of its original drug in the study period. Dosage forms are specifically marked in parentheses for international nonproprietary names (INNs) that were not oral regular-release dosage forms. Pharmacology classification is based on ATC-3 codes and therapeutic subgroup is based on ATC-1 or ATC-2 codes.
. | . | . | Original drugs . | Generic drugs . | ||||||
---|---|---|---|---|---|---|---|---|---|---|
. | . | . | . | Before RP . | After RP . | . | . | . | ||
International nonproprietary names . | Pharmacology classification . | Therapeutic subgroup . | Number of producers . | Median . | Range . | Median . | Range . | Number of producers . | Median . | Range . |
Acarbose | Blood glucose–lowering drugs, excluding insulins | Drugs used in diabetes | 1 | 12.4 | (12.4–14.2) | 12.4 | (12.4–12.4) | 3 | 7.7 | (7.7–8.1) |
Cefoperazone and sulbactam | Other beta-lactam antibacterials | Antibacterials for systemic use | 1 | 183.7 | (183.7–211.2) | 183.7 | (183.7–183.7) | 1 | 11.4 | (11.4–11.4) |
Cefuroxime sodium (injection) | Other beta-lactam antibacterials | 2 | 103.2 | (103.2–103.2) | 103.2 | (103.2–103.2) | 7 | 4.8 | (4.6–19.3) | |
Aspirin | Antithrombotic agents | Drugs used for blood and blood–forming organs | 1 | 0.4 | (0.4–0.5) | 0.4 | (0.4–0.4) | 8 | 0.4 | (0.3–0.4) |
Peritoneal dialysis solution | Peritoneal dialytics | 1 | 3.1 | (3.1–3.1) | 3.1 | (3.1–3.1) | 3 | 2.5 | (2.5–2.9) | |
Human albumin (injection) | Blood substitutes and plasma protein fractions | 1 | 1132.5 | (1132.5–1132.5) | 1132.5 | (1132.5–1260.0) | 4 | 1080.0 | (1080.0–1260.0) | |
Mecobalamin (injection) | Vitamin B12 and folic acid | 1 | 6.0 | (6.0–7.2) | 6.0 | (5.7–6.0) | 3 | 1.3 | (1.1–6.0) | |
Clopidogrel | Antithrombotic agents | 1 | 16.5 | (16.5–19.0) | 15.5 | (15.5–16.5) | 2 | 6.6 | (6.6–8.6) | |
Atorvastatin calcium | Lipid–modifying agents, plain | Drugs used for cardiovascular system | 1 | 8.6 | (8.6–9.9) | 7.9 | (7.9–8.6) | 3 | 5.7 | (5.7–8.9) |
Amlodipine | Selective calcium channel blockers with mainly vascular effects (CCBs) | 1 | 4.7 | (4.7–5.4) | 4.7 | (4.3–4.7) | 10 | 0.2 | (0.2–2.2) | |
Nifedipine (controlled release) | Selective calcium channel blockers with mainly vascular effects (CCBs) | 1 | 4.3 | (4.3–5.1) | 3.8 | (3.8–4.3) | 5 | 2.8 | (2.3–2.8) | |
Omeprazole (injection) | Drugs for peptic ulcer and gastroesophageal reflux disease (PPIs) | Drugs for acid-related disorders | 1 | 43.5 | (43.5–61.0) | 41.0 | (41.0–44.0) | 9 | 0.7 | (0.7–70.2) |
Pantoprazole (injection) | Drugs for peptic ulcer and gastroesophageal reflux disease (PPIs) | 1 | 108.7 | (108.7–125.0) | 108.7 | (108.7–108.7) | 5 | 2.0 | (2.0–17.1) | |
Ambroxol hydrochloride (injection) | Expectorants, excluding combinations with cough suppressants | Drugs for respiratory system | 1 | 16.2 | (16.2–18.6) | 16.2 | (15.3–16.2) | 7 | 8.0 | (8.0–13.7) |
. | . | . | Original drugs . | Generic drugs . | ||||||
---|---|---|---|---|---|---|---|---|---|---|
. | . | . | . | Before RP . | After RP . | . | . | . | ||
International nonproprietary names . | Pharmacology classification . | Therapeutic subgroup . | Number of producers . | Median . | Range . | Median . | Range . | Number of producers . | Median . | Range . |
Acarbose | Blood glucose–lowering drugs, excluding insulins | Drugs used in diabetes | 1 | 12.4 | (12.4–14.2) | 12.4 | (12.4–12.4) | 3 | 7.7 | (7.7–8.1) |
Cefoperazone and sulbactam | Other beta-lactam antibacterials | Antibacterials for systemic use | 1 | 183.7 | (183.7–211.2) | 183.7 | (183.7–183.7) | 1 | 11.4 | (11.4–11.4) |
Cefuroxime sodium (injection) | Other beta-lactam antibacterials | 2 | 103.2 | (103.2–103.2) | 103.2 | (103.2–103.2) | 7 | 4.8 | (4.6–19.3) | |
Aspirin | Antithrombotic agents | Drugs used for blood and blood–forming organs | 1 | 0.4 | (0.4–0.5) | 0.4 | (0.4–0.4) | 8 | 0.4 | (0.3–0.4) |
Peritoneal dialysis solution | Peritoneal dialytics | 1 | 3.1 | (3.1–3.1) | 3.1 | (3.1–3.1) | 3 | 2.5 | (2.5–2.9) | |
Human albumin (injection) | Blood substitutes and plasma protein fractions | 1 | 1132.5 | (1132.5–1132.5) | 1132.5 | (1132.5–1260.0) | 4 | 1080.0 | (1080.0–1260.0) | |
Mecobalamin (injection) | Vitamin B12 and folic acid | 1 | 6.0 | (6.0–7.2) | 6.0 | (5.7–6.0) | 3 | 1.3 | (1.1–6.0) | |
Clopidogrel | Antithrombotic agents | 1 | 16.5 | (16.5–19.0) | 15.5 | (15.5–16.5) | 2 | 6.6 | (6.6–8.6) | |
Atorvastatin calcium | Lipid–modifying agents, plain | Drugs used for cardiovascular system | 1 | 8.6 | (8.6–9.9) | 7.9 | (7.9–8.6) | 3 | 5.7 | (5.7–8.9) |
Amlodipine | Selective calcium channel blockers with mainly vascular effects (CCBs) | 1 | 4.7 | (4.7–5.4) | 4.7 | (4.3–4.7) | 10 | 0.2 | (0.2–2.2) | |
Nifedipine (controlled release) | Selective calcium channel blockers with mainly vascular effects (CCBs) | 1 | 4.3 | (4.3–5.1) | 3.8 | (3.8–4.3) | 5 | 2.8 | (2.3–2.8) | |
Omeprazole (injection) | Drugs for peptic ulcer and gastroesophageal reflux disease (PPIs) | Drugs for acid-related disorders | 1 | 43.5 | (43.5–61.0) | 41.0 | (41.0–44.0) | 9 | 0.7 | (0.7–70.2) |
Pantoprazole (injection) | Drugs for peptic ulcer and gastroesophageal reflux disease (PPIs) | 1 | 108.7 | (108.7–125.0) | 108.7 | (108.7–108.7) | 5 | 2.0 | (2.0–17.1) | |
Ambroxol hydrochloride (injection) | Expectorants, excluding combinations with cough suppressants | Drugs for respiratory system | 1 | 16.2 | (16.2–18.6) | 16.2 | (15.3–16.2) | 7 | 8.0 | (8.0–13.7) |
Notes: Procurement price per DDD for each drug was reported in Chinese Yuan, CNY. In 2014: 1 United States dollar = 6.14 Chinese yuan. RP policy for 15 drug substances was piloted in Sanming, China in September 2014 to December 2016. Amoxicillin sodium and clavulanate potassium (injection) are excluded because of no procurement of its original drug in the study period. Dosage forms are specifically marked in parentheses for international nonproprietary names (INNs) that were not oral regular-release dosage forms. Pharmacology classification is based on ATC-3 codes and therapeutic subgroup is based on ATC-1 or ATC-2 codes.
Trends in drug volumes
Figure 2 visualizes the trends in total monthly procurement volumes, and those of the original and generic versions among the 14 drug substances, by Sanming and Longyan, and by the hospital inpatient and outpatient sectors. Again, the initiation of the ZMDP did not appear to be associated with monthly drug procurement volumes in Sanming. However, after September 2014 when RP was implemented in the inpatient sector of Sanming’s hospitals, the monthly procurement volumes of the 14 drug substances decreased by 3.1%, with a 40.8% decrease for the original drugs and a corresponding 118.4% increase for the generic drugs. Notably, the changes in trends of drug procurement volumes started about a quarter before official implementation of RP. However, in the outpatient sector of Sanming’s hospitals and in both the outpatient and inpatient sectors of Longyan’s hospitals, where RP policies were not implemented, the volumes of total procurement, and of original versions of the 14 drug substances increased by 29.8–49.8% after the initiation of RP. Simultaneously, volumes of the generic drugs increased by 10.0–31.2% in Longyan, and by 125.7% in Sanming’s outpatient sectors, with a much faster speed after the introduction of RP.
Common trends tests for DID
Appendix 5 reports the common trends tests. Appendix 5-1 shows chronologically the point estimates and 95% confidence intervals (CIs) of the interaction terms’ coefficients of the monthly dummy variables with the intervention dummy variable. It appears that trends in all outcomes were similar before the introduction of the RP and ZMDP. As shown in Appendix 5-2, the coefficients of the interaction terms were either small in absolute magnitudes (|$|{\beta _4}| \leqslant 0.01$| for most specifications) or statistically insignificant (P > 0.2 for most specifications, and P > 0.05 for all). These findings clearly suggest that the common trends assumption could not be rejected.
Impact of the ZMDP
As shown in Table 2, the DID estimation found insignificant associations of the ZMDP with drug procurement volumes, costs and hospital visits in Sanming’s tertiary hospitals (P > 0.6686 for volumes, P > 0.4240 for costs and P > 0.3123 for hospital visits, testing for the DID interaction coefficient). The DID estimation of the impacts of the ZMDP on each of the 14 drug substances’ unit price was also insignificant (Appendix 3).
. | ZMDP hospital (Sanming tertiary hospital) . | Control hospital (Longyan tertiary hospital) . | DID estimate . | |||||
---|---|---|---|---|---|---|---|---|
. | Before ZMDP . | After ZMDP . | Change (%) . | Before ZMDP . | After ZMDP . | Change (%) . | Change (%) . | 95% CI . |
Monthly volumes and costs | ||||||||
Inpatient sector | ||||||||
Total volumes (1000 DDDs) | 56.3 | 62.1 | 10.4 | 35.9 | 42.0 | 16.9 | −5.6 | [−27.3%, 22.7%] |
Original drugs | 47.7 | 55.4 | 16.2 | 31.0 | 38.1 | 23.1 | −5.6 | [−28.6%, 25.0%] |
Generic drugs | 8.6 | 6.7 | −21.9 | 4.9 | 3.8 | −22.0 | 0.1 | [−40.1%, 67.2%] |
Total costs (1000 CNY) | 1532.1 | 1245.2 | −18.7 | 1271.8 | 1205.1 | −5.2 | −14.2 | [−41.1%, 25.0%] |
Outpatient sector | ||||||||
Total volumes (1000 DDDs) | 150.3 | 190.8 | 26.9 | 117.0 | 149.8 | 28.0 | −0.8 | [−37.2%, 56.6%] |
Original drugs | 131.5 | 168.8 | 28.4 | 88.7 | 118.6 | 33.6 | −3.9 | [−43.1%, 62.3%] |
Generic drugs | 18.8 | 22.0 | 16.5 | 28.3 | 31.2 | 10.2 | 5.7 | [−45.9%, 106.5%] |
Total costs (1000 CNY) | 686.6 | 864.3 | 25.9 | 530.2 | 708.3 | 33.6 | −5.8 | [−32.8%, 32.0%] |
Monthly hospital visits | ||||||||
Inpatient admissions (1000) | 4.4 | 4.8 | 9.3 | 4.2 | 4.7 | 13.2 | −3.4 | [−11.8%, 5.7%] |
Outpatient visits (1000) | 66.2 | 75.0 | 13.3 | 100.7 | 119.3 | 18.4 | −4.4 | [−12.3%, 4.3%] |
. | ZMDP hospital (Sanming tertiary hospital) . | Control hospital (Longyan tertiary hospital) . | DID estimate . | |||||
---|---|---|---|---|---|---|---|---|
. | Before ZMDP . | After ZMDP . | Change (%) . | Before ZMDP . | After ZMDP . | Change (%) . | Change (%) . | 95% CI . |
Monthly volumes and costs | ||||||||
Inpatient sector | ||||||||
Total volumes (1000 DDDs) | 56.3 | 62.1 | 10.4 | 35.9 | 42.0 | 16.9 | −5.6 | [−27.3%, 22.7%] |
Original drugs | 47.7 | 55.4 | 16.2 | 31.0 | 38.1 | 23.1 | −5.6 | [−28.6%, 25.0%] |
Generic drugs | 8.6 | 6.7 | −21.9 | 4.9 | 3.8 | −22.0 | 0.1 | [−40.1%, 67.2%] |
Total costs (1000 CNY) | 1532.1 | 1245.2 | −18.7 | 1271.8 | 1205.1 | −5.2 | −14.2 | [−41.1%, 25.0%] |
Outpatient sector | ||||||||
Total volumes (1000 DDDs) | 150.3 | 190.8 | 26.9 | 117.0 | 149.8 | 28.0 | −0.8 | [−37.2%, 56.6%] |
Original drugs | 131.5 | 168.8 | 28.4 | 88.7 | 118.6 | 33.6 | −3.9 | [−43.1%, 62.3%] |
Generic drugs | 18.8 | 22.0 | 16.5 | 28.3 | 31.2 | 10.2 | 5.7 | [−45.9%, 106.5%] |
Total costs (1000 CNY) | 686.6 | 864.3 | 25.9 | 530.2 | 708.3 | 33.6 | −5.8 | [−32.8%, 32.0%] |
Monthly hospital visits | ||||||||
Inpatient admissions (1000) | 4.4 | 4.8 | 9.3 | 4.2 | 4.7 | 13.2 | −3.4 | [−11.8%, 5.7%] |
Outpatient visits (1000) | 66.2 | 75.0 | 13.3 | 100.7 | 119.3 | 18.4 | −4.4 | [−12.3%, 4.3%] |
Notes: Monetary values were reported in Chinese Yuan, CNY. In 2014, 1 United States dollar = 6.14 Chinese yuan. The ZMDP has been implemented in Sanming hospitals since January 2013, in Longyan’s secondary hospitals since January 2013 and in Longyan’s tertiary hospitals since June 2015.
. | ZMDP hospital (Sanming tertiary hospital) . | Control hospital (Longyan tertiary hospital) . | DID estimate . | |||||
---|---|---|---|---|---|---|---|---|
. | Before ZMDP . | After ZMDP . | Change (%) . | Before ZMDP . | After ZMDP . | Change (%) . | Change (%) . | 95% CI . |
Monthly volumes and costs | ||||||||
Inpatient sector | ||||||||
Total volumes (1000 DDDs) | 56.3 | 62.1 | 10.4 | 35.9 | 42.0 | 16.9 | −5.6 | [−27.3%, 22.7%] |
Original drugs | 47.7 | 55.4 | 16.2 | 31.0 | 38.1 | 23.1 | −5.6 | [−28.6%, 25.0%] |
Generic drugs | 8.6 | 6.7 | −21.9 | 4.9 | 3.8 | −22.0 | 0.1 | [−40.1%, 67.2%] |
Total costs (1000 CNY) | 1532.1 | 1245.2 | −18.7 | 1271.8 | 1205.1 | −5.2 | −14.2 | [−41.1%, 25.0%] |
Outpatient sector | ||||||||
Total volumes (1000 DDDs) | 150.3 | 190.8 | 26.9 | 117.0 | 149.8 | 28.0 | −0.8 | [−37.2%, 56.6%] |
Original drugs | 131.5 | 168.8 | 28.4 | 88.7 | 118.6 | 33.6 | −3.9 | [−43.1%, 62.3%] |
Generic drugs | 18.8 | 22.0 | 16.5 | 28.3 | 31.2 | 10.2 | 5.7 | [−45.9%, 106.5%] |
Total costs (1000 CNY) | 686.6 | 864.3 | 25.9 | 530.2 | 708.3 | 33.6 | −5.8 | [−32.8%, 32.0%] |
Monthly hospital visits | ||||||||
Inpatient admissions (1000) | 4.4 | 4.8 | 9.3 | 4.2 | 4.7 | 13.2 | −3.4 | [−11.8%, 5.7%] |
Outpatient visits (1000) | 66.2 | 75.0 | 13.3 | 100.7 | 119.3 | 18.4 | −4.4 | [−12.3%, 4.3%] |
. | ZMDP hospital (Sanming tertiary hospital) . | Control hospital (Longyan tertiary hospital) . | DID estimate . | |||||
---|---|---|---|---|---|---|---|---|
. | Before ZMDP . | After ZMDP . | Change (%) . | Before ZMDP . | After ZMDP . | Change (%) . | Change (%) . | 95% CI . |
Monthly volumes and costs | ||||||||
Inpatient sector | ||||||||
Total volumes (1000 DDDs) | 56.3 | 62.1 | 10.4 | 35.9 | 42.0 | 16.9 | −5.6 | [−27.3%, 22.7%] |
Original drugs | 47.7 | 55.4 | 16.2 | 31.0 | 38.1 | 23.1 | −5.6 | [−28.6%, 25.0%] |
Generic drugs | 8.6 | 6.7 | −21.9 | 4.9 | 3.8 | −22.0 | 0.1 | [−40.1%, 67.2%] |
Total costs (1000 CNY) | 1532.1 | 1245.2 | −18.7 | 1271.8 | 1205.1 | −5.2 | −14.2 | [−41.1%, 25.0%] |
Outpatient sector | ||||||||
Total volumes (1000 DDDs) | 150.3 | 190.8 | 26.9 | 117.0 | 149.8 | 28.0 | −0.8 | [−37.2%, 56.6%] |
Original drugs | 131.5 | 168.8 | 28.4 | 88.7 | 118.6 | 33.6 | −3.9 | [−43.1%, 62.3%] |
Generic drugs | 18.8 | 22.0 | 16.5 | 28.3 | 31.2 | 10.2 | 5.7 | [−45.9%, 106.5%] |
Total costs (1000 CNY) | 686.6 | 864.3 | 25.9 | 530.2 | 708.3 | 33.6 | −5.8 | [−32.8%, 32.0%] |
Monthly hospital visits | ||||||||
Inpatient admissions (1000) | 4.4 | 4.8 | 9.3 | 4.2 | 4.7 | 13.2 | −3.4 | [−11.8%, 5.7%] |
Outpatient visits (1000) | 66.2 | 75.0 | 13.3 | 100.7 | 119.3 | 18.4 | −4.4 | [−12.3%, 4.3%] |
Notes: Monetary values were reported in Chinese Yuan, CNY. In 2014, 1 United States dollar = 6.14 Chinese yuan. The ZMDP has been implemented in Sanming hospitals since January 2013, in Longyan’s secondary hospitals since January 2013 and in Longyan’s tertiary hospitals since June 2015.
Impact of the RP
Prices
Similar to the ZMDP, the DID estimation of the impact of RP on drug unit prices was insignificant for all of the 14 drug substances (Appendix 3).
Volumes
The DID estimations of the impact of RP on drug procurement volumes are shown in Table 3. In the inpatient sector of Sanming, the initiation of RP was associated with a decrease in the total monthly procurement volumes for the 14 drug substances by 25.9% (95% CI, 12.9–37.0%), with a decrease in their original versions by 56.8% (95% CI, 47.0–64.7%), and an increase in their generic versions by 98.6% (95% CI, 60.5–145.8%). In the outpatient sector in Sanming, where the RP policy had not been implemented, the RP was not associated with total procurement volumes for the 14 drug substances, nor the procurement volumes for their original versions (P = 0.7217 and 0.4505, respectively). Interestingly, however, the introduction of RP was associated with an increase in the procurement volumes of the generic versions by 72.0% (95% CI, 28.6–130.0%).
. | RP region (Sanming) . | Control region (Longyan) . | DID estimates . | |||||
---|---|---|---|---|---|---|---|---|
. | Before RP . | After RP . | Change (%) . | Before RP . | After RP . | Change (%) . | Change (%) . | 95% CI . |
Monthly volumes and costs | ||||||||
Inpatient sector | ||||||||
Total volumes (1000 DDDs) | 38.9 | 37.7 | −3.1 | 26.4 | 34.6 | 30.9 | −25.9 | [−37.0%, −12.9%] |
Original drugs | 29.7 | 17.6 | −40.8 | 20.5 | 28.0 | 37.0 | −56.8 | [−64.7%, −47.0%] |
Generic drugs | 9.2 | 20.1 | 118.4 | 6.0 | 6.6 | 10.0 | 98.6 | [60.5%, 145.8%] |
Total costs (1000 CNY) | 777.6 | 482.0 | −38.0 | 676.1 | 800.9 | 18.4 | −47.7 | [−58.7%, −33.7%] |
Outpatient sector | ||||||||
Total volumes (1000 DDDs) | 102.5 | 153.5 | 49.8 | 91.3 | 130.4 | 42.8 | 4.9 | [−19.4%, 36.6%] |
Original drugs | 81.1 | 105.3 | 29.8 | 64.3 | 94.9 | 47.6 | −12.1 | [−37.1%, 22.9%] |
Generic drugs | 21.4 | 48.3 | 125.7 | 27.0 | 35.5 | 31.2 | 72.0 | [28.6%, 130.0%] |
Total costs (1000 CNY) | 456.7 | 627.4 | 37.4 | 384.4 | 625.2 | 62.6% | −15.5 | [−32.0%, 5.0%] |
Monthly hospital visits | ||||||||
Inpatient admissions (1000) | 2.9 | 3.5 | 21.2 | 3.2 | 3.6 | 11.0 | 9.4 | [−16.7%, 43.7%] |
Outpatient visits (1000) | 45.6 | 58.3 | 27.6 | 75.3 | 81.1 | 7.6 | 18.8 | [−14.2%, 64.4%] |
. | RP region (Sanming) . | Control region (Longyan) . | DID estimates . | |||||
---|---|---|---|---|---|---|---|---|
. | Before RP . | After RP . | Change (%) . | Before RP . | After RP . | Change (%) . | Change (%) . | 95% CI . |
Monthly volumes and costs | ||||||||
Inpatient sector | ||||||||
Total volumes (1000 DDDs) | 38.9 | 37.7 | −3.1 | 26.4 | 34.6 | 30.9 | −25.9 | [−37.0%, −12.9%] |
Original drugs | 29.7 | 17.6 | −40.8 | 20.5 | 28.0 | 37.0 | −56.8 | [−64.7%, −47.0%] |
Generic drugs | 9.2 | 20.1 | 118.4 | 6.0 | 6.6 | 10.0 | 98.6 | [60.5%, 145.8%] |
Total costs (1000 CNY) | 777.6 | 482.0 | −38.0 | 676.1 | 800.9 | 18.4 | −47.7 | [−58.7%, −33.7%] |
Outpatient sector | ||||||||
Total volumes (1000 DDDs) | 102.5 | 153.5 | 49.8 | 91.3 | 130.4 | 42.8 | 4.9 | [−19.4%, 36.6%] |
Original drugs | 81.1 | 105.3 | 29.8 | 64.3 | 94.9 | 47.6 | −12.1 | [−37.1%, 22.9%] |
Generic drugs | 21.4 | 48.3 | 125.7 | 27.0 | 35.5 | 31.2 | 72.0 | [28.6%, 130.0%] |
Total costs (1000 CNY) | 456.7 | 627.4 | 37.4 | 384.4 | 625.2 | 62.6% | −15.5 | [−32.0%, 5.0%] |
Monthly hospital visits | ||||||||
Inpatient admissions (1000) | 2.9 | 3.5 | 21.2 | 3.2 | 3.6 | 11.0 | 9.4 | [−16.7%, 43.7%] |
Outpatient visits (1000) | 45.6 | 58.3 | 27.6 | 75.3 | 81.1 | 7.6 | 18.8 | [−14.2%, 64.4%] |
Notes: Monetary values were reported in Chinese Yuan, CNY. In 2014, 1 United States dollar = 6.14 Chinese yuan. RP policy for 15 drug substances was piloted in Sanming, China in September 2014 to December 2016. Amoxicillin sodium and clavulanate potassium (injection) were excluded from the table because of no procurement of its original drug in the study period.
. | RP region (Sanming) . | Control region (Longyan) . | DID estimates . | |||||
---|---|---|---|---|---|---|---|---|
. | Before RP . | After RP . | Change (%) . | Before RP . | After RP . | Change (%) . | Change (%) . | 95% CI . |
Monthly volumes and costs | ||||||||
Inpatient sector | ||||||||
Total volumes (1000 DDDs) | 38.9 | 37.7 | −3.1 | 26.4 | 34.6 | 30.9 | −25.9 | [−37.0%, −12.9%] |
Original drugs | 29.7 | 17.6 | −40.8 | 20.5 | 28.0 | 37.0 | −56.8 | [−64.7%, −47.0%] |
Generic drugs | 9.2 | 20.1 | 118.4 | 6.0 | 6.6 | 10.0 | 98.6 | [60.5%, 145.8%] |
Total costs (1000 CNY) | 777.6 | 482.0 | −38.0 | 676.1 | 800.9 | 18.4 | −47.7 | [−58.7%, −33.7%] |
Outpatient sector | ||||||||
Total volumes (1000 DDDs) | 102.5 | 153.5 | 49.8 | 91.3 | 130.4 | 42.8 | 4.9 | [−19.4%, 36.6%] |
Original drugs | 81.1 | 105.3 | 29.8 | 64.3 | 94.9 | 47.6 | −12.1 | [−37.1%, 22.9%] |
Generic drugs | 21.4 | 48.3 | 125.7 | 27.0 | 35.5 | 31.2 | 72.0 | [28.6%, 130.0%] |
Total costs (1000 CNY) | 456.7 | 627.4 | 37.4 | 384.4 | 625.2 | 62.6% | −15.5 | [−32.0%, 5.0%] |
Monthly hospital visits | ||||||||
Inpatient admissions (1000) | 2.9 | 3.5 | 21.2 | 3.2 | 3.6 | 11.0 | 9.4 | [−16.7%, 43.7%] |
Outpatient visits (1000) | 45.6 | 58.3 | 27.6 | 75.3 | 81.1 | 7.6 | 18.8 | [−14.2%, 64.4%] |
. | RP region (Sanming) . | Control region (Longyan) . | DID estimates . | |||||
---|---|---|---|---|---|---|---|---|
. | Before RP . | After RP . | Change (%) . | Before RP . | After RP . | Change (%) . | Change (%) . | 95% CI . |
Monthly volumes and costs | ||||||||
Inpatient sector | ||||||||
Total volumes (1000 DDDs) | 38.9 | 37.7 | −3.1 | 26.4 | 34.6 | 30.9 | −25.9 | [−37.0%, −12.9%] |
Original drugs | 29.7 | 17.6 | −40.8 | 20.5 | 28.0 | 37.0 | −56.8 | [−64.7%, −47.0%] |
Generic drugs | 9.2 | 20.1 | 118.4 | 6.0 | 6.6 | 10.0 | 98.6 | [60.5%, 145.8%] |
Total costs (1000 CNY) | 777.6 | 482.0 | −38.0 | 676.1 | 800.9 | 18.4 | −47.7 | [−58.7%, −33.7%] |
Outpatient sector | ||||||||
Total volumes (1000 DDDs) | 102.5 | 153.5 | 49.8 | 91.3 | 130.4 | 42.8 | 4.9 | [−19.4%, 36.6%] |
Original drugs | 81.1 | 105.3 | 29.8 | 64.3 | 94.9 | 47.6 | −12.1 | [−37.1%, 22.9%] |
Generic drugs | 21.4 | 48.3 | 125.7 | 27.0 | 35.5 | 31.2 | 72.0 | [28.6%, 130.0%] |
Total costs (1000 CNY) | 456.7 | 627.4 | 37.4 | 384.4 | 625.2 | 62.6% | −15.5 | [−32.0%, 5.0%] |
Monthly hospital visits | ||||||||
Inpatient admissions (1000) | 2.9 | 3.5 | 21.2 | 3.2 | 3.6 | 11.0 | 9.4 | [−16.7%, 43.7%] |
Outpatient visits (1000) | 45.6 | 58.3 | 27.6 | 75.3 | 81.1 | 7.6 | 18.8 | [−14.2%, 64.4%] |
Notes: Monetary values were reported in Chinese Yuan, CNY. In 2014, 1 United States dollar = 6.14 Chinese yuan. RP policy for 15 drug substances was piloted in Sanming, China in September 2014 to December 2016. Amoxicillin sodium and clavulanate potassium (injection) were excluded from the table because of no procurement of its original drug in the study period.
Costs
The total monthly procurement costs for these 14 drug substances decreased by 295 600 CNY on average in Sanming’s inpatient sector, after the implementation of the RP policy. In relative terms, the RP policy was associated with a decrease in the total costs by 47.7% (95% CI, 33.7–58.7%) in Sanming’s inpatient sector. While in the outpatient sector, the RP policy was not associated with any changes in costs (P = 0.1282) (Table 3).
Hospital visits
As shown in Table 3, monthly hospital visits, including both inpatient admissions and outpatient visits, were not associated with the initiation of the RP policy (P = 0.5164 and 0.2993, respectively).
Subgroup analyses
Table 4 shows the subgroup analyses on the impact of RP, stratified by hospital level. Each reported similar effects to those from the aggregated analyses. Subgroup analyses by the drugs’ therapeutic classes are shown in Figure 3 and Appendix 4. For each therapeutic class, the RP policy was associated with a decrease in the procurement volumes of the original versions, and an increase in their generic counterparts. The effect size of RP policy on drug volumes was more profound among drugs for acid-related disorders (−59.2%), and antibacterials for systemic use (−55.7%).
. | RP region (Sanming) . | Control region (Longyan) . | DID estimates . | |||||
---|---|---|---|---|---|---|---|---|
. | Before RP . | After RP . | Change (%) . | Before RP . | After RP . | Change (%) . | Change (%) . | 95% CI . |
Monthly volumes and costs | ||||||||
Tertiary hospitals | ||||||||
Total volumes (1000 DDDs) | 60.0 | 50.9 | −15.2 | 39.7 | 54.3 | 36.9 | −38.0 | [−48.2%, −25.8%] |
Original drugs | 52.5 | 29.9 | −43.2 | 35.5 | 48.8 | 37.7 | −58.7 | [−66.0%, −49.8%] |
Generic drugs | 7.4 | 21.0 | 182.5 | 4.3 | 5.5 | 30.1 | 117.2 | [57.4%, 199.6%] |
Total costs (1000 CNY) | 1352.8 | 824.9 | −39.0 | 1230.1 | 1383.0 | 12.4 | −45.8 | [−57.4%, −31.0%] |
Secondary hospitals | ||||||||
Total volumes (1000 DDDs) | 17.9 | 24.6 | 37.6 | 13.2 | 14.9 | 12.8 | 22.0 | [−3.1%, 53.6%] |
Original drugs | 6.9 | 5.3 | −22.5 | 5.5 | 7.2 | 32.4 | −41.4 | [−60.5%, −13.1%] |
Generic drugs | 11.0 | 19.2 | 75.0 | 7.7 | 7.6 | −1.1 | 77.0 | [33.4%, 134.9%] |
Total costs (1000 CNY) | 202.4 | 139.0 | −31.3 | 122.2 | 218.8 | 79.1 | −61.6 | [−71.0%, −49.3%] |
Monthly inpatient admission | ||||||||
Tertiary hospitals (1000) | 4.7 | 5.0 | 5.2 | 4.7 | 5.1 | 9.8 | −3.8 | [−11.4%, 4.4%] |
Secondary hospitals (1000) | 1.7 | 2.0 | 16.7 | 1.7 | 2.2 | 26.3 | −7.8 | [−15.2%, 0.1%] |
. | RP region (Sanming) . | Control region (Longyan) . | DID estimates . | |||||
---|---|---|---|---|---|---|---|---|
. | Before RP . | After RP . | Change (%) . | Before RP . | After RP . | Change (%) . | Change (%) . | 95% CI . |
Monthly volumes and costs | ||||||||
Tertiary hospitals | ||||||||
Total volumes (1000 DDDs) | 60.0 | 50.9 | −15.2 | 39.7 | 54.3 | 36.9 | −38.0 | [−48.2%, −25.8%] |
Original drugs | 52.5 | 29.9 | −43.2 | 35.5 | 48.8 | 37.7 | −58.7 | [−66.0%, −49.8%] |
Generic drugs | 7.4 | 21.0 | 182.5 | 4.3 | 5.5 | 30.1 | 117.2 | [57.4%, 199.6%] |
Total costs (1000 CNY) | 1352.8 | 824.9 | −39.0 | 1230.1 | 1383.0 | 12.4 | −45.8 | [−57.4%, −31.0%] |
Secondary hospitals | ||||||||
Total volumes (1000 DDDs) | 17.9 | 24.6 | 37.6 | 13.2 | 14.9 | 12.8 | 22.0 | [−3.1%, 53.6%] |
Original drugs | 6.9 | 5.3 | −22.5 | 5.5 | 7.2 | 32.4 | −41.4 | [−60.5%, −13.1%] |
Generic drugs | 11.0 | 19.2 | 75.0 | 7.7 | 7.6 | −1.1 | 77.0 | [33.4%, 134.9%] |
Total costs (1000 CNY) | 202.4 | 139.0 | −31.3 | 122.2 | 218.8 | 79.1 | −61.6 | [−71.0%, −49.3%] |
Monthly inpatient admission | ||||||||
Tertiary hospitals (1000) | 4.7 | 5.0 | 5.2 | 4.7 | 5.1 | 9.8 | −3.8 | [−11.4%, 4.4%] |
Secondary hospitals (1000) | 1.7 | 2.0 | 16.7 | 1.7 | 2.2 | 26.3 | −7.8 | [−15.2%, 0.1%] |
Notes: Monetary values were reported in Chinese Yuan, CNY. In 2014, 1 United States dollar = 6.14 Chinese yuan. RP policy for 15 drug substances was piloted in Sanming, China in September 2014 to December 2016. Amoxicillin sodium and clavulanate potassium (injection) were excluded from the table because of no procurement of its original drug in the study period.
. | RP region (Sanming) . | Control region (Longyan) . | DID estimates . | |||||
---|---|---|---|---|---|---|---|---|
. | Before RP . | After RP . | Change (%) . | Before RP . | After RP . | Change (%) . | Change (%) . | 95% CI . |
Monthly volumes and costs | ||||||||
Tertiary hospitals | ||||||||
Total volumes (1000 DDDs) | 60.0 | 50.9 | −15.2 | 39.7 | 54.3 | 36.9 | −38.0 | [−48.2%, −25.8%] |
Original drugs | 52.5 | 29.9 | −43.2 | 35.5 | 48.8 | 37.7 | −58.7 | [−66.0%, −49.8%] |
Generic drugs | 7.4 | 21.0 | 182.5 | 4.3 | 5.5 | 30.1 | 117.2 | [57.4%, 199.6%] |
Total costs (1000 CNY) | 1352.8 | 824.9 | −39.0 | 1230.1 | 1383.0 | 12.4 | −45.8 | [−57.4%, −31.0%] |
Secondary hospitals | ||||||||
Total volumes (1000 DDDs) | 17.9 | 24.6 | 37.6 | 13.2 | 14.9 | 12.8 | 22.0 | [−3.1%, 53.6%] |
Original drugs | 6.9 | 5.3 | −22.5 | 5.5 | 7.2 | 32.4 | −41.4 | [−60.5%, −13.1%] |
Generic drugs | 11.0 | 19.2 | 75.0 | 7.7 | 7.6 | −1.1 | 77.0 | [33.4%, 134.9%] |
Total costs (1000 CNY) | 202.4 | 139.0 | −31.3 | 122.2 | 218.8 | 79.1 | −61.6 | [−71.0%, −49.3%] |
Monthly inpatient admission | ||||||||
Tertiary hospitals (1000) | 4.7 | 5.0 | 5.2 | 4.7 | 5.1 | 9.8 | −3.8 | [−11.4%, 4.4%] |
Secondary hospitals (1000) | 1.7 | 2.0 | 16.7 | 1.7 | 2.2 | 26.3 | −7.8 | [−15.2%, 0.1%] |
. | RP region (Sanming) . | Control region (Longyan) . | DID estimates . | |||||
---|---|---|---|---|---|---|---|---|
. | Before RP . | After RP . | Change (%) . | Before RP . | After RP . | Change (%) . | Change (%) . | 95% CI . |
Monthly volumes and costs | ||||||||
Tertiary hospitals | ||||||||
Total volumes (1000 DDDs) | 60.0 | 50.9 | −15.2 | 39.7 | 54.3 | 36.9 | −38.0 | [−48.2%, −25.8%] |
Original drugs | 52.5 | 29.9 | −43.2 | 35.5 | 48.8 | 37.7 | −58.7 | [−66.0%, −49.8%] |
Generic drugs | 7.4 | 21.0 | 182.5 | 4.3 | 5.5 | 30.1 | 117.2 | [57.4%, 199.6%] |
Total costs (1000 CNY) | 1352.8 | 824.9 | −39.0 | 1230.1 | 1383.0 | 12.4 | −45.8 | [−57.4%, −31.0%] |
Secondary hospitals | ||||||||
Total volumes (1000 DDDs) | 17.9 | 24.6 | 37.6 | 13.2 | 14.9 | 12.8 | 22.0 | [−3.1%, 53.6%] |
Original drugs | 6.9 | 5.3 | −22.5 | 5.5 | 7.2 | 32.4 | −41.4 | [−60.5%, −13.1%] |
Generic drugs | 11.0 | 19.2 | 75.0 | 7.7 | 7.6 | −1.1 | 77.0 | [33.4%, 134.9%] |
Total costs (1000 CNY) | 202.4 | 139.0 | −31.3 | 122.2 | 218.8 | 79.1 | −61.6 | [−71.0%, −49.3%] |
Monthly inpatient admission | ||||||||
Tertiary hospitals (1000) | 4.7 | 5.0 | 5.2 | 4.7 | 5.1 | 9.8 | −3.8 | [−11.4%, 4.4%] |
Secondary hospitals (1000) | 1.7 | 2.0 | 16.7 | 1.7 | 2.2 | 26.3 | −7.8 | [−15.2%, 0.1%] |
Notes: Monetary values were reported in Chinese Yuan, CNY. In 2014, 1 United States dollar = 6.14 Chinese yuan. RP policy for 15 drug substances was piloted in Sanming, China in September 2014 to December 2016. Amoxicillin sodium and clavulanate potassium (injection) were excluded from the table because of no procurement of its original drug in the study period.
Sensitivity analyses
Appendix 6 reported the estimation from the time–varying DID specification, reporting consist findings that the ZMDP was not associated with drug volumes and costs but RP was. Appendix 7 compares the estimated impacts of RP between the whole database and the excluded data before January 2013. Appendix 8 compares the estimated impacts of RP, using data from January 2012 to January 2013 as the pre-policy period and from June 2015 to December 2016 as the post-policy period, stratified by hospital level. In both cases, the Hausman test could not reject the null hypothesis that the two groups of estimations are indifferent (P > 0.19 for all outcome variables in Appendix 7, and P > 0.21 for most in Appendix 8).
The placebo test that artificially changed the initiation month of RP yielded consistent estimations ( Appendix 9), where for each outcome, the largest policy effect was around the month of September 2014, when the RP was initiated.
Discussion
Using drug procurement data from a large-scale natural experiment, we found that the pilot of RP in Sanming, China, targeting the 15 most-used drug substances, was not associated with changes in drug prices, but substantially decreased drug volumes by 25.9% and decreased total costs by 47.7%. The impact of Sanming’s RP pilot could be largely attributed to its effects in shifting the volumes of high-price original versions (decreased by 56.8%) to generic versions. The ZMDP, however, seemed be associated with none of the outcomes that were ascertained.
Acosta et al.’s (2014) systematic review summarized two modalities of RP, i.e. external and internal RP, which are widely carried out internationally. In this study, we were actually exploring the type of internal RP, which has been used in countries in Europe, Canada and New Zealand (Aaserud et al., 2006). Notably, we have provided the first report on the impact of RP on drug prices, volumes and costs in China. Although RP had been implemented in Japan and South Korea, we only identified one empirical study without controls, which investigated the impact of a comprehensive drug pricing reform in South Korea, with RP as one of its components. The study found an associated 9.3% decrease in drug costs, much lower than ours (Cho et al., 2015). In other low- and middle-income settings, we only identified one study from South Africa that examined the impact of RP (de Jager and Suleman, 2019). However, using before-versus-after designs without a parallel control group, the study yielded discrepant results for the impact of RP across two different drug substances.
In theory, RP takes effect by disincentivizing physicians’ preference for or patients’ acceptance of high-priced drugs. Since RP can direct patient demand to low-priced counterparts of drug substances with low profit margins, the rebates for physicians decrease, which in turn eliminates the financial incentives of prescribing high–priced drugs. Our findings corroborate the literatures from high-income countries showing that RP is an effective means to contain high-priced drug volumes and drug costs (Aaserud et al., 2006; Galizzi et al., 2011). In these systems, RP is used to target so–called original drugs or cost-sharing drugs, where health schemes normally group interchangeable drug substances and set standard rates of reimbursement, usually the lowest price, within the same group (Acosta et al., 2014). With this financial incentive, RP was believed to be able to steer demand to low-priced drugs and push drug manufacturers to reduce unit prices (Aaserud et al., 2006; Galizzi et al., 2011). Empirically, with the exception of the USA (Kibicho and Pinkerton, 2012; Robinson et al., 2017), studies from high-income countries have consistently revealed that RP is associated with decreases in targeted drugs’ unit prices, consumption volumes and costs, but not with patients’ health outcomes (Aaserud et al., 2006; Galizzi et al., 2011; Acosta et al., 2014).
We found that the introduction of RP in Sanming, China, was associated with substantial decreases in drug volumes and costs, which corroborated the findings from most high-income countries. However, such a policy seemed not to be associated with changes in drug prices, in contrast to research findings from Europe and Canada, where the implementation of RP was associated with a decrease in the targeted drugs’ prices by 18–33% (Pavenik, 2002; Brekke et al., 2009, 2011). Within China’s Essential Medicines Program, drug procurements were carried out at the provincial level and the unit price of each individual drug product must be the same within a province (Hu and Mossialos, 2016). Given such a context, it is not a surprise that the DID specification that compared Sanming and Longyan could not identify any significant impact of RP on drug prices. However, even within Sanming, our analyses identified no trends that were associated with the introduction of RP either, suggesting that RP may not have any effects on drug prices. Indeed, the policy design of RP in Sanming was similar to that in the high-income settings, which do introduce incentives for manufactures to decrease drugs’ unit prices. However, Sanming is just one of nine cities in Fujian Province and thus has very limited influence on the entire market, in particular for high-priced original brands. Thus, a larger scale pilot is warranted to obtain more robust evidence.
Adopting the same evaluation strategy, and data with the exclusion of the period after RP implementation, we found that the ZMDP in Sanming played no role in changing the prices, volumes and costs for the same 14 most-used drug substances. Previous studies that investigated the ZMDP in China offer inconsistent findings regarding its effects in controlling drug volumes and costs (Li et al., 2013, 2018; Yang et al., 2013; Chen et al., 2014; Ding and Wu, 2017; Wei et al., 2017; He et al., 2018; Tang et al., 2018; Zeng et al., 2019). For those indicating that the policy was successful, outcome indicators were all broadly defined as per visit/prescription volumes and/or costs, due to the lack of retail data from hospital pharmacies (Li et al., 2013, 2018; Yang et al., 2013; Chen et al., 2014; Wei et al., 2017; He et al., 2018; Zeng et al., 2019). However, for those reporting negative findings of the policy’s impact, drug volumes and costs were all measured monthly or annually (Ding and Wu, 2017; Tang et al., 2018). One reason for these inconsistencies might be attributed to physicians’ incentives to increase the number of patients’ total visits and encourage re-admissions. Indeed, China’s health system brings health providers’ distorted incentives to generate revenues. Previous studies on the impact of the ZMDP in China revealed substantial spill-over effects as well. For example, both Yi et al. (2015) and Ding and Wu (2017) found that healthcare providers may have strong incentives to increase inpatient volumes against the reduction in drug revenues under the ZMDP. Our findings provide new evidence of spill-overs. As we found, the policy in Sanming had a substantial spill-over effect on the hospitals’ outpatient sector, where no RP was carried out, which was associated with a 125.7% increase in the volumes of the generic drugs. As physicians are not permanently fixed in a specific medical sector, they may smoothly take their prescribing habits from one sector (i.e. inpatient sector) to another (i.e. outpatient sector). Additionally, although the ZMDP may have eliminated physicians’ direct financial incentives from prescribing drugs, the implementation of the ZMDP was still subsidized by governments against the losses in drug sales (Hu and Mossialos, 2016). In particular, all previous studies reporting positive effects of the ZMDP consistently used claim/discharge data that relies on information from drug retail costs. However, one important component of the ZMDP’s measures was to remove the 15% markup of drug retail prices from procurement prices, which would obviously lead to reductions in drug costs as reported in claim/discharge data, even if it had no impact on reversing physicians’ incentives to overprescribe. We only identified one other study from China that used drug procurement data to evaluate the impact of the ZMDP, which also found that such a policy had no effect on reducing drug volumes and costs (Tang et al., 2018). Importantly, the ZMDP’s policy intention lies in eliminating physicians’ financial incentives to overprescribe high-priced drugs (Li et al., 2012; Yip et al., 2012). However, upon evaluating the evidence, we found no indication that the ZMDP had achieved its goal.
Sanming implemented systemic healthcare reforms in January 2013 by making changes to the hospitals’ governance structure, payment system and compensation method, with the ZMDP one of its components. Current evidence from Sanming primarily pertains to its general reforms of the health systems, without specifying the mechanisms that drove down the drug costs. For example, using the hospitals’ claim data from 2008 to 2014, Fu et al. (2017) found that this round of reform was associated with a 15.4% reduction of per inpatient admission drug costs. Upon focusing on the ZMDP and the 14 specific drug substances that were most used in Sanming, we found no effect of the ZMDP on drug procurement volumes or costs. RP might be one reason for Sanming’s success in containing drug costs, but further studies with more deliberate designs are warranted.
Our study should be interpreted considering its limitations. First, we used a non-randomized processing method in building counterfactuals. Luckily, however, the common trends tests support the model specifications of the DID. While Longyan is the best choice as a control considering its location, economic development and population, and the several sensitivity and subgroup analyses yielded consistent findings, we are still confronted with risks in drawing causal conclusions regarding the policies’ impacts. In particular, we might have overestimated the impact of RP in Sanming, because as a benchmark city of China’s health reform, systemic and comprehensive reforms were prompted in Sanming, which were believed to have reduced drug costs in general (Yip et al., 2019). Second, our study only used data from two hospitals in Sanming, which is a very small sample size. These hospitals were geographically representative on this occasion, however, because they were the only tertiary and secondary hospitals in the specific city or county that was investigated. Third, RP in Sanming was only implemented on 15 drug substances in clinical use, which might increase the usage of interchangeable drug substances not subjected to RP and increase their costs. Thus, we might have overestimated the policy impact. However, the definition and extent of interchangeable drug substances remain unclear, and because these 15 drug substances were based on the largest consumption volumes, the probability of patients shifting to other drug substances seemed to be low. Fourth, we used the hospitals’ procurement data, which might be affected by hospital visits and thus may not completely reflect drug utilization. Our analysis of the hospital visits data from the same hospitals found that no significant changes in either the inpatient or outpatient sectors were associated with the implementation of RP and the ZMDP. Acknowledging RP, hospitals may tend to procure more generic drugs and fewer original versions in advance, to prepare for the policy changes. However, our placebo test does not support such a hypothesis.
The Chinese government is currently consolidating its health insurance schemes and piloting various efforts to control drug prices and costs. In Sanming, although we found that the implementation of RP was strongly associated with decreases in drugs’ procurement volumes and costs, such a policy had no effect on drug prices. RP might have more potential to cause a change should the Chinese government decide to experiment on a larger scale or even make it national. In fact, despite the 15 drug substances, most off-patent drugs had generic counterparts in China’s domestic drug market. Although the bioequivalence evaluation of listed drugs is still in process in China (Chen et al., 2016), the Chinese government has never limited the market entrance of generic drugs, and there is no clear evidence indicating different performance in real-world clinical outcomes between original and generic drugs in China. Therefore, the implementation of RP at least would not be harmful as it could save drug costs with no or little quality risks.
Conclusion
The Chinese government is currently consolidating its health systems and searching for solutions to make healthcare more affordable. The RP pilot in Sanming was associated with a substantial decrease in drug volumes and costs. With consideration of the development and regulation of the generic market, RP could be scaled up to better control drug costs in China.
Author contribution
B.J. designed the study and collected the data. X.L.F. and B.J. conceived the paper. R.J.Z. reviewed literatures and analyzed the data with X.L.F.’s guidance. X.L.F. made the first draft of the paper and all other authors contributed to the critical interpretation of the findings.
Funding
This work was supported by the China National Natural Science Foundation (71761130083, 71422009).
Conflict of interest statement. None declared.
Ethical approval. No ethical approval was required for this study.
References
Appendix 1
Trends in prices of each drug substance that was subjected to RP from January 2012 to December 2016, measured as median monthly procurement price per DDD.
. | . | . | Original drug . | |||
---|---|---|---|---|---|---|
. | . | . | Before ZMDP . | After ZMDP . | ||
INN . | Pharmacology classification . | Therapeutic subgroup . | Median price . | Price range . | Median price . | Price range . |
Acarbose | Blood glucose–lowering drugs, excluding insulins | Drugs used in diabetes | 12.4 | (9.2–14.7) | 12.4 | (12.4–12.4) |
Cefoperazone and sulbactam | Other beta-lactam antibacterials | Antibacterials for systemic use | 211.2 | (211.2–211.2) | 183.7 | (183.7–211.2) |
Cefuroxime sodium (injection) | Other beta-lactam antibacterials | 103.2 | (103.2–103.2) | 103.2 | (103.2–103.2) | |
Aspirin | Antithrombotic agents | Drugs used for blood and blood–forming organs | 0.4 | (0.4–0.5) | 0.4 | (0.4–0.5) |
Peritoneal dialysis solution | Peritoneal dialytics | 3.1 | (3.1–3.1) | 3.1 | (3.1–3.1) | |
Human albumin (injection) | Blood substitutes and plasma protein fractions | 1132.5 | (1132.5–1132.5) | 1132.5 | (1132.5–1132.5) | |
Mecobalamin (injection) | Vitamin B12 and folic acid | 6.9 | (6.9–7.2) | 6.0 | (6.0–6.0) | |
Clopidogrel | Antithrombotic agents | 20.5 | (19.0–20.5) | 16.5 | (16.5–19.0) | |
Atorvastatin calcium | Lipid modifying agents, plain | Drugs used for cardiovascular system | 9.9 | (9.9–9.9) | 8.6 | (8.6–9.9) |
Amlodipine | Selective calcium channel blockers with mainly vascular effects (CCBs) | 5.0 | (5.0–5.4) | 4.7 | (4.7–5.0) | |
Nifedipine (controlled-release) | Selective calcium channel blockers with mainly vascular effects (CCBs) | 4.9 | (4.9–5.1) | 4.3 | (4.3–4.9) | |
Omeprazole (injection) | Drugs for peptic ulcer and gastro-oesophageal reflux disease (PPIs) | Drugs for acid-related disorders | NA | NA | 43.5 | (43.5–61.0) |
Pantoprazole (injection) | Drugs for peptic ulcer and gastro-oesophageal reflux disease (PPIs) | 125.0 | (125.0–125.0) | 108.7 | (108.7–125.0) | |
Ambroxol hydrochloride (injection) | Expectorants, excluding combinations with cough suppressants | Drugs for respiratory system | 18.6 | (18.6–18.6) | 16.2 | (16.2–18.6) |
. | . | . | Original drug . | |||
---|---|---|---|---|---|---|
. | . | . | Before ZMDP . | After ZMDP . | ||
INN . | Pharmacology classification . | Therapeutic subgroup . | Median price . | Price range . | Median price . | Price range . |
Acarbose | Blood glucose–lowering drugs, excluding insulins | Drugs used in diabetes | 12.4 | (9.2–14.7) | 12.4 | (12.4–12.4) |
Cefoperazone and sulbactam | Other beta-lactam antibacterials | Antibacterials for systemic use | 211.2 | (211.2–211.2) | 183.7 | (183.7–211.2) |
Cefuroxime sodium (injection) | Other beta-lactam antibacterials | 103.2 | (103.2–103.2) | 103.2 | (103.2–103.2) | |
Aspirin | Antithrombotic agents | Drugs used for blood and blood–forming organs | 0.4 | (0.4–0.5) | 0.4 | (0.4–0.5) |
Peritoneal dialysis solution | Peritoneal dialytics | 3.1 | (3.1–3.1) | 3.1 | (3.1–3.1) | |
Human albumin (injection) | Blood substitutes and plasma protein fractions | 1132.5 | (1132.5–1132.5) | 1132.5 | (1132.5–1132.5) | |
Mecobalamin (injection) | Vitamin B12 and folic acid | 6.9 | (6.9–7.2) | 6.0 | (6.0–6.0) | |
Clopidogrel | Antithrombotic agents | 20.5 | (19.0–20.5) | 16.5 | (16.5–19.0) | |
Atorvastatin calcium | Lipid modifying agents, plain | Drugs used for cardiovascular system | 9.9 | (9.9–9.9) | 8.6 | (8.6–9.9) |
Amlodipine | Selective calcium channel blockers with mainly vascular effects (CCBs) | 5.0 | (5.0–5.4) | 4.7 | (4.7–5.0) | |
Nifedipine (controlled-release) | Selective calcium channel blockers with mainly vascular effects (CCBs) | 4.9 | (4.9–5.1) | 4.3 | (4.3–4.9) | |
Omeprazole (injection) | Drugs for peptic ulcer and gastro-oesophageal reflux disease (PPIs) | Drugs for acid-related disorders | NA | NA | 43.5 | (43.5–61.0) |
Pantoprazole (injection) | Drugs for peptic ulcer and gastro-oesophageal reflux disease (PPIs) | 125.0 | (125.0–125.0) | 108.7 | (108.7–125.0) | |
Ambroxol hydrochloride (injection) | Expectorants, excluding combinations with cough suppressants | Drugs for respiratory system | 18.6 | (18.6–18.6) | 16.2 | (16.2–18.6) |
Notes: Shown are prices of drugs that were subjected to RP before and after the implementation of the ZMDP, during the period before the implementation of RP. Procurement price per DDD for each drug is reported. Exchange rates to US dollars in 2014: 1 United States dollar = 6.14 Chinese yuan. The ZMDP had been implemented in Sanming since January 2013, in Longyan’s secondary hospital since January 2013 and in Longyan’s tertiary hospital since June 2015. Dosage forms are specifically marked in parentheses for INNs that were not oral regular-release dosage forms. Pharmacology classification is based on ATC-3 codes and therapeutic subgroup is based on ATC-1 or ATC-2 codes.
. | . | . | Original drug . | |||
---|---|---|---|---|---|---|
. | . | . | Before ZMDP . | After ZMDP . | ||
INN . | Pharmacology classification . | Therapeutic subgroup . | Median price . | Price range . | Median price . | Price range . |
Acarbose | Blood glucose–lowering drugs, excluding insulins | Drugs used in diabetes | 12.4 | (9.2–14.7) | 12.4 | (12.4–12.4) |
Cefoperazone and sulbactam | Other beta-lactam antibacterials | Antibacterials for systemic use | 211.2 | (211.2–211.2) | 183.7 | (183.7–211.2) |
Cefuroxime sodium (injection) | Other beta-lactam antibacterials | 103.2 | (103.2–103.2) | 103.2 | (103.2–103.2) | |
Aspirin | Antithrombotic agents | Drugs used for blood and blood–forming organs | 0.4 | (0.4–0.5) | 0.4 | (0.4–0.5) |
Peritoneal dialysis solution | Peritoneal dialytics | 3.1 | (3.1–3.1) | 3.1 | (3.1–3.1) | |
Human albumin (injection) | Blood substitutes and plasma protein fractions | 1132.5 | (1132.5–1132.5) | 1132.5 | (1132.5–1132.5) | |
Mecobalamin (injection) | Vitamin B12 and folic acid | 6.9 | (6.9–7.2) | 6.0 | (6.0–6.0) | |
Clopidogrel | Antithrombotic agents | 20.5 | (19.0–20.5) | 16.5 | (16.5–19.0) | |
Atorvastatin calcium | Lipid modifying agents, plain | Drugs used for cardiovascular system | 9.9 | (9.9–9.9) | 8.6 | (8.6–9.9) |
Amlodipine | Selective calcium channel blockers with mainly vascular effects (CCBs) | 5.0 | (5.0–5.4) | 4.7 | (4.7–5.0) | |
Nifedipine (controlled-release) | Selective calcium channel blockers with mainly vascular effects (CCBs) | 4.9 | (4.9–5.1) | 4.3 | (4.3–4.9) | |
Omeprazole (injection) | Drugs for peptic ulcer and gastro-oesophageal reflux disease (PPIs) | Drugs for acid-related disorders | NA | NA | 43.5 | (43.5–61.0) |
Pantoprazole (injection) | Drugs for peptic ulcer and gastro-oesophageal reflux disease (PPIs) | 125.0 | (125.0–125.0) | 108.7 | (108.7–125.0) | |
Ambroxol hydrochloride (injection) | Expectorants, excluding combinations with cough suppressants | Drugs for respiratory system | 18.6 | (18.6–18.6) | 16.2 | (16.2–18.6) |
. | . | . | Original drug . | |||
---|---|---|---|---|---|---|
. | . | . | Before ZMDP . | After ZMDP . | ||
INN . | Pharmacology classification . | Therapeutic subgroup . | Median price . | Price range . | Median price . | Price range . |
Acarbose | Blood glucose–lowering drugs, excluding insulins | Drugs used in diabetes | 12.4 | (9.2–14.7) | 12.4 | (12.4–12.4) |
Cefoperazone and sulbactam | Other beta-lactam antibacterials | Antibacterials for systemic use | 211.2 | (211.2–211.2) | 183.7 | (183.7–211.2) |
Cefuroxime sodium (injection) | Other beta-lactam antibacterials | 103.2 | (103.2–103.2) | 103.2 | (103.2–103.2) | |
Aspirin | Antithrombotic agents | Drugs used for blood and blood–forming organs | 0.4 | (0.4–0.5) | 0.4 | (0.4–0.5) |
Peritoneal dialysis solution | Peritoneal dialytics | 3.1 | (3.1–3.1) | 3.1 | (3.1–3.1) | |
Human albumin (injection) | Blood substitutes and plasma protein fractions | 1132.5 | (1132.5–1132.5) | 1132.5 | (1132.5–1132.5) | |
Mecobalamin (injection) | Vitamin B12 and folic acid | 6.9 | (6.9–7.2) | 6.0 | (6.0–6.0) | |
Clopidogrel | Antithrombotic agents | 20.5 | (19.0–20.5) | 16.5 | (16.5–19.0) | |
Atorvastatin calcium | Lipid modifying agents, plain | Drugs used for cardiovascular system | 9.9 | (9.9–9.9) | 8.6 | (8.6–9.9) |
Amlodipine | Selective calcium channel blockers with mainly vascular effects (CCBs) | 5.0 | (5.0–5.4) | 4.7 | (4.7–5.0) | |
Nifedipine (controlled-release) | Selective calcium channel blockers with mainly vascular effects (CCBs) | 4.9 | (4.9–5.1) | 4.3 | (4.3–4.9) | |
Omeprazole (injection) | Drugs for peptic ulcer and gastro-oesophageal reflux disease (PPIs) | Drugs for acid-related disorders | NA | NA | 43.5 | (43.5–61.0) |
Pantoprazole (injection) | Drugs for peptic ulcer and gastro-oesophageal reflux disease (PPIs) | 125.0 | (125.0–125.0) | 108.7 | (108.7–125.0) | |
Ambroxol hydrochloride (injection) | Expectorants, excluding combinations with cough suppressants | Drugs for respiratory system | 18.6 | (18.6–18.6) | 16.2 | (16.2–18.6) |
Notes: Shown are prices of drugs that were subjected to RP before and after the implementation of the ZMDP, during the period before the implementation of RP. Procurement price per DDD for each drug is reported. Exchange rates to US dollars in 2014: 1 United States dollar = 6.14 Chinese yuan. The ZMDP had been implemented in Sanming since January 2013, in Longyan’s secondary hospital since January 2013 and in Longyan’s tertiary hospital since June 2015. Dosage forms are specifically marked in parentheses for INNs that were not oral regular-release dosage forms. Pharmacology classification is based on ATC-3 codes and therapeutic subgroup is based on ATC-1 or ATC-2 codes.
. | Impacts of ZMDP—DID estimates . | Impacts of RP—DID estimates . | ||
---|---|---|---|---|
. | Change (%) . | 95% CI . | Change (%) . | 95% CI . |
Acarbose | ||||
Original drugs | −8.1 | [−10.3%, 6.2%] | 3.4 | [−0.2%, 6.7%] |
Generic drugs | 6.3 | [−3.6%, 9.0%] | 14.1 | [−10.5%, 16.5%] |
Cefoperazone and sulbactam | ||||
Original drugs | −11.0 | [−13.5%, −8.4%] | −6.2 | [−9.2%, 3.1%] |
Generic drugs | −13.9 | [−41.1%, 25.9%] | −21.1 | [−39.8%, 3.5%] |
Cefuroxime sodium (injection) | ||||
Original drugs | −9.3 | [12.8%, −7.2%] | 1.0 | [−1.5%, 3.6%] |
Generic drugs | −10.5 | [−17.0%, −3.5%] | −7.1 | [13.3%, −2.2%] |
Aspirin | ||||
Original drugs | −8.4 | [−10.3%, 6.9%] | −0.3 | [−2.8%, 2.2%] |
Generic drugs | −29.7 | [−51.7%, 2.4%] | −66.0 | [−72.3%, −58.4%] |
Peritoneal dialysis solution | ||||
Original drugs | −14.9 | [−16.7%, 15.1%] | 2.0 | [−0.8%, 4.9%] |
Generic drugs | 2.4 | [−16.3%, 25.2%] | −65.5 | [−72.5%, 130.4%] |
Human albumin (injection) | ||||
Original drugs | −16.2 | [−20.3%, 13.4%] | −14.8 | [−22.2%, 7.2%] |
Generic drugs | −77.3 | [−93.3%, −22.9%] | −94.4 | [−97.8%, 21.3%] |
Mecobalamin (injection) | ||||
Original drugs | −17.4 | [−24.8%, 10.3%] | 9.4 | [−1.4%, 18.0%] |
Generic drugs | NA | NA | 9.1 | [−1.5%, 20.8%] |
Clopidogrel | ||||
Original drugs | −13.3 | [−14.1%, −12.4%] | 1.3 | [−1.4%, 4.0%] |
Generic drugs | NA | NA | −64.9 | [−74.4%, 7.6%] |
Atorvastatin calcium | ||||
Original drugs | −11.5 | [−13.9%, 10.0%] | 1.5 | [−1.4%, 4.4%] |
Generic drugs | −10.3 | [−12.3%, −8.3%] | −9.2 | [−14.5%, 3.8%] |
Amlodipine | ||||
Original drugs | −13.2 | [−15.9%, 11.5%] | 4.1 | [−0.2%, 8.5%] |
Generic drugs | −41.5 | [−48.8%, 49.6%] | −50.7 | [−56.7%, −43.8%] |
Nifedipine (controlled release) | ||||
Original drugs | −10.0 | [−20.8%, 2.2%] | 13.3 | [−2.6%, 25.1%] |
Generic drugs | NA | NA | NA | NA |
Omeprazole (injection) | ||||
Original drugs | −7.2 | [−9.9%, 4.5%] | 7.8 | [−4.2%, 11.3%] |
Generic drugs | 267.8 | [236.4%, 302.2%] | −43.4 | [−58.1%, 2.7%] |
Pantoprazole (injection) | ||||
Original drugs | NA | NA | 13.4 | [−2.5%, 25.3%] |
Generic drugs | NA | NA | 450.2 | [296.9%, 662.8%] |
Ambroxol hydrochloride (injection) | ||||
Original drugs | NA | NA | −3.0 | [−6.8%, 0.9%] |
Generic drugs | NA | NA | NA | NA |
. | Impacts of ZMDP—DID estimates . | Impacts of RP—DID estimates . | ||
---|---|---|---|---|
. | Change (%) . | 95% CI . | Change (%) . | 95% CI . |
Acarbose | ||||
Original drugs | −8.1 | [−10.3%, 6.2%] | 3.4 | [−0.2%, 6.7%] |
Generic drugs | 6.3 | [−3.6%, 9.0%] | 14.1 | [−10.5%, 16.5%] |
Cefoperazone and sulbactam | ||||
Original drugs | −11.0 | [−13.5%, −8.4%] | −6.2 | [−9.2%, 3.1%] |
Generic drugs | −13.9 | [−41.1%, 25.9%] | −21.1 | [−39.8%, 3.5%] |
Cefuroxime sodium (injection) | ||||
Original drugs | −9.3 | [12.8%, −7.2%] | 1.0 | [−1.5%, 3.6%] |
Generic drugs | −10.5 | [−17.0%, −3.5%] | −7.1 | [13.3%, −2.2%] |
Aspirin | ||||
Original drugs | −8.4 | [−10.3%, 6.9%] | −0.3 | [−2.8%, 2.2%] |
Generic drugs | −29.7 | [−51.7%, 2.4%] | −66.0 | [−72.3%, −58.4%] |
Peritoneal dialysis solution | ||||
Original drugs | −14.9 | [−16.7%, 15.1%] | 2.0 | [−0.8%, 4.9%] |
Generic drugs | 2.4 | [−16.3%, 25.2%] | −65.5 | [−72.5%, 130.4%] |
Human albumin (injection) | ||||
Original drugs | −16.2 | [−20.3%, 13.4%] | −14.8 | [−22.2%, 7.2%] |
Generic drugs | −77.3 | [−93.3%, −22.9%] | −94.4 | [−97.8%, 21.3%] |
Mecobalamin (injection) | ||||
Original drugs | −17.4 | [−24.8%, 10.3%] | 9.4 | [−1.4%, 18.0%] |
Generic drugs | NA | NA | 9.1 | [−1.5%, 20.8%] |
Clopidogrel | ||||
Original drugs | −13.3 | [−14.1%, −12.4%] | 1.3 | [−1.4%, 4.0%] |
Generic drugs | NA | NA | −64.9 | [−74.4%, 7.6%] |
Atorvastatin calcium | ||||
Original drugs | −11.5 | [−13.9%, 10.0%] | 1.5 | [−1.4%, 4.4%] |
Generic drugs | −10.3 | [−12.3%, −8.3%] | −9.2 | [−14.5%, 3.8%] |
Amlodipine | ||||
Original drugs | −13.2 | [−15.9%, 11.5%] | 4.1 | [−0.2%, 8.5%] |
Generic drugs | −41.5 | [−48.8%, 49.6%] | −50.7 | [−56.7%, −43.8%] |
Nifedipine (controlled release) | ||||
Original drugs | −10.0 | [−20.8%, 2.2%] | 13.3 | [−2.6%, 25.1%] |
Generic drugs | NA | NA | NA | NA |
Omeprazole (injection) | ||||
Original drugs | −7.2 | [−9.9%, 4.5%] | 7.8 | [−4.2%, 11.3%] |
Generic drugs | 267.8 | [236.4%, 302.2%] | −43.4 | [−58.1%, 2.7%] |
Pantoprazole (injection) | ||||
Original drugs | NA | NA | 13.4 | [−2.5%, 25.3%] |
Generic drugs | NA | NA | 450.2 | [296.9%, 662.8%] |
Ambroxol hydrochloride (injection) | ||||
Original drugs | NA | NA | −3.0 | [−6.8%, 0.9%] |
Generic drugs | NA | NA | NA | NA |
Notes: We performed GLMs to estimate the impacts of ZMDP and RP, respectively. For impacts of RP, regressing monthly median prices of each drug substance on the interaction term between a dummy variable equalling 1 after 2014.9 and 0 before 2014.9, and a dummy variable indicating RP intervention, during 2012.1–2016.12. For impacts of ZMDP, regressing monthly median prices on the interaction term between a dummy variable equalling 1 after 2013.1 and 0 before 2013.1, and a dummy variable indicating ZMDP intervention, during 2012.1–2014.9.
. | Impacts of ZMDP—DID estimates . | Impacts of RP—DID estimates . | ||
---|---|---|---|---|
. | Change (%) . | 95% CI . | Change (%) . | 95% CI . |
Acarbose | ||||
Original drugs | −8.1 | [−10.3%, 6.2%] | 3.4 | [−0.2%, 6.7%] |
Generic drugs | 6.3 | [−3.6%, 9.0%] | 14.1 | [−10.5%, 16.5%] |
Cefoperazone and sulbactam | ||||
Original drugs | −11.0 | [−13.5%, −8.4%] | −6.2 | [−9.2%, 3.1%] |
Generic drugs | −13.9 | [−41.1%, 25.9%] | −21.1 | [−39.8%, 3.5%] |
Cefuroxime sodium (injection) | ||||
Original drugs | −9.3 | [12.8%, −7.2%] | 1.0 | [−1.5%, 3.6%] |
Generic drugs | −10.5 | [−17.0%, −3.5%] | −7.1 | [13.3%, −2.2%] |
Aspirin | ||||
Original drugs | −8.4 | [−10.3%, 6.9%] | −0.3 | [−2.8%, 2.2%] |
Generic drugs | −29.7 | [−51.7%, 2.4%] | −66.0 | [−72.3%, −58.4%] |
Peritoneal dialysis solution | ||||
Original drugs | −14.9 | [−16.7%, 15.1%] | 2.0 | [−0.8%, 4.9%] |
Generic drugs | 2.4 | [−16.3%, 25.2%] | −65.5 | [−72.5%, 130.4%] |
Human albumin (injection) | ||||
Original drugs | −16.2 | [−20.3%, 13.4%] | −14.8 | [−22.2%, 7.2%] |
Generic drugs | −77.3 | [−93.3%, −22.9%] | −94.4 | [−97.8%, 21.3%] |
Mecobalamin (injection) | ||||
Original drugs | −17.4 | [−24.8%, 10.3%] | 9.4 | [−1.4%, 18.0%] |
Generic drugs | NA | NA | 9.1 | [−1.5%, 20.8%] |
Clopidogrel | ||||
Original drugs | −13.3 | [−14.1%, −12.4%] | 1.3 | [−1.4%, 4.0%] |
Generic drugs | NA | NA | −64.9 | [−74.4%, 7.6%] |
Atorvastatin calcium | ||||
Original drugs | −11.5 | [−13.9%, 10.0%] | 1.5 | [−1.4%, 4.4%] |
Generic drugs | −10.3 | [−12.3%, −8.3%] | −9.2 | [−14.5%, 3.8%] |
Amlodipine | ||||
Original drugs | −13.2 | [−15.9%, 11.5%] | 4.1 | [−0.2%, 8.5%] |
Generic drugs | −41.5 | [−48.8%, 49.6%] | −50.7 | [−56.7%, −43.8%] |
Nifedipine (controlled release) | ||||
Original drugs | −10.0 | [−20.8%, 2.2%] | 13.3 | [−2.6%, 25.1%] |
Generic drugs | NA | NA | NA | NA |
Omeprazole (injection) | ||||
Original drugs | −7.2 | [−9.9%, 4.5%] | 7.8 | [−4.2%, 11.3%] |
Generic drugs | 267.8 | [236.4%, 302.2%] | −43.4 | [−58.1%, 2.7%] |
Pantoprazole (injection) | ||||
Original drugs | NA | NA | 13.4 | [−2.5%, 25.3%] |
Generic drugs | NA | NA | 450.2 | [296.9%, 662.8%] |
Ambroxol hydrochloride (injection) | ||||
Original drugs | NA | NA | −3.0 | [−6.8%, 0.9%] |
Generic drugs | NA | NA | NA | NA |
. | Impacts of ZMDP—DID estimates . | Impacts of RP—DID estimates . | ||
---|---|---|---|---|
. | Change (%) . | 95% CI . | Change (%) . | 95% CI . |
Acarbose | ||||
Original drugs | −8.1 | [−10.3%, 6.2%] | 3.4 | [−0.2%, 6.7%] |
Generic drugs | 6.3 | [−3.6%, 9.0%] | 14.1 | [−10.5%, 16.5%] |
Cefoperazone and sulbactam | ||||
Original drugs | −11.0 | [−13.5%, −8.4%] | −6.2 | [−9.2%, 3.1%] |
Generic drugs | −13.9 | [−41.1%, 25.9%] | −21.1 | [−39.8%, 3.5%] |
Cefuroxime sodium (injection) | ||||
Original drugs | −9.3 | [12.8%, −7.2%] | 1.0 | [−1.5%, 3.6%] |
Generic drugs | −10.5 | [−17.0%, −3.5%] | −7.1 | [13.3%, −2.2%] |
Aspirin | ||||
Original drugs | −8.4 | [−10.3%, 6.9%] | −0.3 | [−2.8%, 2.2%] |
Generic drugs | −29.7 | [−51.7%, 2.4%] | −66.0 | [−72.3%, −58.4%] |
Peritoneal dialysis solution | ||||
Original drugs | −14.9 | [−16.7%, 15.1%] | 2.0 | [−0.8%, 4.9%] |
Generic drugs | 2.4 | [−16.3%, 25.2%] | −65.5 | [−72.5%, 130.4%] |
Human albumin (injection) | ||||
Original drugs | −16.2 | [−20.3%, 13.4%] | −14.8 | [−22.2%, 7.2%] |
Generic drugs | −77.3 | [−93.3%, −22.9%] | −94.4 | [−97.8%, 21.3%] |
Mecobalamin (injection) | ||||
Original drugs | −17.4 | [−24.8%, 10.3%] | 9.4 | [−1.4%, 18.0%] |
Generic drugs | NA | NA | 9.1 | [−1.5%, 20.8%] |
Clopidogrel | ||||
Original drugs | −13.3 | [−14.1%, −12.4%] | 1.3 | [−1.4%, 4.0%] |
Generic drugs | NA | NA | −64.9 | [−74.4%, 7.6%] |
Atorvastatin calcium | ||||
Original drugs | −11.5 | [−13.9%, 10.0%] | 1.5 | [−1.4%, 4.4%] |
Generic drugs | −10.3 | [−12.3%, −8.3%] | −9.2 | [−14.5%, 3.8%] |
Amlodipine | ||||
Original drugs | −13.2 | [−15.9%, 11.5%] | 4.1 | [−0.2%, 8.5%] |
Generic drugs | −41.5 | [−48.8%, 49.6%] | −50.7 | [−56.7%, −43.8%] |
Nifedipine (controlled release) | ||||
Original drugs | −10.0 | [−20.8%, 2.2%] | 13.3 | [−2.6%, 25.1%] |
Generic drugs | NA | NA | NA | NA |
Omeprazole (injection) | ||||
Original drugs | −7.2 | [−9.9%, 4.5%] | 7.8 | [−4.2%, 11.3%] |
Generic drugs | 267.8 | [236.4%, 302.2%] | −43.4 | [−58.1%, 2.7%] |
Pantoprazole (injection) | ||||
Original drugs | NA | NA | 13.4 | [−2.5%, 25.3%] |
Generic drugs | NA | NA | 450.2 | [296.9%, 662.8%] |
Ambroxol hydrochloride (injection) | ||||
Original drugs | NA | NA | −3.0 | [−6.8%, 0.9%] |
Generic drugs | NA | NA | NA | NA |
Notes: We performed GLMs to estimate the impacts of ZMDP and RP, respectively. For impacts of RP, regressing monthly median prices of each drug substance on the interaction term between a dummy variable equalling 1 after 2014.9 and 0 before 2014.9, and a dummy variable indicating RP intervention, during 2012.1–2016.12. For impacts of ZMDP, regressing monthly median prices on the interaction term between a dummy variable equalling 1 after 2013.1 and 0 before 2013.1, and a dummy variable indicating ZMDP intervention, during 2012.1–2014.9.
. | RP region (Sanming) . | Control region (Longyan) . | DID estimates . | |||||
---|---|---|---|---|---|---|---|---|
. | Before RP . | After RP . | Change (%) . | Before RP . | After RP . | Change (%) . | Change (%) . | 95% CI . |
Monthly volumes and costs | ||||||||
1. Drugs for diabetes | ||||||||
Total volumes (1000 DDDs) | 9.7 | 9.5 | −1.5 | 6.2 | 7.3 | 17.5 | −16.2 | [−40.6%, 18.3%] |
Original drugs | 8.4 | 5.9 | −29.2 | 5.1 | 6.0 | 17.2 | −39.6 | [−60.8%, −6.9%] |
Generic drugs | 1.3 | 3.6 | 176.8 | 1.1 | 1.3 | 19.0 | 132.7 | [42.3%, 280.5%] |
Total costs (1000 CNY) | 121.5 | 106.5 | −12.3 | 82.4 | 81.4 | −1.2 | −11.3 | [−38.8%, 28.6%] |
2. Antibacterials for systemic use | ||||||||
Total volumes (1000 DDDs) | 6.4 | 6.9 | 8.1 | 6.5 | 9.2 | 41.3 | −23.5 | [−43.1%, 2.9%] |
Original drugs | 4.0 | 1.9 | −51.8 | 6.5 | 9.2 | 41.3 | −65.9 | [−77.0%, −49.5%] |
Generic drugs | 2.3 | 4.9 | 111.2 | NA | NA | NA | NA | NA |
Total costs (1000 CNY) | 915.2 | 516.5 | −43.6 | 1409.3 | 1796.5 | 27.5 | −55.7 | [−70.8%, −33.0%] |
3. Drugs for blood and blood–forming organs | ||||||||
Total volumes (1000 DDDs) | 30.8 | 32.7 | 6.0 | 24.6 | 33.9 | 37.5 | −23.0 | [−42.0%, 2.3%] |
Original drugs | 25.5 | 17.6 | −31.0 | 21.7 | 30.1 | 38.7 | −50.2 | [−64.3%, −30.7%] |
Generic drugs | 5.3 | 15.0 | 184.3 | 2.9 | 3.7 | 28.6 | 121.1 | [40.6%, 247.9%] |
Total costs (1000 CNY) | 245.1 | 280.3 | 14.4 | 506.8 | 609.4 | 20.2 | −4.9 | [−34.9%, 39.0%] |
4. Drugs for cardiovascular system | ||||||||
Total volumes (1000 DDDs) | 41.2 | 42.4 | 2.8 | 40.0 | 56.5 | 41.2 | −27.2 | [−42.1%, −8.4%] |
Original drugs | 34.4 | 24.3 | −29.4 | 26.9 | 44.8 | 66.3 | −57.6 | [−68.9%, −42.0%] |
Generic drugs | 6.8 | 18.1 | 164.8 | 13.1 | 11.7 | −10.5 | 195.9 | [108.8%, 319.2%] |
Total costs (1000 CNY) | 259.7 | 196.2 | −24.4 | 237.6 | 279.7 | 17.7% | −35.8% | [−53.6%, −11.3%] |
5. Drugs for acid-related disorders | ||||||||
Total volumes (1000 DDDs) | 65.0 | 47.9 | −26.3 | 18.1 | 24.6 | 36.1 | −45.9 | [−59.7%, −27.2%] |
Original drugs | 35.6 | 17.2 | −51.6 | 8.0 | 11.5 | 43.8 | −66.3 | [−79.9%, −43.6%] |
Generic drugs | 29.4 | 30.7 | 4.2 | 10.0 | 13.0 | 29.9 | −19.8 | [−42.7%, 12.2%] |
Total costs (1000 CNY) | 2521.7 | 1244.7 | −50.6 | 1037.5 | 1256.4 | 21.1 | −59.2 | [−74.9%, −33.8%] |
6. Drugs for respiratory system | ||||||||
Total volumes (1000 DDDs) | 114.5 | 114.4 | −0.1% | 71.7 | 70.8 | −1.3% | 1.2% | [−25.7%, 37.9%] |
Original drugs | 97.3 | 39.3 | −59.5% | 62.9 | 59.9 | −4.8% | −57.5% | [−73.3%, −32.4%] |
Generic drugs | 17.2 | 75.0 | 336.0% | 8.7 | 10.8 | 24.0% | 251.5% | [116.7%, 470.3%] |
Total costs (1000 CNY) | 1886.6 | 1074.7 | −43.0% | 1243.0 | 1138.9 | −8.4% | −37.8% | [−57.0%, −10.2%] |
. | RP region (Sanming) . | Control region (Longyan) . | DID estimates . | |||||
---|---|---|---|---|---|---|---|---|
. | Before RP . | After RP . | Change (%) . | Before RP . | After RP . | Change (%) . | Change (%) . | 95% CI . |
Monthly volumes and costs | ||||||||
1. Drugs for diabetes | ||||||||
Total volumes (1000 DDDs) | 9.7 | 9.5 | −1.5 | 6.2 | 7.3 | 17.5 | −16.2 | [−40.6%, 18.3%] |
Original drugs | 8.4 | 5.9 | −29.2 | 5.1 | 6.0 | 17.2 | −39.6 | [−60.8%, −6.9%] |
Generic drugs | 1.3 | 3.6 | 176.8 | 1.1 | 1.3 | 19.0 | 132.7 | [42.3%, 280.5%] |
Total costs (1000 CNY) | 121.5 | 106.5 | −12.3 | 82.4 | 81.4 | −1.2 | −11.3 | [−38.8%, 28.6%] |
2. Antibacterials for systemic use | ||||||||
Total volumes (1000 DDDs) | 6.4 | 6.9 | 8.1 | 6.5 | 9.2 | 41.3 | −23.5 | [−43.1%, 2.9%] |
Original drugs | 4.0 | 1.9 | −51.8 | 6.5 | 9.2 | 41.3 | −65.9 | [−77.0%, −49.5%] |
Generic drugs | 2.3 | 4.9 | 111.2 | NA | NA | NA | NA | NA |
Total costs (1000 CNY) | 915.2 | 516.5 | −43.6 | 1409.3 | 1796.5 | 27.5 | −55.7 | [−70.8%, −33.0%] |
3. Drugs for blood and blood–forming organs | ||||||||
Total volumes (1000 DDDs) | 30.8 | 32.7 | 6.0 | 24.6 | 33.9 | 37.5 | −23.0 | [−42.0%, 2.3%] |
Original drugs | 25.5 | 17.6 | −31.0 | 21.7 | 30.1 | 38.7 | −50.2 | [−64.3%, −30.7%] |
Generic drugs | 5.3 | 15.0 | 184.3 | 2.9 | 3.7 | 28.6 | 121.1 | [40.6%, 247.9%] |
Total costs (1000 CNY) | 245.1 | 280.3 | 14.4 | 506.8 | 609.4 | 20.2 | −4.9 | [−34.9%, 39.0%] |
4. Drugs for cardiovascular system | ||||||||
Total volumes (1000 DDDs) | 41.2 | 42.4 | 2.8 | 40.0 | 56.5 | 41.2 | −27.2 | [−42.1%, −8.4%] |
Original drugs | 34.4 | 24.3 | −29.4 | 26.9 | 44.8 | 66.3 | −57.6 | [−68.9%, −42.0%] |
Generic drugs | 6.8 | 18.1 | 164.8 | 13.1 | 11.7 | −10.5 | 195.9 | [108.8%, 319.2%] |
Total costs (1000 CNY) | 259.7 | 196.2 | −24.4 | 237.6 | 279.7 | 17.7% | −35.8% | [−53.6%, −11.3%] |
5. Drugs for acid-related disorders | ||||||||
Total volumes (1000 DDDs) | 65.0 | 47.9 | −26.3 | 18.1 | 24.6 | 36.1 | −45.9 | [−59.7%, −27.2%] |
Original drugs | 35.6 | 17.2 | −51.6 | 8.0 | 11.5 | 43.8 | −66.3 | [−79.9%, −43.6%] |
Generic drugs | 29.4 | 30.7 | 4.2 | 10.0 | 13.0 | 29.9 | −19.8 | [−42.7%, 12.2%] |
Total costs (1000 CNY) | 2521.7 | 1244.7 | −50.6 | 1037.5 | 1256.4 | 21.1 | −59.2 | [−74.9%, −33.8%] |
6. Drugs for respiratory system | ||||||||
Total volumes (1000 DDDs) | 114.5 | 114.4 | −0.1% | 71.7 | 70.8 | −1.3% | 1.2% | [−25.7%, 37.9%] |
Original drugs | 97.3 | 39.3 | −59.5% | 62.9 | 59.9 | −4.8% | −57.5% | [−73.3%, −32.4%] |
Generic drugs | 17.2 | 75.0 | 336.0% | 8.7 | 10.8 | 24.0% | 251.5% | [116.7%, 470.3%] |
Total costs (1000 CNY) | 1886.6 | 1074.7 | −43.0% | 1243.0 | 1138.9 | −8.4% | −37.8% | [−57.0%, −10.2%] |
Notes: Exchange rates to US dollars in 2014: 1 United States dollar = 6.14 Chinese yuan. Therapeutic subgroup was based on ATC-1 or ATC-2 codes. RP policy for 15 drug substances was piloted in Sanming, China from September 2014 to December 2016. Amoxicillin sodium and clavulanate potassium (injection) was excluded from the table because of no procurement of its original drug in the study period.
. | RP region (Sanming) . | Control region (Longyan) . | DID estimates . | |||||
---|---|---|---|---|---|---|---|---|
. | Before RP . | After RP . | Change (%) . | Before RP . | After RP . | Change (%) . | Change (%) . | 95% CI . |
Monthly volumes and costs | ||||||||
1. Drugs for diabetes | ||||||||
Total volumes (1000 DDDs) | 9.7 | 9.5 | −1.5 | 6.2 | 7.3 | 17.5 | −16.2 | [−40.6%, 18.3%] |
Original drugs | 8.4 | 5.9 | −29.2 | 5.1 | 6.0 | 17.2 | −39.6 | [−60.8%, −6.9%] |
Generic drugs | 1.3 | 3.6 | 176.8 | 1.1 | 1.3 | 19.0 | 132.7 | [42.3%, 280.5%] |
Total costs (1000 CNY) | 121.5 | 106.5 | −12.3 | 82.4 | 81.4 | −1.2 | −11.3 | [−38.8%, 28.6%] |
2. Antibacterials for systemic use | ||||||||
Total volumes (1000 DDDs) | 6.4 | 6.9 | 8.1 | 6.5 | 9.2 | 41.3 | −23.5 | [−43.1%, 2.9%] |
Original drugs | 4.0 | 1.9 | −51.8 | 6.5 | 9.2 | 41.3 | −65.9 | [−77.0%, −49.5%] |
Generic drugs | 2.3 | 4.9 | 111.2 | NA | NA | NA | NA | NA |
Total costs (1000 CNY) | 915.2 | 516.5 | −43.6 | 1409.3 | 1796.5 | 27.5 | −55.7 | [−70.8%, −33.0%] |
3. Drugs for blood and blood–forming organs | ||||||||
Total volumes (1000 DDDs) | 30.8 | 32.7 | 6.0 | 24.6 | 33.9 | 37.5 | −23.0 | [−42.0%, 2.3%] |
Original drugs | 25.5 | 17.6 | −31.0 | 21.7 | 30.1 | 38.7 | −50.2 | [−64.3%, −30.7%] |
Generic drugs | 5.3 | 15.0 | 184.3 | 2.9 | 3.7 | 28.6 | 121.1 | [40.6%, 247.9%] |
Total costs (1000 CNY) | 245.1 | 280.3 | 14.4 | 506.8 | 609.4 | 20.2 | −4.9 | [−34.9%, 39.0%] |
4. Drugs for cardiovascular system | ||||||||
Total volumes (1000 DDDs) | 41.2 | 42.4 | 2.8 | 40.0 | 56.5 | 41.2 | −27.2 | [−42.1%, −8.4%] |
Original drugs | 34.4 | 24.3 | −29.4 | 26.9 | 44.8 | 66.3 | −57.6 | [−68.9%, −42.0%] |
Generic drugs | 6.8 | 18.1 | 164.8 | 13.1 | 11.7 | −10.5 | 195.9 | [108.8%, 319.2%] |
Total costs (1000 CNY) | 259.7 | 196.2 | −24.4 | 237.6 | 279.7 | 17.7% | −35.8% | [−53.6%, −11.3%] |
5. Drugs for acid-related disorders | ||||||||
Total volumes (1000 DDDs) | 65.0 | 47.9 | −26.3 | 18.1 | 24.6 | 36.1 | −45.9 | [−59.7%, −27.2%] |
Original drugs | 35.6 | 17.2 | −51.6 | 8.0 | 11.5 | 43.8 | −66.3 | [−79.9%, −43.6%] |
Generic drugs | 29.4 | 30.7 | 4.2 | 10.0 | 13.0 | 29.9 | −19.8 | [−42.7%, 12.2%] |
Total costs (1000 CNY) | 2521.7 | 1244.7 | −50.6 | 1037.5 | 1256.4 | 21.1 | −59.2 | [−74.9%, −33.8%] |
6. Drugs for respiratory system | ||||||||
Total volumes (1000 DDDs) | 114.5 | 114.4 | −0.1% | 71.7 | 70.8 | −1.3% | 1.2% | [−25.7%, 37.9%] |
Original drugs | 97.3 | 39.3 | −59.5% | 62.9 | 59.9 | −4.8% | −57.5% | [−73.3%, −32.4%] |
Generic drugs | 17.2 | 75.0 | 336.0% | 8.7 | 10.8 | 24.0% | 251.5% | [116.7%, 470.3%] |
Total costs (1000 CNY) | 1886.6 | 1074.7 | −43.0% | 1243.0 | 1138.9 | −8.4% | −37.8% | [−57.0%, −10.2%] |
. | RP region (Sanming) . | Control region (Longyan) . | DID estimates . | |||||
---|---|---|---|---|---|---|---|---|
. | Before RP . | After RP . | Change (%) . | Before RP . | After RP . | Change (%) . | Change (%) . | 95% CI . |
Monthly volumes and costs | ||||||||
1. Drugs for diabetes | ||||||||
Total volumes (1000 DDDs) | 9.7 | 9.5 | −1.5 | 6.2 | 7.3 | 17.5 | −16.2 | [−40.6%, 18.3%] |
Original drugs | 8.4 | 5.9 | −29.2 | 5.1 | 6.0 | 17.2 | −39.6 | [−60.8%, −6.9%] |
Generic drugs | 1.3 | 3.6 | 176.8 | 1.1 | 1.3 | 19.0 | 132.7 | [42.3%, 280.5%] |
Total costs (1000 CNY) | 121.5 | 106.5 | −12.3 | 82.4 | 81.4 | −1.2 | −11.3 | [−38.8%, 28.6%] |
2. Antibacterials for systemic use | ||||||||
Total volumes (1000 DDDs) | 6.4 | 6.9 | 8.1 | 6.5 | 9.2 | 41.3 | −23.5 | [−43.1%, 2.9%] |
Original drugs | 4.0 | 1.9 | −51.8 | 6.5 | 9.2 | 41.3 | −65.9 | [−77.0%, −49.5%] |
Generic drugs | 2.3 | 4.9 | 111.2 | NA | NA | NA | NA | NA |
Total costs (1000 CNY) | 915.2 | 516.5 | −43.6 | 1409.3 | 1796.5 | 27.5 | −55.7 | [−70.8%, −33.0%] |
3. Drugs for blood and blood–forming organs | ||||||||
Total volumes (1000 DDDs) | 30.8 | 32.7 | 6.0 | 24.6 | 33.9 | 37.5 | −23.0 | [−42.0%, 2.3%] |
Original drugs | 25.5 | 17.6 | −31.0 | 21.7 | 30.1 | 38.7 | −50.2 | [−64.3%, −30.7%] |
Generic drugs | 5.3 | 15.0 | 184.3 | 2.9 | 3.7 | 28.6 | 121.1 | [40.6%, 247.9%] |
Total costs (1000 CNY) | 245.1 | 280.3 | 14.4 | 506.8 | 609.4 | 20.2 | −4.9 | [−34.9%, 39.0%] |
4. Drugs for cardiovascular system | ||||||||
Total volumes (1000 DDDs) | 41.2 | 42.4 | 2.8 | 40.0 | 56.5 | 41.2 | −27.2 | [−42.1%, −8.4%] |
Original drugs | 34.4 | 24.3 | −29.4 | 26.9 | 44.8 | 66.3 | −57.6 | [−68.9%, −42.0%] |
Generic drugs | 6.8 | 18.1 | 164.8 | 13.1 | 11.7 | −10.5 | 195.9 | [108.8%, 319.2%] |
Total costs (1000 CNY) | 259.7 | 196.2 | −24.4 | 237.6 | 279.7 | 17.7% | −35.8% | [−53.6%, −11.3%] |
5. Drugs for acid-related disorders | ||||||||
Total volumes (1000 DDDs) | 65.0 | 47.9 | −26.3 | 18.1 | 24.6 | 36.1 | −45.9 | [−59.7%, −27.2%] |
Original drugs | 35.6 | 17.2 | −51.6 | 8.0 | 11.5 | 43.8 | −66.3 | [−79.9%, −43.6%] |
Generic drugs | 29.4 | 30.7 | 4.2 | 10.0 | 13.0 | 29.9 | −19.8 | [−42.7%, 12.2%] |
Total costs (1000 CNY) | 2521.7 | 1244.7 | −50.6 | 1037.5 | 1256.4 | 21.1 | −59.2 | [−74.9%, −33.8%] |
6. Drugs for respiratory system | ||||||||
Total volumes (1000 DDDs) | 114.5 | 114.4 | −0.1% | 71.7 | 70.8 | −1.3% | 1.2% | [−25.7%, 37.9%] |
Original drugs | 97.3 | 39.3 | −59.5% | 62.9 | 59.9 | −4.8% | −57.5% | [−73.3%, −32.4%] |
Generic drugs | 17.2 | 75.0 | 336.0% | 8.7 | 10.8 | 24.0% | 251.5% | [116.7%, 470.3%] |
Total costs (1000 CNY) | 1886.6 | 1074.7 | −43.0% | 1243.0 | 1138.9 | −8.4% | −37.8% | [−57.0%, −10.2%] |
Notes: Exchange rates to US dollars in 2014: 1 United States dollar = 6.14 Chinese yuan. Therapeutic subgroup was based on ATC-1 or ATC-2 codes. RP policy for 15 drug substances was piloted in Sanming, China from September 2014 to December 2016. Amoxicillin sodium and clavulanate potassium (injection) was excluded from the table because of no procurement of its original drug in the study period.
Appendix 5
Common trends tests
Outcome indicators . | β2 . | P . | β4 . | P . |
---|---|---|---|---|
Table 3 | ||||
Monthly volumes and costs | ||||
Inpatient sector | ||||
Volumes | 0.01 | 0.0261 | 0.00 | 0.8659 |
Costs | 0.00 | 0.8515 | −0.01 | 0.5198 |
Outpatient sector | ||||
Volumes | 0.02 | 0.0290 | 0.00 | 0.9724 |
Costs | 0.02 | 0.0002 | −0.01 | 0.4301 |
Monthly hospital visits | ||||
Inpatient admissions | 0.01 | 0.1293 | 0.01 | 0.2471 |
Outpatient visits | 0.02 | 0.1121 | 0.01 | 0.6004 |
Table 2 | ||||
Monthly volumes and costs | ||||
Inpatient sector | ||||
Volumes | 0.00 | 0.9317 | 0.00 | 0.9396 |
Costs | 0.00 | 0.9123 | −0.03 | 0.5498 |
Outpatient sector | ||||
Volumes | 0.04 | 0.3208 | −0.01 | 0.7978 |
Costs | 0.05 | 0.0561 | −0.02 | 0.6624 |
Monthly hospital visits | ||||
Inpatient admissions | 0.03 | 0.0210 | 0.05 | 0.1002 |
Outpatient visits | 0.02 | 0.1516 | −0.03 | 0.1259 |
Table 4 | ||||
Monthly volumes and costs | ||||
Tertiary hospitals | ||||
Volumes | 0.01 | 0.0252 | −0.01 | 0.3314 |
Costs | 0.00 | 0.9177 | −0.01 | 0.3907 |
Secondary hospitals | ||||
Volumes | 0.01 | 0.3160 | 0.02 | 0.0525 |
Costs | 0.02 | 0.0293 | 0.00 | 0.7548 |
Monthly inpatient admission | ||||
Tertiary hospitals | 0.01 | 0.0000 | 0.00 | 0.9872 |
Secondary hospitals | 0.01 | 0.0000 | 0.00 | 0.0542 |
Appendix 4 | ||||
Monthly volumes and costs | ||||
1. Drugs for diabetes | ||||
Volumes | 0.01 | 0.2025 | −0.01 | 0.6668 |
Costs | 0.01 | 0.3300 | −0.01 | 0.4080 |
2. Antibacterials for systemic use | ||||
Volumes | 0.02 | 0.0386 | 0.04 | 0.0605 |
Costs | 0.02 | 0.1156 | 0.01 | 0.4314 |
3. Drugs for blood and blood–forming organs | ||||
Volumes | 0.01 | 0.2422 | 0.01 | 0.3635 |
Costs | 0.01 | 0.6513 | 0.05 | 0.0610 |
4. Drugs for cardiovascular system | ||||
Volumes | 0.02 | 0.0002 | −0.02 | 0.0773 |
Costs | 0.02 | 0.0171 | −0.02 | 0.1220 |
5. Drugs for acid-related disorders | ||||
Volumes | −0.02 | 0.0759 | 0.02 | 0.1250 |
Costs | −0.03 | 0.0327 | 0.00 | 0.9899 |
6. Drugs for respiratory system | ||||
Volumes | 0.00 | 0.6542 | 0.00 | 0.8181 |
Costs | 0.00 | 0.7517 | −0.01 | 0.4635 |
Outcome indicators . | β2 . | P . | β4 . | P . |
---|---|---|---|---|
Table 3 | ||||
Monthly volumes and costs | ||||
Inpatient sector | ||||
Volumes | 0.01 | 0.0261 | 0.00 | 0.8659 |
Costs | 0.00 | 0.8515 | −0.01 | 0.5198 |
Outpatient sector | ||||
Volumes | 0.02 | 0.0290 | 0.00 | 0.9724 |
Costs | 0.02 | 0.0002 | −0.01 | 0.4301 |
Monthly hospital visits | ||||
Inpatient admissions | 0.01 | 0.1293 | 0.01 | 0.2471 |
Outpatient visits | 0.02 | 0.1121 | 0.01 | 0.6004 |
Table 2 | ||||
Monthly volumes and costs | ||||
Inpatient sector | ||||
Volumes | 0.00 | 0.9317 | 0.00 | 0.9396 |
Costs | 0.00 | 0.9123 | −0.03 | 0.5498 |
Outpatient sector | ||||
Volumes | 0.04 | 0.3208 | −0.01 | 0.7978 |
Costs | 0.05 | 0.0561 | −0.02 | 0.6624 |
Monthly hospital visits | ||||
Inpatient admissions | 0.03 | 0.0210 | 0.05 | 0.1002 |
Outpatient visits | 0.02 | 0.1516 | −0.03 | 0.1259 |
Table 4 | ||||
Monthly volumes and costs | ||||
Tertiary hospitals | ||||
Volumes | 0.01 | 0.0252 | −0.01 | 0.3314 |
Costs | 0.00 | 0.9177 | −0.01 | 0.3907 |
Secondary hospitals | ||||
Volumes | 0.01 | 0.3160 | 0.02 | 0.0525 |
Costs | 0.02 | 0.0293 | 0.00 | 0.7548 |
Monthly inpatient admission | ||||
Tertiary hospitals | 0.01 | 0.0000 | 0.00 | 0.9872 |
Secondary hospitals | 0.01 | 0.0000 | 0.00 | 0.0542 |
Appendix 4 | ||||
Monthly volumes and costs | ||||
1. Drugs for diabetes | ||||
Volumes | 0.01 | 0.2025 | −0.01 | 0.6668 |
Costs | 0.01 | 0.3300 | −0.01 | 0.4080 |
2. Antibacterials for systemic use | ||||
Volumes | 0.02 | 0.0386 | 0.04 | 0.0605 |
Costs | 0.02 | 0.1156 | 0.01 | 0.4314 |
3. Drugs for blood and blood–forming organs | ||||
Volumes | 0.01 | 0.2422 | 0.01 | 0.3635 |
Costs | 0.01 | 0.6513 | 0.05 | 0.0610 |
4. Drugs for cardiovascular system | ||||
Volumes | 0.02 | 0.0002 | −0.02 | 0.0773 |
Costs | 0.02 | 0.0171 | −0.02 | 0.1220 |
5. Drugs for acid-related disorders | ||||
Volumes | −0.02 | 0.0759 | 0.02 | 0.1250 |
Costs | −0.03 | 0.0327 | 0.00 | 0.9899 |
6. Drugs for respiratory system | ||||
Volumes | 0.00 | 0.6542 | 0.00 | 0.8181 |
Costs | 0.00 | 0.7517 | −0.01 | 0.4635 |
Notes: We performed GLM estimation with a log-link to testify common trends of outcome indicators in the intervention group and control group, and the basic regression model was set as follows:
|$E\left[ {{Y_{ijt}}} \right] = {\beta _1} + {\beta _2} \times {\text{Mont}}{{\text{h}}_t} + {\beta _3} \times {\text{Treatmen}}{{\text{t}}_i} + {\beta _4} \times {\text{Mont}}{{\text{h}}_t} \times {\text{Treatmen}}{{\text{t}}_i} + {_{ijt}}$|,
where Yijt refers to outcome indicators in our study like drug procurement volumes, costs and hospital visits, Treatmenti is a binary variable equalling 1 for Sanming and 0 for Longyan and Montht is a monthly time variable covering only the study period before policy intervention. β4 is the regression coefficient of interaction term, indicating common trends of corresponding indicators in Sanming and Longyan if statistically insignificant. We conducted the common trends test on drug procurement data (i.e. volumes and costs) in every data-independent regression model (i.e. the regression model in Tables 2–4 and Appendix 4, respectively). We did not conduct such tests for models in Appendix 6–8, as data in these models were actually a subset of data in models above.
Outcome indicators . | β2 . | P . | β4 . | P . |
---|---|---|---|---|
Table 3 | ||||
Monthly volumes and costs | ||||
Inpatient sector | ||||
Volumes | 0.01 | 0.0261 | 0.00 | 0.8659 |
Costs | 0.00 | 0.8515 | −0.01 | 0.5198 |
Outpatient sector | ||||
Volumes | 0.02 | 0.0290 | 0.00 | 0.9724 |
Costs | 0.02 | 0.0002 | −0.01 | 0.4301 |
Monthly hospital visits | ||||
Inpatient admissions | 0.01 | 0.1293 | 0.01 | 0.2471 |
Outpatient visits | 0.02 | 0.1121 | 0.01 | 0.6004 |
Table 2 | ||||
Monthly volumes and costs | ||||
Inpatient sector | ||||
Volumes | 0.00 | 0.9317 | 0.00 | 0.9396 |
Costs | 0.00 | 0.9123 | −0.03 | 0.5498 |
Outpatient sector | ||||
Volumes | 0.04 | 0.3208 | −0.01 | 0.7978 |
Costs | 0.05 | 0.0561 | −0.02 | 0.6624 |
Monthly hospital visits | ||||
Inpatient admissions | 0.03 | 0.0210 | 0.05 | 0.1002 |
Outpatient visits | 0.02 | 0.1516 | −0.03 | 0.1259 |
Table 4 | ||||
Monthly volumes and costs | ||||
Tertiary hospitals | ||||
Volumes | 0.01 | 0.0252 | −0.01 | 0.3314 |
Costs | 0.00 | 0.9177 | −0.01 | 0.3907 |
Secondary hospitals | ||||
Volumes | 0.01 | 0.3160 | 0.02 | 0.0525 |
Costs | 0.02 | 0.0293 | 0.00 | 0.7548 |
Monthly inpatient admission | ||||
Tertiary hospitals | 0.01 | 0.0000 | 0.00 | 0.9872 |
Secondary hospitals | 0.01 | 0.0000 | 0.00 | 0.0542 |
Appendix 4 | ||||
Monthly volumes and costs | ||||
1. Drugs for diabetes | ||||
Volumes | 0.01 | 0.2025 | −0.01 | 0.6668 |
Costs | 0.01 | 0.3300 | −0.01 | 0.4080 |
2. Antibacterials for systemic use | ||||
Volumes | 0.02 | 0.0386 | 0.04 | 0.0605 |
Costs | 0.02 | 0.1156 | 0.01 | 0.4314 |
3. Drugs for blood and blood–forming organs | ||||
Volumes | 0.01 | 0.2422 | 0.01 | 0.3635 |
Costs | 0.01 | 0.6513 | 0.05 | 0.0610 |
4. Drugs for cardiovascular system | ||||
Volumes | 0.02 | 0.0002 | −0.02 | 0.0773 |
Costs | 0.02 | 0.0171 | −0.02 | 0.1220 |
5. Drugs for acid-related disorders | ||||
Volumes | −0.02 | 0.0759 | 0.02 | 0.1250 |
Costs | −0.03 | 0.0327 | 0.00 | 0.9899 |
6. Drugs for respiratory system | ||||
Volumes | 0.00 | 0.6542 | 0.00 | 0.8181 |
Costs | 0.00 | 0.7517 | −0.01 | 0.4635 |
Outcome indicators . | β2 . | P . | β4 . | P . |
---|---|---|---|---|
Table 3 | ||||
Monthly volumes and costs | ||||
Inpatient sector | ||||
Volumes | 0.01 | 0.0261 | 0.00 | 0.8659 |
Costs | 0.00 | 0.8515 | −0.01 | 0.5198 |
Outpatient sector | ||||
Volumes | 0.02 | 0.0290 | 0.00 | 0.9724 |
Costs | 0.02 | 0.0002 | −0.01 | 0.4301 |
Monthly hospital visits | ||||
Inpatient admissions | 0.01 | 0.1293 | 0.01 | 0.2471 |
Outpatient visits | 0.02 | 0.1121 | 0.01 | 0.6004 |
Table 2 | ||||
Monthly volumes and costs | ||||
Inpatient sector | ||||
Volumes | 0.00 | 0.9317 | 0.00 | 0.9396 |
Costs | 0.00 | 0.9123 | −0.03 | 0.5498 |
Outpatient sector | ||||
Volumes | 0.04 | 0.3208 | −0.01 | 0.7978 |
Costs | 0.05 | 0.0561 | −0.02 | 0.6624 |
Monthly hospital visits | ||||
Inpatient admissions | 0.03 | 0.0210 | 0.05 | 0.1002 |
Outpatient visits | 0.02 | 0.1516 | −0.03 | 0.1259 |
Table 4 | ||||
Monthly volumes and costs | ||||
Tertiary hospitals | ||||
Volumes | 0.01 | 0.0252 | −0.01 | 0.3314 |
Costs | 0.00 | 0.9177 | −0.01 | 0.3907 |
Secondary hospitals | ||||
Volumes | 0.01 | 0.3160 | 0.02 | 0.0525 |
Costs | 0.02 | 0.0293 | 0.00 | 0.7548 |
Monthly inpatient admission | ||||
Tertiary hospitals | 0.01 | 0.0000 | 0.00 | 0.9872 |
Secondary hospitals | 0.01 | 0.0000 | 0.00 | 0.0542 |
Appendix 4 | ||||
Monthly volumes and costs | ||||
1. Drugs for diabetes | ||||
Volumes | 0.01 | 0.2025 | −0.01 | 0.6668 |
Costs | 0.01 | 0.3300 | −0.01 | 0.4080 |
2. Antibacterials for systemic use | ||||
Volumes | 0.02 | 0.0386 | 0.04 | 0.0605 |
Costs | 0.02 | 0.1156 | 0.01 | 0.4314 |
3. Drugs for blood and blood–forming organs | ||||
Volumes | 0.01 | 0.2422 | 0.01 | 0.3635 |
Costs | 0.01 | 0.6513 | 0.05 | 0.0610 |
4. Drugs for cardiovascular system | ||||
Volumes | 0.02 | 0.0002 | −0.02 | 0.0773 |
Costs | 0.02 | 0.0171 | −0.02 | 0.1220 |
5. Drugs for acid-related disorders | ||||
Volumes | −0.02 | 0.0759 | 0.02 | 0.1250 |
Costs | −0.03 | 0.0327 | 0.00 | 0.9899 |
6. Drugs for respiratory system | ||||
Volumes | 0.00 | 0.6542 | 0.00 | 0.8181 |
Costs | 0.00 | 0.7517 | −0.01 | 0.4635 |
Notes: We performed GLM estimation with a log-link to testify common trends of outcome indicators in the intervention group and control group, and the basic regression model was set as follows:
|$E\left[ {{Y_{ijt}}} \right] = {\beta _1} + {\beta _2} \times {\text{Mont}}{{\text{h}}_t} + {\beta _3} \times {\text{Treatmen}}{{\text{t}}_i} + {\beta _4} \times {\text{Mont}}{{\text{h}}_t} \times {\text{Treatmen}}{{\text{t}}_i} + {_{ijt}}$|,
where Yijt refers to outcome indicators in our study like drug procurement volumes, costs and hospital visits, Treatmenti is a binary variable equalling 1 for Sanming and 0 for Longyan and Montht is a monthly time variable covering only the study period before policy intervention. β4 is the regression coefficient of interaction term, indicating common trends of corresponding indicators in Sanming and Longyan if statistically insignificant. We conducted the common trends test on drug procurement data (i.e. volumes and costs) in every data-independent regression model (i.e. the regression model in Tables 2–4 and Appendix 4, respectively). We did not conduct such tests for models in Appendix 6–8, as data in these models were actually a subset of data in models above.
. | Impact of ZMDP . | Impact of RP . | ||
---|---|---|---|---|
. | Impact . | 95% CI . | Impact . | 95% CI . |
Monthly volumes and costs | ||||
Total volumes (1000 DDDs) | 18.1% | [−9.8%, 54.6%] | −72.5% | [−77.2%, −67.0%] |
Original drugs | 9.4% | [−21.3%, 52.2%] | −81.6% | [−85.3%, −77.0%] |
Generic drugs | 51.7% | [−9.2%, 108.8%] | −51.5% | [−61.1%, −39.6%] |
Total costs (1000 CNY) | −10.5% | [−29.3%, 13.3%] | −24.1% | [−35.6%, −10.6%] |
. | Impact of ZMDP . | Impact of RP . | ||
---|---|---|---|---|
. | Impact . | 95% CI . | Impact . | 95% CI . |
Monthly volumes and costs | ||||
Total volumes (1000 DDDs) | 18.1% | [−9.8%, 54.6%] | −72.5% | [−77.2%, −67.0%] |
Original drugs | 9.4% | [−21.3%, 52.2%] | −81.6% | [−85.3%, −77.0%] |
Generic drugs | 51.7% | [−9.2%, 108.8%] | −51.5% | [−61.1%, −39.6%] |
Total costs (1000 CNY) | −10.5% | [−29.3%, 13.3%] | −24.1% | [−35.6%, −10.6%] |
Notes: We reported the impact of the ZMDP and RP policy in Sanming, by regarding Sanming as the treatment group and Longyan as the control group, adopting the time–varying DID design. We reported the estimated coefficients (α2 and α3) and their 95% CIs from the following regression:
|$E\left[ {{Y_{ijt}}} \right] = {\alpha _1} + {\alpha _2} \times {\text{policy}}{1_{it}} + {\alpha _3} \times {\text{policy}}{2_{it}} + {\alpha _4} \times {\text{Treatmen}}{{\text{t}}_i} + {\alpha _5} \times {\text{Post}}{1_t} + {\alpha _6} \times {\text{Post}}{2_t} + {\varepsilon _{ijt}}$|.
The policy1it equals 1 if hospital i has ZMDP at time t, policy2it equals 1 if hospital i has RP at time t, Treatmenti equals 1 if hospital i is located in Sanming, Post1t and Post2t equal 1 if time t is after the implementation of ZMDP or RP. Otherwise, all these variables equal 0.
. | Impact of ZMDP . | Impact of RP . | ||
---|---|---|---|---|
. | Impact . | 95% CI . | Impact . | 95% CI . |
Monthly volumes and costs | ||||
Total volumes (1000 DDDs) | 18.1% | [−9.8%, 54.6%] | −72.5% | [−77.2%, −67.0%] |
Original drugs | 9.4% | [−21.3%, 52.2%] | −81.6% | [−85.3%, −77.0%] |
Generic drugs | 51.7% | [−9.2%, 108.8%] | −51.5% | [−61.1%, −39.6%] |
Total costs (1000 CNY) | −10.5% | [−29.3%, 13.3%] | −24.1% | [−35.6%, −10.6%] |
. | Impact of ZMDP . | Impact of RP . | ||
---|---|---|---|---|
. | Impact . | 95% CI . | Impact . | 95% CI . |
Monthly volumes and costs | ||||
Total volumes (1000 DDDs) | 18.1% | [−9.8%, 54.6%] | −72.5% | [−77.2%, −67.0%] |
Original drugs | 9.4% | [−21.3%, 52.2%] | −81.6% | [−85.3%, −77.0%] |
Generic drugs | 51.7% | [−9.2%, 108.8%] | −51.5% | [−61.1%, −39.6%] |
Total costs (1000 CNY) | −10.5% | [−29.3%, 13.3%] | −24.1% | [−35.6%, −10.6%] |
Notes: We reported the impact of the ZMDP and RP policy in Sanming, by regarding Sanming as the treatment group and Longyan as the control group, adopting the time–varying DID design. We reported the estimated coefficients (α2 and α3) and their 95% CIs from the following regression:
|$E\left[ {{Y_{ijt}}} \right] = {\alpha _1} + {\alpha _2} \times {\text{policy}}{1_{it}} + {\alpha _3} \times {\text{policy}}{2_{it}} + {\alpha _4} \times {\text{Treatmen}}{{\text{t}}_i} + {\alpha _5} \times {\text{Post}}{1_t} + {\alpha _6} \times {\text{Post}}{2_t} + {\varepsilon _{ijt}}$|.
The policy1it equals 1 if hospital i has ZMDP at time t, policy2it equals 1 if hospital i has RP at time t, Treatmenti equals 1 if hospital i is located in Sanming, Post1t and Post2t equal 1 if time t is after the implementation of ZMDP or RP. Otherwise, all these variables equal 0.
. | (1) Results from the sensitivity analyses . | (2) Results from the main analysesa . | Hausman tests for the difference between (1) and (2) . | |||
---|---|---|---|---|---|---|
. | Impact . | 95% CI . | Impact . | 95% CI . | Chi2 . | P . |
Monthly volumes and costs | ||||||
Inpatient sector | ||||||
Total volumes (1000 DDDs) | −26.3% | [−38.6%, −11.5%] | −25.9% | [−37.0%, −12.9%] | 0.01 | 0.9177 |
Original drugs | −56.1% | [−65.2%, −44.6%] | −56.8% | [−64.7%, −47.0%] | 0.10 | 0.7545 |
Generic drugs | 84.6% | [44.8%, 135.4%] | 98.6% | [60.5%, 145.8%] | 1.51 | 0.2191 |
Total costs (1000 CNY) | −45.2% | [−57.8%, −28.8%] | −47.7% | [−58.7%, −33.7%] | 0.37 | 0.5455 |
Outpatient sector | ||||||
Total volumes (1000 DDDs) | 5.0% | [−22.1%, 41.7%] | 4.9% | [−19.4%, 36.6%] | 0.00 | 0.9873 |
Original drugs | −10.2% | [−38.4%, 30.9%] | −12.1% | [−37.1%, 22.9%] | 0.07 | 0.7907 |
Generic drugs | 63.6% | [17.9%, 127.0%] | 72.0% | [28.6%, 130.0%] | 0.40 | 0.5273 |
Total costs (1000 CNY) | −11.8% | [−31.1%, 13.0%] | −15.5% | [−32.0%, 5.0%] | 0.66 | 0.4168 |
Monthly hospital visits | ||||||
Inpatient admissions (1000) | 2.5% | [−22.56%, 35.76%] | 9.4% | [−16.7%, 43.7%] | 1.65 | 0.1986 |
Outpatient visits (1000) | 12.4% | [−19.72%, 57.38%] | 18.8% | [−14.2%, 64.4%] | 1.13 | 0.2882 |
. | (1) Results from the sensitivity analyses . | (2) Results from the main analysesa . | Hausman tests for the difference between (1) and (2) . | |||
---|---|---|---|---|---|---|
. | Impact . | 95% CI . | Impact . | 95% CI . | Chi2 . | P . |
Monthly volumes and costs | ||||||
Inpatient sector | ||||||
Total volumes (1000 DDDs) | −26.3% | [−38.6%, −11.5%] | −25.9% | [−37.0%, −12.9%] | 0.01 | 0.9177 |
Original drugs | −56.1% | [−65.2%, −44.6%] | −56.8% | [−64.7%, −47.0%] | 0.10 | 0.7545 |
Generic drugs | 84.6% | [44.8%, 135.4%] | 98.6% | [60.5%, 145.8%] | 1.51 | 0.2191 |
Total costs (1000 CNY) | −45.2% | [−57.8%, −28.8%] | −47.7% | [−58.7%, −33.7%] | 0.37 | 0.5455 |
Outpatient sector | ||||||
Total volumes (1000 DDDs) | 5.0% | [−22.1%, 41.7%] | 4.9% | [−19.4%, 36.6%] | 0.00 | 0.9873 |
Original drugs | −10.2% | [−38.4%, 30.9%] | −12.1% | [−37.1%, 22.9%] | 0.07 | 0.7907 |
Generic drugs | 63.6% | [17.9%, 127.0%] | 72.0% | [28.6%, 130.0%] | 0.40 | 0.5273 |
Total costs (1000 CNY) | −11.8% | [−31.1%, 13.0%] | −15.5% | [−32.0%, 5.0%] | 0.66 | 0.4168 |
Monthly hospital visits | ||||||
Inpatient admissions (1000) | 2.5% | [−22.56%, 35.76%] | 9.4% | [−16.7%, 43.7%] | 1.65 | 0.1986 |
Outpatient visits (1000) | 12.4% | [−19.72%, 57.38%] | 18.8% | [−14.2%, 64.4%] | 1.13 | 0.2882 |
Notes: We performed GLMs to estimate the sensitivity model, by regressing each outcome indicator on the interaction term between a dummy variable equalling 1 after 2014.9 and 0 within 2013.1–2014.9, and a dummy variable indicating intervention. Then, the Hausman tests were conducted for the significance of differences between results from the sensitivity model and the main model.
Results from the main analyses are shown in Table 3.
. | (1) Results from the sensitivity analyses . | (2) Results from the main analysesa . | Hausman tests for the difference between (1) and (2) . | |||
---|---|---|---|---|---|---|
. | Impact . | 95% CI . | Impact . | 95% CI . | Chi2 . | P . |
Monthly volumes and costs | ||||||
Inpatient sector | ||||||
Total volumes (1000 DDDs) | −26.3% | [−38.6%, −11.5%] | −25.9% | [−37.0%, −12.9%] | 0.01 | 0.9177 |
Original drugs | −56.1% | [−65.2%, −44.6%] | −56.8% | [−64.7%, −47.0%] | 0.10 | 0.7545 |
Generic drugs | 84.6% | [44.8%, 135.4%] | 98.6% | [60.5%, 145.8%] | 1.51 | 0.2191 |
Total costs (1000 CNY) | −45.2% | [−57.8%, −28.8%] | −47.7% | [−58.7%, −33.7%] | 0.37 | 0.5455 |
Outpatient sector | ||||||
Total volumes (1000 DDDs) | 5.0% | [−22.1%, 41.7%] | 4.9% | [−19.4%, 36.6%] | 0.00 | 0.9873 |
Original drugs | −10.2% | [−38.4%, 30.9%] | −12.1% | [−37.1%, 22.9%] | 0.07 | 0.7907 |
Generic drugs | 63.6% | [17.9%, 127.0%] | 72.0% | [28.6%, 130.0%] | 0.40 | 0.5273 |
Total costs (1000 CNY) | −11.8% | [−31.1%, 13.0%] | −15.5% | [−32.0%, 5.0%] | 0.66 | 0.4168 |
Monthly hospital visits | ||||||
Inpatient admissions (1000) | 2.5% | [−22.56%, 35.76%] | 9.4% | [−16.7%, 43.7%] | 1.65 | 0.1986 |
Outpatient visits (1000) | 12.4% | [−19.72%, 57.38%] | 18.8% | [−14.2%, 64.4%] | 1.13 | 0.2882 |
. | (1) Results from the sensitivity analyses . | (2) Results from the main analysesa . | Hausman tests for the difference between (1) and (2) . | |||
---|---|---|---|---|---|---|
. | Impact . | 95% CI . | Impact . | 95% CI . | Chi2 . | P . |
Monthly volumes and costs | ||||||
Inpatient sector | ||||||
Total volumes (1000 DDDs) | −26.3% | [−38.6%, −11.5%] | −25.9% | [−37.0%, −12.9%] | 0.01 | 0.9177 |
Original drugs | −56.1% | [−65.2%, −44.6%] | −56.8% | [−64.7%, −47.0%] | 0.10 | 0.7545 |
Generic drugs | 84.6% | [44.8%, 135.4%] | 98.6% | [60.5%, 145.8%] | 1.51 | 0.2191 |
Total costs (1000 CNY) | −45.2% | [−57.8%, −28.8%] | −47.7% | [−58.7%, −33.7%] | 0.37 | 0.5455 |
Outpatient sector | ||||||
Total volumes (1000 DDDs) | 5.0% | [−22.1%, 41.7%] | 4.9% | [−19.4%, 36.6%] | 0.00 | 0.9873 |
Original drugs | −10.2% | [−38.4%, 30.9%] | −12.1% | [−37.1%, 22.9%] | 0.07 | 0.7907 |
Generic drugs | 63.6% | [17.9%, 127.0%] | 72.0% | [28.6%, 130.0%] | 0.40 | 0.5273 |
Total costs (1000 CNY) | −11.8% | [−31.1%, 13.0%] | −15.5% | [−32.0%, 5.0%] | 0.66 | 0.4168 |
Monthly hospital visits | ||||||
Inpatient admissions (1000) | 2.5% | [−22.56%, 35.76%] | 9.4% | [−16.7%, 43.7%] | 1.65 | 0.1986 |
Outpatient visits (1000) | 12.4% | [−19.72%, 57.38%] | 18.8% | [−14.2%, 64.4%] | 1.13 | 0.2882 |
Notes: We performed GLMs to estimate the sensitivity model, by regressing each outcome indicator on the interaction term between a dummy variable equalling 1 after 2014.9 and 0 within 2013.1–2014.9, and a dummy variable indicating intervention. Then, the Hausman tests were conducted for the significance of differences between results from the sensitivity model and the main model.
Results from the main analyses are shown in Table 3.
Appendix 8
Sensitivity analyses using 2012.1–2013.1 as the pre-policy period and 2015.6–2016.12 as the post-policy period.
. | (1) Results from the sensitivity analyses . | (2) Results from the main analysesa . | Hausman tests for the difference between (1) and (2) . | |||
---|---|---|---|---|---|---|
. | Impact . | 95% CI . | Impact . | 95% CI . | Chi2 . | P . |
Monthly volumes and costs | ||||||
Inpatient sector | ||||||
Total volumes (1000 DDDs) | −24.5% | [−40.5%, −4.2%] | −25.9% | [−37.0%, −12.9%] | 0.05 | 0.8317 |
Original drugs | −56.5% | [−67.4%, −41.8%] | −56.8% | [−64.7%, −47.0%] | 0.00 | 0.9533 |
Generic drugs | 131.2% | [67.5%, 219.2%] | 98.6% | [60.5%, 145.8%] | 1.56 | 0.2124 |
Total costs (1000 CNY) | −52.3% | [−66.8%, −31.5%] | −47.7% | [−58.7%, −33.7%] | 0.54 | 0.4604 |
Outpatient sector | ||||||
Total volumes (1000 DDDs) | 5.5% | [−28.4%, 55.5%] | 4.9% | [−19.4%, 36.6%] | 0.00 | 0.9687 |
Original drugs | −17.3% | [−49.3%, 34.9%] | −12.1% | [−37.1%, 22.9%] | 0.10 | 0.7464 |
Generic drugs | 102.1% | [29.3%, 216.0%] | 72.0% | [28.6%, 130.0%] | 0.92 | 0.3379 |
Total costs (1000 CNY) | −21.4% | [−42.6%, 7.7%] | −15.5% | [−32.0%, 5.0%] | 0.35 | 0.5531 |
Monthly hospital visits | ||||||
Inpatient admissions (1000) | 35.97% | [−17.1%, 123.1%] | 9.4% | [−16.7%, 43.7%] | 0.82 | 0.3639 |
Outpatient visits (1000) | 38.44% | [−23.3%, 149.8%] | 18.8% | [−14.2%, 64.4%] | 0.34 | 0.5574 |
. | (1) Results from the sensitivity analyses . | (2) Results from the main analysesa . | Hausman tests for the difference between (1) and (2) . | |||
---|---|---|---|---|---|---|
. | Impact . | 95% CI . | Impact . | 95% CI . | Chi2 . | P . |
Monthly volumes and costs | ||||||
Inpatient sector | ||||||
Total volumes (1000 DDDs) | −24.5% | [−40.5%, −4.2%] | −25.9% | [−37.0%, −12.9%] | 0.05 | 0.8317 |
Original drugs | −56.5% | [−67.4%, −41.8%] | −56.8% | [−64.7%, −47.0%] | 0.00 | 0.9533 |
Generic drugs | 131.2% | [67.5%, 219.2%] | 98.6% | [60.5%, 145.8%] | 1.56 | 0.2124 |
Total costs (1000 CNY) | −52.3% | [−66.8%, −31.5%] | −47.7% | [−58.7%, −33.7%] | 0.54 | 0.4604 |
Outpatient sector | ||||||
Total volumes (1000 DDDs) | 5.5% | [−28.4%, 55.5%] | 4.9% | [−19.4%, 36.6%] | 0.00 | 0.9687 |
Original drugs | −17.3% | [−49.3%, 34.9%] | −12.1% | [−37.1%, 22.9%] | 0.10 | 0.7464 |
Generic drugs | 102.1% | [29.3%, 216.0%] | 72.0% | [28.6%, 130.0%] | 0.92 | 0.3379 |
Total costs (1000 CNY) | −21.4% | [−42.6%, 7.7%] | −15.5% | [−32.0%, 5.0%] | 0.35 | 0.5531 |
Monthly hospital visits | ||||||
Inpatient admissions (1000) | 35.97% | [−17.1%, 123.1%] | 9.4% | [−16.7%, 43.7%] | 0.82 | 0.3639 |
Outpatient visits (1000) | 38.44% | [−23.3%, 149.8%] | 18.8% | [−14.2%, 64.4%] | 0.34 | 0.5574 |
Notes: We performed GLMs to estimate the sensitivity model, by regressing each outcome indicator on the interaction term between a dummy variable equalling 1 after 2015.6 and 0 before 2013.1, and a dummy variable indicating intervention. Then, the Hausman tests were conducted for the significance of differences between results from the sensitivity model and the main model.
Results from the main analyses are shown in Table 3.
. | (1) Results from the sensitivity analyses . | (2) Results from the main analysesa . | Hausman tests for the difference between (1) and (2) . | |||
---|---|---|---|---|---|---|
. | Impact . | 95% CI . | Impact . | 95% CI . | Chi2 . | P . |
Monthly volumes and costs | ||||||
Inpatient sector | ||||||
Total volumes (1000 DDDs) | −24.5% | [−40.5%, −4.2%] | −25.9% | [−37.0%, −12.9%] | 0.05 | 0.8317 |
Original drugs | −56.5% | [−67.4%, −41.8%] | −56.8% | [−64.7%, −47.0%] | 0.00 | 0.9533 |
Generic drugs | 131.2% | [67.5%, 219.2%] | 98.6% | [60.5%, 145.8%] | 1.56 | 0.2124 |
Total costs (1000 CNY) | −52.3% | [−66.8%, −31.5%] | −47.7% | [−58.7%, −33.7%] | 0.54 | 0.4604 |
Outpatient sector | ||||||
Total volumes (1000 DDDs) | 5.5% | [−28.4%, 55.5%] | 4.9% | [−19.4%, 36.6%] | 0.00 | 0.9687 |
Original drugs | −17.3% | [−49.3%, 34.9%] | −12.1% | [−37.1%, 22.9%] | 0.10 | 0.7464 |
Generic drugs | 102.1% | [29.3%, 216.0%] | 72.0% | [28.6%, 130.0%] | 0.92 | 0.3379 |
Total costs (1000 CNY) | −21.4% | [−42.6%, 7.7%] | −15.5% | [−32.0%, 5.0%] | 0.35 | 0.5531 |
Monthly hospital visits | ||||||
Inpatient admissions (1000) | 35.97% | [−17.1%, 123.1%] | 9.4% | [−16.7%, 43.7%] | 0.82 | 0.3639 |
Outpatient visits (1000) | 38.44% | [−23.3%, 149.8%] | 18.8% | [−14.2%, 64.4%] | 0.34 | 0.5574 |
. | (1) Results from the sensitivity analyses . | (2) Results from the main analysesa . | Hausman tests for the difference between (1) and (2) . | |||
---|---|---|---|---|---|---|
. | Impact . | 95% CI . | Impact . | 95% CI . | Chi2 . | P . |
Monthly volumes and costs | ||||||
Inpatient sector | ||||||
Total volumes (1000 DDDs) | −24.5% | [−40.5%, −4.2%] | −25.9% | [−37.0%, −12.9%] | 0.05 | 0.8317 |
Original drugs | −56.5% | [−67.4%, −41.8%] | −56.8% | [−64.7%, −47.0%] | 0.00 | 0.9533 |
Generic drugs | 131.2% | [67.5%, 219.2%] | 98.6% | [60.5%, 145.8%] | 1.56 | 0.2124 |
Total costs (1000 CNY) | −52.3% | [−66.8%, −31.5%] | −47.7% | [−58.7%, −33.7%] | 0.54 | 0.4604 |
Outpatient sector | ||||||
Total volumes (1000 DDDs) | 5.5% | [−28.4%, 55.5%] | 4.9% | [−19.4%, 36.6%] | 0.00 | 0.9687 |
Original drugs | −17.3% | [−49.3%, 34.9%] | −12.1% | [−37.1%, 22.9%] | 0.10 | 0.7464 |
Generic drugs | 102.1% | [29.3%, 216.0%] | 72.0% | [28.6%, 130.0%] | 0.92 | 0.3379 |
Total costs (1000 CNY) | −21.4% | [−42.6%, 7.7%] | −15.5% | [−32.0%, 5.0%] | 0.35 | 0.5531 |
Monthly hospital visits | ||||||
Inpatient admissions (1000) | 35.97% | [−17.1%, 123.1%] | 9.4% | [−16.7%, 43.7%] | 0.82 | 0.3639 |
Outpatient visits (1000) | 38.44% | [−23.3%, 149.8%] | 18.8% | [−14.2%, 64.4%] | 0.34 | 0.5574 |
Notes: We performed GLMs to estimate the sensitivity model, by regressing each outcome indicator on the interaction term between a dummy variable equalling 1 after 2015.6 and 0 before 2013.1, and a dummy variable indicating intervention. Then, the Hausman tests were conducted for the significance of differences between results from the sensitivity model and the main model.
Results from the main analyses are shown in Table 3.
. | (1) Results from the tertiary hospitals . | (2) Results from the secondary hospitals . | Hausman tests for the difference between (1) and (2) . | |||
---|---|---|---|---|---|---|
. | Impact . | 95% CI . | Impact . | 95% CI . | Chi2 . | P . |
Monthly volumes and costs | ||||||
Inpatient sector | ||||||
Total volumes (1000 DDDs) | −40.7% | [−54.2%, −23.4%] | 52.3% | [9.5%, 111.8%] | 18.14 | 0.0000 |
Original drugs | −58.3% | [−68.4%, −45.1%] | −42.0% | [−65.7%, −1.9%] | 1.03 | 0.3097 |
Generic drugs | 103.7% | [28.3%, 223.4%] | 163.6% | [74.7%, 297.9%] | 0.61 | 0.4347 |
Total costs (1000 CNY) | −52.6% | [−66.4%, −33.1%] | −56.5% | [−70.9%, −34.8%] | 0.09 | 0.7629 |
Outpatient sector | ||||||
Total volumes (1000 DDDs) | −1.0% | [−36.3%, 54.1%] | 49.2% | [−10.2%, 147.8%] | 1.34 | 0.2477 |
Original drugs | −10.7% | [−46.7%, 49.6%] | −26.5% | [−59.5%, 33.2%] | 0.22 | 0.6385 |
Generic drugs | 64.0% | [−9.0%, 195.9%] | 152.4% | [37.9%, 362.2%] | 0.90 | 0.3434 |
Total costs (1000 CNY) | −21.7% | [−44.0%, 9.4%] | −20.7% | [−48.2%, 21.4%] | 0.00 | 0.9644 |
Monthly hospital visits | ||||||
Inpatient admissions (1000) | −5.87% | [−19.2%, 9.6%] | −4.56% | [−13.9%, 5.8%] | 0.06 | 0.8052 |
Outpatient visits (1000) | −0.79% | [−15.1%, 16.0%] | 4.86% | [−7.7%, 19.1%] | 0.86 | 0.3541 |
. | (1) Results from the tertiary hospitals . | (2) Results from the secondary hospitals . | Hausman tests for the difference between (1) and (2) . | |||
---|---|---|---|---|---|---|
. | Impact . | 95% CI . | Impact . | 95% CI . | Chi2 . | P . |
Monthly volumes and costs | ||||||
Inpatient sector | ||||||
Total volumes (1000 DDDs) | −40.7% | [−54.2%, −23.4%] | 52.3% | [9.5%, 111.8%] | 18.14 | 0.0000 |
Original drugs | −58.3% | [−68.4%, −45.1%] | −42.0% | [−65.7%, −1.9%] | 1.03 | 0.3097 |
Generic drugs | 103.7% | [28.3%, 223.4%] | 163.6% | [74.7%, 297.9%] | 0.61 | 0.4347 |
Total costs (1000 CNY) | −52.6% | [−66.4%, −33.1%] | −56.5% | [−70.9%, −34.8%] | 0.09 | 0.7629 |
Outpatient sector | ||||||
Total volumes (1000 DDDs) | −1.0% | [−36.3%, 54.1%] | 49.2% | [−10.2%, 147.8%] | 1.34 | 0.2477 |
Original drugs | −10.7% | [−46.7%, 49.6%] | −26.5% | [−59.5%, 33.2%] | 0.22 | 0.6385 |
Generic drugs | 64.0% | [−9.0%, 195.9%] | 152.4% | [37.9%, 362.2%] | 0.90 | 0.3434 |
Total costs (1000 CNY) | −21.7% | [−44.0%, 9.4%] | −20.7% | [−48.2%, 21.4%] | 0.00 | 0.9644 |
Monthly hospital visits | ||||||
Inpatient admissions (1000) | −5.87% | [−19.2%, 9.6%] | −4.56% | [−13.9%, 5.8%] | 0.06 | 0.8052 |
Outpatient visits (1000) | −0.79% | [−15.1%, 16.0%] | 4.86% | [−7.7%, 19.1%] | 0.86 | 0.3541 |
Notes: We performed GLMs to estimate the sensitivity model, by regressing each outcome indicator on the interaction term between a dummy variable equalling 1 after 2015.6 and 0 before 2013.1, and a dummy variable indicating intervention. Then, the Hausman tests were conducted for the significance of differences between results from the sensitivity model and the main model.
a Results from the main analyses are shown in Table 3.
. | (1) Results from the tertiary hospitals . | (2) Results from the secondary hospitals . | Hausman tests for the difference between (1) and (2) . | |||
---|---|---|---|---|---|---|
. | Impact . | 95% CI . | Impact . | 95% CI . | Chi2 . | P . |
Monthly volumes and costs | ||||||
Inpatient sector | ||||||
Total volumes (1000 DDDs) | −40.7% | [−54.2%, −23.4%] | 52.3% | [9.5%, 111.8%] | 18.14 | 0.0000 |
Original drugs | −58.3% | [−68.4%, −45.1%] | −42.0% | [−65.7%, −1.9%] | 1.03 | 0.3097 |
Generic drugs | 103.7% | [28.3%, 223.4%] | 163.6% | [74.7%, 297.9%] | 0.61 | 0.4347 |
Total costs (1000 CNY) | −52.6% | [−66.4%, −33.1%] | −56.5% | [−70.9%, −34.8%] | 0.09 | 0.7629 |
Outpatient sector | ||||||
Total volumes (1000 DDDs) | −1.0% | [−36.3%, 54.1%] | 49.2% | [−10.2%, 147.8%] | 1.34 | 0.2477 |
Original drugs | −10.7% | [−46.7%, 49.6%] | −26.5% | [−59.5%, 33.2%] | 0.22 | 0.6385 |
Generic drugs | 64.0% | [−9.0%, 195.9%] | 152.4% | [37.9%, 362.2%] | 0.90 | 0.3434 |
Total costs (1000 CNY) | −21.7% | [−44.0%, 9.4%] | −20.7% | [−48.2%, 21.4%] | 0.00 | 0.9644 |
Monthly hospital visits | ||||||
Inpatient admissions (1000) | −5.87% | [−19.2%, 9.6%] | −4.56% | [−13.9%, 5.8%] | 0.06 | 0.8052 |
Outpatient visits (1000) | −0.79% | [−15.1%, 16.0%] | 4.86% | [−7.7%, 19.1%] | 0.86 | 0.3541 |
. | (1) Results from the tertiary hospitals . | (2) Results from the secondary hospitals . | Hausman tests for the difference between (1) and (2) . | |||
---|---|---|---|---|---|---|
. | Impact . | 95% CI . | Impact . | 95% CI . | Chi2 . | P . |
Monthly volumes and costs | ||||||
Inpatient sector | ||||||
Total volumes (1000 DDDs) | −40.7% | [−54.2%, −23.4%] | 52.3% | [9.5%, 111.8%] | 18.14 | 0.0000 |
Original drugs | −58.3% | [−68.4%, −45.1%] | −42.0% | [−65.7%, −1.9%] | 1.03 | 0.3097 |
Generic drugs | 103.7% | [28.3%, 223.4%] | 163.6% | [74.7%, 297.9%] | 0.61 | 0.4347 |
Total costs (1000 CNY) | −52.6% | [−66.4%, −33.1%] | −56.5% | [−70.9%, −34.8%] | 0.09 | 0.7629 |
Outpatient sector | ||||||
Total volumes (1000 DDDs) | −1.0% | [−36.3%, 54.1%] | 49.2% | [−10.2%, 147.8%] | 1.34 | 0.2477 |
Original drugs | −10.7% | [−46.7%, 49.6%] | −26.5% | [−59.5%, 33.2%] | 0.22 | 0.6385 |
Generic drugs | 64.0% | [−9.0%, 195.9%] | 152.4% | [37.9%, 362.2%] | 0.90 | 0.3434 |
Total costs (1000 CNY) | −21.7% | [−44.0%, 9.4%] | −20.7% | [−48.2%, 21.4%] | 0.00 | 0.9644 |
Monthly hospital visits | ||||||
Inpatient admissions (1000) | −5.87% | [−19.2%, 9.6%] | −4.56% | [−13.9%, 5.8%] | 0.06 | 0.8052 |
Outpatient visits (1000) | −0.79% | [−15.1%, 16.0%] | 4.86% | [−7.7%, 19.1%] | 0.86 | 0.3541 |
Notes: We performed GLMs to estimate the sensitivity model, by regressing each outcome indicator on the interaction term between a dummy variable equalling 1 after 2015.6 and 0 before 2013.1, and a dummy variable indicating intervention. Then, the Hausman tests were conducted for the significance of differences between results from the sensitivity model and the main model.
a Results from the main analyses are shown in Table 3.
Appendix 9
Placebo test by artificially changing the timing of RP’s implementation step-wised by each month, from six months before to six months after the real timing of event
Outcome indicators . | Coefficient . | [95% confidence interval] . | Family distribution . |
---|---|---|---|
Table 3 | |||
Monthly volumes and costs | |||
Inpatient sector | |||
Total volumes (1000 DDDs) | 1.6801 | [1.4317, 1.9285] | Gamma |
Original drugs | 2.1136 | [1.9351, 2.2921] | Gamma |
Generic drugs | 1.6803 | [1.6017, 1.7589] | Gamma |
Total costs (1000 CNY) | 1.9728 | [1.8083, 2.1372] | Gamma |
Outpatient sector | |||
Total volumes (1000 DDDs) | 1.8875 | [1.7577, 2.0173] | Gamma |
Original drugs | 1.9879 | [1.8516, 2.1242] | Gamma |
Generic drugs | 1.8546 | [1.7616, 1.9476] | Gamma |
Total costs (1000 CNY) | 1.565 | [1.3980, 1.7321] | Gamma |
Monthly hospital visits | |||
Inpatient admissions (1000) | 0.4086 | [−0.8173, 1.6345] | Gaussian |
Outpatient visits (1000) | 2.8278 | [2.3383, 3.3174] | Gamma |
Table 2 | |||
Monthly volumes and costs | |||
Inpatient sector | |||
Total volumes (1000 DDDs) | 1.6041 | [1.4364, 1.7717] | Gamma |
Original drugs | 1.4902 | [1.3509, 1.6296] | Poisson |
Generic drugs | 1.5274 | [1.3869, 1.6679] | Gamma |
Total costs (1000 CNY) | 2.4078 | [2.0566, 2.7589] | Gamma |
Outpatient sector | |||
Total volumes (1000 DDDs) | 1.7935 | [1.6355, 1.9515] | Gamma |
Original drugs | 2.2072 | [2.1101, 2.3043] | Gamma |
Generic drugs | 1.4688 | [1.3157, 1.6218] | Poisson |
Total costs (1000 CNY) | 1.1798 | [0.9846, 1.3750] | Poisson |
Monthly hospital visits | |||
Inpatient admissions (1000) | −4.5472 | [−18.5043, 9.4100] | Gamma |
Outpatient visits (1000) | 1.952 | [0.3596, 3.5443] | Gamma |
Table 4 | |||
Monthly volumes and costs | |||
Tertiary hospitals | |||
Total volumes (1000 DDDs) | 1.7848 | [1.5821, 1.9876] | Gamma |
Original drugs | 2.3403 | [2.1901, 2.4904] | Gamma |
Generic drugs | 1.6492 | [1.5853, 1.7132] | Gamma |
Total costs (1000 CNY) | 2.1202 | [1.9897, 2.2506] | Gamma |
Secondary hospitals | |||
Total volumes (1000 DDDs) | 1.6157 | [1.4194, 1.8120] | Gamma |
Original drugs | 1.3015 | [0.9119, 1.6911] | Poisson |
Generic drugs | 1.7168 | [1.6194, 1.8141] | Gamma |
Total costs (1000 CNY) | 1.5269 | [1.3214, 1.7324] | Gamma |
Monthly inpatient admission | |||
Tertiary hospitals (1000) | −4.8928 | [−16.4130, 6.6275] | Gamma |
Secondary hospitals (1000) | −0.3252 | [−5.1622, 4.5118] | Gaussian |
Appendix 4 | |||
Monthly volumes and costs | |||
1. Drugs for diabetes | |||
Total volumes (1000 DDDs) | 1.1418 | [0.8484, 1.4352] | Poisson |
Original drugs | 0.9481 | [0.6779, 1.2182] | Poisson |
Generic drugs | 1.8813 | [1.7517, 2.0110] | Gamma |
Total costs (1000 CNY) | −0.5801 | [−0.8954, −0.2648] | Poisson |
2. Antibacterials for systemic use | |||
Total volumes (1000 DDDs) | 2.1586 | [1.7802, 2.5370] | Gamma |
Original drugs | 0.8977 | [0.7912, 1.0042] | Poisson |
Generic drugs | NA | NA | NA |
Total costs (1000 CNY) | 1.8506 | [1.7859, 1.9153] | Gamma |
3. Drugs for blood and blood–forming organs | |||
Total volumes (1000 DDDs) | 1.9951 | [1.6760, 2.3141] | Gamma |
Original drugs | 2.2049 | [2.0006, 2.4093] | Gamma |
Generic drugs | 1.6835 | [1.6191, 1.7479] | Gamma |
Total costs (1000 CNY) | 1.841 | [1.7391, 1.9429] | Gamma |
4. Drugs for cardiovascular system | |||
Total volumes (1000 DDDs) | 1.7219 | [1.4561, 1.9878] | Gamma |
Original drugs | 2.1481 | [2.0021, 2.2941] | Gamma |
Generic drugs | 1.8361 | [1.7320, 1.9401] | Gamma |
Total costs (1000 CNY) | 2.329 | [1.9596, 2.6984] | Gamma |
5. Drugs for acid–related disorders | |||
Total volumes (1000 DDDs) | 1.55 | [1.4774, 1.6225] | Gamma |
Original drugs | 1.4861 | [1.4208, 1.5515] | Poisson |
Generic drugs | 1.6814 | [1.6206, 1.7423] | Gamma |
Total costs (1000 CNY) | 2.0509 | [1.9857, 2.1160] | Gamma |
6. Drugs for respiratory system | |||
Total volumes (1000 DDDs) | 1.796 | [1.6742, 1.9178] | Gamma |
Original drugs | 2.2033 | [2.1053, 2.3014] | Gamma |
Generic drugs | 1.7448 | [1.7132, 1.7764] | Gamma |
Total costs (1000 CNY) | 2.1855 | [2.0847, 2.2863] | Gamma |
Appendix 7 | |||
Monthly volumes and costs | |||
Inpatient sector | |||
Total volumes (1000 DDDs) | 1.6133 | [1.3573, 1.8693] | Gamma |
Original drugs | 2.1097 | [1.9355, 2.2839] | Gamma |
Generic drugs | 1.6669 | [1.5879, 1.7460] | Gamma |
Total costs (1000 CNY) | 1.9991 | [1.8255, 2.1727] | Gamma |
Outpatient sector | |||
Total volumes (1000 DDDs) | 1.8793 | [1.7202, 2.0385] | Gamma |
Original drugs | 2.083 | [1.9091, 2.2568] | Gamma |
Generic drugs | 1.8484 | [1.7464, 1.9503] | Gamma |
Total costs (1000 CNY) | 1.5257 | [1.3228, 1.7285] | Gamma |
Monthly hospital visits | |||
Inpatient admissions (1000) | 0.448 | [−2.9339, 3.8298] | Gaussian |
Outpatient visits (1000) | 3.1912 | [2.3947, 3.9876] | Gamma |
Outcome indicators . | Coefficient . | [95% confidence interval] . | Family distribution . |
---|---|---|---|
Table 3 | |||
Monthly volumes and costs | |||
Inpatient sector | |||
Total volumes (1000 DDDs) | 1.6801 | [1.4317, 1.9285] | Gamma |
Original drugs | 2.1136 | [1.9351, 2.2921] | Gamma |
Generic drugs | 1.6803 | [1.6017, 1.7589] | Gamma |
Total costs (1000 CNY) | 1.9728 | [1.8083, 2.1372] | Gamma |
Outpatient sector | |||
Total volumes (1000 DDDs) | 1.8875 | [1.7577, 2.0173] | Gamma |
Original drugs | 1.9879 | [1.8516, 2.1242] | Gamma |
Generic drugs | 1.8546 | [1.7616, 1.9476] | Gamma |
Total costs (1000 CNY) | 1.565 | [1.3980, 1.7321] | Gamma |
Monthly hospital visits | |||
Inpatient admissions (1000) | 0.4086 | [−0.8173, 1.6345] | Gaussian |
Outpatient visits (1000) | 2.8278 | [2.3383, 3.3174] | Gamma |
Table 2 | |||
Monthly volumes and costs | |||
Inpatient sector | |||
Total volumes (1000 DDDs) | 1.6041 | [1.4364, 1.7717] | Gamma |
Original drugs | 1.4902 | [1.3509, 1.6296] | Poisson |
Generic drugs | 1.5274 | [1.3869, 1.6679] | Gamma |
Total costs (1000 CNY) | 2.4078 | [2.0566, 2.7589] | Gamma |
Outpatient sector | |||
Total volumes (1000 DDDs) | 1.7935 | [1.6355, 1.9515] | Gamma |
Original drugs | 2.2072 | [2.1101, 2.3043] | Gamma |
Generic drugs | 1.4688 | [1.3157, 1.6218] | Poisson |
Total costs (1000 CNY) | 1.1798 | [0.9846, 1.3750] | Poisson |
Monthly hospital visits | |||
Inpatient admissions (1000) | −4.5472 | [−18.5043, 9.4100] | Gamma |
Outpatient visits (1000) | 1.952 | [0.3596, 3.5443] | Gamma |
Table 4 | |||
Monthly volumes and costs | |||
Tertiary hospitals | |||
Total volumes (1000 DDDs) | 1.7848 | [1.5821, 1.9876] | Gamma |
Original drugs | 2.3403 | [2.1901, 2.4904] | Gamma |
Generic drugs | 1.6492 | [1.5853, 1.7132] | Gamma |
Total costs (1000 CNY) | 2.1202 | [1.9897, 2.2506] | Gamma |
Secondary hospitals | |||
Total volumes (1000 DDDs) | 1.6157 | [1.4194, 1.8120] | Gamma |
Original drugs | 1.3015 | [0.9119, 1.6911] | Poisson |
Generic drugs | 1.7168 | [1.6194, 1.8141] | Gamma |
Total costs (1000 CNY) | 1.5269 | [1.3214, 1.7324] | Gamma |
Monthly inpatient admission | |||
Tertiary hospitals (1000) | −4.8928 | [−16.4130, 6.6275] | Gamma |
Secondary hospitals (1000) | −0.3252 | [−5.1622, 4.5118] | Gaussian |
Appendix 4 | |||
Monthly volumes and costs | |||
1. Drugs for diabetes | |||
Total volumes (1000 DDDs) | 1.1418 | [0.8484, 1.4352] | Poisson |
Original drugs | 0.9481 | [0.6779, 1.2182] | Poisson |
Generic drugs | 1.8813 | [1.7517, 2.0110] | Gamma |
Total costs (1000 CNY) | −0.5801 | [−0.8954, −0.2648] | Poisson |
2. Antibacterials for systemic use | |||
Total volumes (1000 DDDs) | 2.1586 | [1.7802, 2.5370] | Gamma |
Original drugs | 0.8977 | [0.7912, 1.0042] | Poisson |
Generic drugs | NA | NA | NA |
Total costs (1000 CNY) | 1.8506 | [1.7859, 1.9153] | Gamma |
3. Drugs for blood and blood–forming organs | |||
Total volumes (1000 DDDs) | 1.9951 | [1.6760, 2.3141] | Gamma |
Original drugs | 2.2049 | [2.0006, 2.4093] | Gamma |
Generic drugs | 1.6835 | [1.6191, 1.7479] | Gamma |
Total costs (1000 CNY) | 1.841 | [1.7391, 1.9429] | Gamma |
4. Drugs for cardiovascular system | |||
Total volumes (1000 DDDs) | 1.7219 | [1.4561, 1.9878] | Gamma |
Original drugs | 2.1481 | [2.0021, 2.2941] | Gamma |
Generic drugs | 1.8361 | [1.7320, 1.9401] | Gamma |
Total costs (1000 CNY) | 2.329 | [1.9596, 2.6984] | Gamma |
5. Drugs for acid–related disorders | |||
Total volumes (1000 DDDs) | 1.55 | [1.4774, 1.6225] | Gamma |
Original drugs | 1.4861 | [1.4208, 1.5515] | Poisson |
Generic drugs | 1.6814 | [1.6206, 1.7423] | Gamma |
Total costs (1000 CNY) | 2.0509 | [1.9857, 2.1160] | Gamma |
6. Drugs for respiratory system | |||
Total volumes (1000 DDDs) | 1.796 | [1.6742, 1.9178] | Gamma |
Original drugs | 2.2033 | [2.1053, 2.3014] | Gamma |
Generic drugs | 1.7448 | [1.7132, 1.7764] | Gamma |
Total costs (1000 CNY) | 2.1855 | [2.0847, 2.2863] | Gamma |
Appendix 7 | |||
Monthly volumes and costs | |||
Inpatient sector | |||
Total volumes (1000 DDDs) | 1.6133 | [1.3573, 1.8693] | Gamma |
Original drugs | 2.1097 | [1.9355, 2.2839] | Gamma |
Generic drugs | 1.6669 | [1.5879, 1.7460] | Gamma |
Total costs (1000 CNY) | 1.9991 | [1.8255, 2.1727] | Gamma |
Outpatient sector | |||
Total volumes (1000 DDDs) | 1.8793 | [1.7202, 2.0385] | Gamma |
Original drugs | 2.083 | [1.9091, 2.2568] | Gamma |
Generic drugs | 1.8484 | [1.7464, 1.9503] | Gamma |
Total costs (1000 CNY) | 1.5257 | [1.3228, 1.7285] | Gamma |
Monthly hospital visits | |||
Inpatient admissions (1000) | 0.448 | [−2.9339, 3.8298] | Gaussian |
Outpatient visits (1000) | 3.1912 | [2.3947, 3.9876] | Gamma |
Notes: Coefficients derived from the modified Park tests were used to determine family distribution in the generalized linear regression models: 0 indicates Gaussian distribution, 1 indicates Poisson distribution and 2 indicates Gamma distribution.
Outcome indicators . | Coefficient . | [95% confidence interval] . | Family distribution . |
---|---|---|---|
Table 3 | |||
Monthly volumes and costs | |||
Inpatient sector | |||
Total volumes (1000 DDDs) | 1.6801 | [1.4317, 1.9285] | Gamma |
Original drugs | 2.1136 | [1.9351, 2.2921] | Gamma |
Generic drugs | 1.6803 | [1.6017, 1.7589] | Gamma |
Total costs (1000 CNY) | 1.9728 | [1.8083, 2.1372] | Gamma |
Outpatient sector | |||
Total volumes (1000 DDDs) | 1.8875 | [1.7577, 2.0173] | Gamma |
Original drugs | 1.9879 | [1.8516, 2.1242] | Gamma |
Generic drugs | 1.8546 | [1.7616, 1.9476] | Gamma |
Total costs (1000 CNY) | 1.565 | [1.3980, 1.7321] | Gamma |
Monthly hospital visits | |||
Inpatient admissions (1000) | 0.4086 | [−0.8173, 1.6345] | Gaussian |
Outpatient visits (1000) | 2.8278 | [2.3383, 3.3174] | Gamma |
Table 2 | |||
Monthly volumes and costs | |||
Inpatient sector | |||
Total volumes (1000 DDDs) | 1.6041 | [1.4364, 1.7717] | Gamma |
Original drugs | 1.4902 | [1.3509, 1.6296] | Poisson |
Generic drugs | 1.5274 | [1.3869, 1.6679] | Gamma |
Total costs (1000 CNY) | 2.4078 | [2.0566, 2.7589] | Gamma |
Outpatient sector | |||
Total volumes (1000 DDDs) | 1.7935 | [1.6355, 1.9515] | Gamma |
Original drugs | 2.2072 | [2.1101, 2.3043] | Gamma |
Generic drugs | 1.4688 | [1.3157, 1.6218] | Poisson |
Total costs (1000 CNY) | 1.1798 | [0.9846, 1.3750] | Poisson |
Monthly hospital visits | |||
Inpatient admissions (1000) | −4.5472 | [−18.5043, 9.4100] | Gamma |
Outpatient visits (1000) | 1.952 | [0.3596, 3.5443] | Gamma |
Table 4 | |||
Monthly volumes and costs | |||
Tertiary hospitals | |||
Total volumes (1000 DDDs) | 1.7848 | [1.5821, 1.9876] | Gamma |
Original drugs | 2.3403 | [2.1901, 2.4904] | Gamma |
Generic drugs | 1.6492 | [1.5853, 1.7132] | Gamma |
Total costs (1000 CNY) | 2.1202 | [1.9897, 2.2506] | Gamma |
Secondary hospitals | |||
Total volumes (1000 DDDs) | 1.6157 | [1.4194, 1.8120] | Gamma |
Original drugs | 1.3015 | [0.9119, 1.6911] | Poisson |
Generic drugs | 1.7168 | [1.6194, 1.8141] | Gamma |
Total costs (1000 CNY) | 1.5269 | [1.3214, 1.7324] | Gamma |
Monthly inpatient admission | |||
Tertiary hospitals (1000) | −4.8928 | [−16.4130, 6.6275] | Gamma |
Secondary hospitals (1000) | −0.3252 | [−5.1622, 4.5118] | Gaussian |
Appendix 4 | |||
Monthly volumes and costs | |||
1. Drugs for diabetes | |||
Total volumes (1000 DDDs) | 1.1418 | [0.8484, 1.4352] | Poisson |
Original drugs | 0.9481 | [0.6779, 1.2182] | Poisson |
Generic drugs | 1.8813 | [1.7517, 2.0110] | Gamma |
Total costs (1000 CNY) | −0.5801 | [−0.8954, −0.2648] | Poisson |
2. Antibacterials for systemic use | |||
Total volumes (1000 DDDs) | 2.1586 | [1.7802, 2.5370] | Gamma |
Original drugs | 0.8977 | [0.7912, 1.0042] | Poisson |
Generic drugs | NA | NA | NA |
Total costs (1000 CNY) | 1.8506 | [1.7859, 1.9153] | Gamma |
3. Drugs for blood and blood–forming organs | |||
Total volumes (1000 DDDs) | 1.9951 | [1.6760, 2.3141] | Gamma |
Original drugs | 2.2049 | [2.0006, 2.4093] | Gamma |
Generic drugs | 1.6835 | [1.6191, 1.7479] | Gamma |
Total costs (1000 CNY) | 1.841 | [1.7391, 1.9429] | Gamma |
4. Drugs for cardiovascular system | |||
Total volumes (1000 DDDs) | 1.7219 | [1.4561, 1.9878] | Gamma |
Original drugs | 2.1481 | [2.0021, 2.2941] | Gamma |
Generic drugs | 1.8361 | [1.7320, 1.9401] | Gamma |
Total costs (1000 CNY) | 2.329 | [1.9596, 2.6984] | Gamma |
5. Drugs for acid–related disorders | |||
Total volumes (1000 DDDs) | 1.55 | [1.4774, 1.6225] | Gamma |
Original drugs | 1.4861 | [1.4208, 1.5515] | Poisson |
Generic drugs | 1.6814 | [1.6206, 1.7423] | Gamma |
Total costs (1000 CNY) | 2.0509 | [1.9857, 2.1160] | Gamma |
6. Drugs for respiratory system | |||
Total volumes (1000 DDDs) | 1.796 | [1.6742, 1.9178] | Gamma |
Original drugs | 2.2033 | [2.1053, 2.3014] | Gamma |
Generic drugs | 1.7448 | [1.7132, 1.7764] | Gamma |
Total costs (1000 CNY) | 2.1855 | [2.0847, 2.2863] | Gamma |
Appendix 7 | |||
Monthly volumes and costs | |||
Inpatient sector | |||
Total volumes (1000 DDDs) | 1.6133 | [1.3573, 1.8693] | Gamma |
Original drugs | 2.1097 | [1.9355, 2.2839] | Gamma |
Generic drugs | 1.6669 | [1.5879, 1.7460] | Gamma |
Total costs (1000 CNY) | 1.9991 | [1.8255, 2.1727] | Gamma |
Outpatient sector | |||
Total volumes (1000 DDDs) | 1.8793 | [1.7202, 2.0385] | Gamma |
Original drugs | 2.083 | [1.9091, 2.2568] | Gamma |
Generic drugs | 1.8484 | [1.7464, 1.9503] | Gamma |
Total costs (1000 CNY) | 1.5257 | [1.3228, 1.7285] | Gamma |
Monthly hospital visits | |||
Inpatient admissions (1000) | 0.448 | [−2.9339, 3.8298] | Gaussian |
Outpatient visits (1000) | 3.1912 | [2.3947, 3.9876] | Gamma |
Outcome indicators . | Coefficient . | [95% confidence interval] . | Family distribution . |
---|---|---|---|
Table 3 | |||
Monthly volumes and costs | |||
Inpatient sector | |||
Total volumes (1000 DDDs) | 1.6801 | [1.4317, 1.9285] | Gamma |
Original drugs | 2.1136 | [1.9351, 2.2921] | Gamma |
Generic drugs | 1.6803 | [1.6017, 1.7589] | Gamma |
Total costs (1000 CNY) | 1.9728 | [1.8083, 2.1372] | Gamma |
Outpatient sector | |||
Total volumes (1000 DDDs) | 1.8875 | [1.7577, 2.0173] | Gamma |
Original drugs | 1.9879 | [1.8516, 2.1242] | Gamma |
Generic drugs | 1.8546 | [1.7616, 1.9476] | Gamma |
Total costs (1000 CNY) | 1.565 | [1.3980, 1.7321] | Gamma |
Monthly hospital visits | |||
Inpatient admissions (1000) | 0.4086 | [−0.8173, 1.6345] | Gaussian |
Outpatient visits (1000) | 2.8278 | [2.3383, 3.3174] | Gamma |
Table 2 | |||
Monthly volumes and costs | |||
Inpatient sector | |||
Total volumes (1000 DDDs) | 1.6041 | [1.4364, 1.7717] | Gamma |
Original drugs | 1.4902 | [1.3509, 1.6296] | Poisson |
Generic drugs | 1.5274 | [1.3869, 1.6679] | Gamma |
Total costs (1000 CNY) | 2.4078 | [2.0566, 2.7589] | Gamma |
Outpatient sector | |||
Total volumes (1000 DDDs) | 1.7935 | [1.6355, 1.9515] | Gamma |
Original drugs | 2.2072 | [2.1101, 2.3043] | Gamma |
Generic drugs | 1.4688 | [1.3157, 1.6218] | Poisson |
Total costs (1000 CNY) | 1.1798 | [0.9846, 1.3750] | Poisson |
Monthly hospital visits | |||
Inpatient admissions (1000) | −4.5472 | [−18.5043, 9.4100] | Gamma |
Outpatient visits (1000) | 1.952 | [0.3596, 3.5443] | Gamma |
Table 4 | |||
Monthly volumes and costs | |||
Tertiary hospitals | |||
Total volumes (1000 DDDs) | 1.7848 | [1.5821, 1.9876] | Gamma |
Original drugs | 2.3403 | [2.1901, 2.4904] | Gamma |
Generic drugs | 1.6492 | [1.5853, 1.7132] | Gamma |
Total costs (1000 CNY) | 2.1202 | [1.9897, 2.2506] | Gamma |
Secondary hospitals | |||
Total volumes (1000 DDDs) | 1.6157 | [1.4194, 1.8120] | Gamma |
Original drugs | 1.3015 | [0.9119, 1.6911] | Poisson |
Generic drugs | 1.7168 | [1.6194, 1.8141] | Gamma |
Total costs (1000 CNY) | 1.5269 | [1.3214, 1.7324] | Gamma |
Monthly inpatient admission | |||
Tertiary hospitals (1000) | −4.8928 | [−16.4130, 6.6275] | Gamma |
Secondary hospitals (1000) | −0.3252 | [−5.1622, 4.5118] | Gaussian |
Appendix 4 | |||
Monthly volumes and costs | |||
1. Drugs for diabetes | |||
Total volumes (1000 DDDs) | 1.1418 | [0.8484, 1.4352] | Poisson |
Original drugs | 0.9481 | [0.6779, 1.2182] | Poisson |
Generic drugs | 1.8813 | [1.7517, 2.0110] | Gamma |
Total costs (1000 CNY) | −0.5801 | [−0.8954, −0.2648] | Poisson |
2. Antibacterials for systemic use | |||
Total volumes (1000 DDDs) | 2.1586 | [1.7802, 2.5370] | Gamma |
Original drugs | 0.8977 | [0.7912, 1.0042] | Poisson |
Generic drugs | NA | NA | NA |
Total costs (1000 CNY) | 1.8506 | [1.7859, 1.9153] | Gamma |
3. Drugs for blood and blood–forming organs | |||
Total volumes (1000 DDDs) | 1.9951 | [1.6760, 2.3141] | Gamma |
Original drugs | 2.2049 | [2.0006, 2.4093] | Gamma |
Generic drugs | 1.6835 | [1.6191, 1.7479] | Gamma |
Total costs (1000 CNY) | 1.841 | [1.7391, 1.9429] | Gamma |
4. Drugs for cardiovascular system | |||
Total volumes (1000 DDDs) | 1.7219 | [1.4561, 1.9878] | Gamma |
Original drugs | 2.1481 | [2.0021, 2.2941] | Gamma |
Generic drugs | 1.8361 | [1.7320, 1.9401] | Gamma |
Total costs (1000 CNY) | 2.329 | [1.9596, 2.6984] | Gamma |
5. Drugs for acid–related disorders | |||
Total volumes (1000 DDDs) | 1.55 | [1.4774, 1.6225] | Gamma |
Original drugs | 1.4861 | [1.4208, 1.5515] | Poisson |
Generic drugs | 1.6814 | [1.6206, 1.7423] | Gamma |
Total costs (1000 CNY) | 2.0509 | [1.9857, 2.1160] | Gamma |
6. Drugs for respiratory system | |||
Total volumes (1000 DDDs) | 1.796 | [1.6742, 1.9178] | Gamma |
Original drugs | 2.2033 | [2.1053, 2.3014] | Gamma |
Generic drugs | 1.7448 | [1.7132, 1.7764] | Gamma |
Total costs (1000 CNY) | 2.1855 | [2.0847, 2.2863] | Gamma |
Appendix 7 | |||
Monthly volumes and costs | |||
Inpatient sector | |||
Total volumes (1000 DDDs) | 1.6133 | [1.3573, 1.8693] | Gamma |
Original drugs | 2.1097 | [1.9355, 2.2839] | Gamma |
Generic drugs | 1.6669 | [1.5879, 1.7460] | Gamma |
Total costs (1000 CNY) | 1.9991 | [1.8255, 2.1727] | Gamma |
Outpatient sector | |||
Total volumes (1000 DDDs) | 1.8793 | [1.7202, 2.0385] | Gamma |
Original drugs | 2.083 | [1.9091, 2.2568] | Gamma |
Generic drugs | 1.8484 | [1.7464, 1.9503] | Gamma |
Total costs (1000 CNY) | 1.5257 | [1.3228, 1.7285] | Gamma |
Monthly hospital visits | |||
Inpatient admissions (1000) | 0.448 | [−2.9339, 3.8298] | Gaussian |
Outpatient visits (1000) | 3.1912 | [2.3947, 3.9876] | Gamma |
Notes: Coefficients derived from the modified Park tests were used to determine family distribution in the generalized linear regression models: 0 indicates Gaussian distribution, 1 indicates Poisson distribution and 2 indicates Gamma distribution.
. | DID estimates by OLS . | DID estimates by GLM . | Differences of DID estimates between OLS versus GLS . | |||
---|---|---|---|---|---|---|
. | Impacts of RP . | 95% CI . | Impacts of RP . | 95% CI . | Chi2 . | P . |
Monthly volumes and costs | ||||||
Inpatient sector | ||||||
Total volumes (1000 DDDs) | −9.4 | [−15.0, −3.8] | −9.4 | [−15.0, −3.8] | 0.0 | 0.9845 |
Original drugs | −19.6 | [−24.6, −14.7] | −19.6 | [−24.6, −14.7] | 0.0 | 0.9985 |
Generic drugs | 10.3 | [8.1, 12.5] | 10.3 | [8.1, 12.5] | 0.0 | 0.9941 |
Total costs (1000 CNY) | −420.0 | [−586.6, −253.4] | −420.0 | [−586.6, −254.8] | 0.0 | 0.9986 |
Outpatient sector | ||||||
Total volumes (1000 DDDs) | 12.0 | [−19.9, 43.8] | 12.0 | [−19.9, 43.8] | 0.0 | 0.9854 |
Original drugs | −6.5 | [−35.8, 22.9] | −6.5 | [−35.8, 22.9] | 0.0 | 0.9564 |
Generic drugs | 18.4 | [8.8, 28.1] | 18.4 | [8.8, 28.0] | 0.0 | 0.9996 |
Total costs (1000 CNY) | −70.1 | [−186.2, 45.7] | −70.1 | [−186.2, 45.6] | 0.0 | 0.9976 |
Monthly hospital visits | ||||||
Inpatient admissions (1000) | 0.3 | [−0.6, 1.2] | 0.3 | [−0.6, 1.2] | 0.0 | 0.9931 |
Outpatient visits (1000) | 7.0 | [−16.0, 30.0] | 7.0 | [−15.7, 29.6] | 0.0 | 0.9996 |
. | DID estimates by OLS . | DID estimates by GLM . | Differences of DID estimates between OLS versus GLS . | |||
---|---|---|---|---|---|---|
. | Impacts of RP . | 95% CI . | Impacts of RP . | 95% CI . | Chi2 . | P . |
Monthly volumes and costs | ||||||
Inpatient sector | ||||||
Total volumes (1000 DDDs) | −9.4 | [−15.0, −3.8] | −9.4 | [−15.0, −3.8] | 0.0 | 0.9845 |
Original drugs | −19.6 | [−24.6, −14.7] | −19.6 | [−24.6, −14.7] | 0.0 | 0.9985 |
Generic drugs | 10.3 | [8.1, 12.5] | 10.3 | [8.1, 12.5] | 0.0 | 0.9941 |
Total costs (1000 CNY) | −420.0 | [−586.6, −253.4] | −420.0 | [−586.6, −254.8] | 0.0 | 0.9986 |
Outpatient sector | ||||||
Total volumes (1000 DDDs) | 12.0 | [−19.9, 43.8] | 12.0 | [−19.9, 43.8] | 0.0 | 0.9854 |
Original drugs | −6.5 | [−35.8, 22.9] | −6.5 | [−35.8, 22.9] | 0.0 | 0.9564 |
Generic drugs | 18.4 | [8.8, 28.1] | 18.4 | [8.8, 28.0] | 0.0 | 0.9996 |
Total costs (1000 CNY) | −70.1 | [−186.2, 45.7] | −70.1 | [−186.2, 45.6] | 0.0 | 0.9976 |
Monthly hospital visits | ||||||
Inpatient admissions (1000) | 0.3 | [−0.6, 1.2] | 0.3 | [−0.6, 1.2] | 0.0 | 0.9931 |
Outpatient visits (1000) | 7.0 | [−16.0, 30.0] | 7.0 | [−15.7, 29.6] | 0.0 | 0.9996 |
. | DID estimates by OLS . | DID estimates by GLM . | Differences of DID estimates between OLS versus GLS . | |||
---|---|---|---|---|---|---|
. | Impacts of RP . | 95% CI . | Impacts of RP . | 95% CI . | Chi2 . | P . |
Monthly volumes and costs | ||||||
Inpatient sector | ||||||
Total volumes (1000 DDDs) | −9.4 | [−15.0, −3.8] | −9.4 | [−15.0, −3.8] | 0.0 | 0.9845 |
Original drugs | −19.6 | [−24.6, −14.7] | −19.6 | [−24.6, −14.7] | 0.0 | 0.9985 |
Generic drugs | 10.3 | [8.1, 12.5] | 10.3 | [8.1, 12.5] | 0.0 | 0.9941 |
Total costs (1000 CNY) | −420.0 | [−586.6, −253.4] | −420.0 | [−586.6, −254.8] | 0.0 | 0.9986 |
Outpatient sector | ||||||
Total volumes (1000 DDDs) | 12.0 | [−19.9, 43.8] | 12.0 | [−19.9, 43.8] | 0.0 | 0.9854 |
Original drugs | −6.5 | [−35.8, 22.9] | −6.5 | [−35.8, 22.9] | 0.0 | 0.9564 |
Generic drugs | 18.4 | [8.8, 28.1] | 18.4 | [8.8, 28.0] | 0.0 | 0.9996 |
Total costs (1000 CNY) | −70.1 | [−186.2, 45.7] | −70.1 | [−186.2, 45.6] | 0.0 | 0.9976 |
Monthly hospital visits | ||||||
Inpatient admissions (1000) | 0.3 | [−0.6, 1.2] | 0.3 | [−0.6, 1.2] | 0.0 | 0.9931 |
Outpatient visits (1000) | 7.0 | [−16.0, 30.0] | 7.0 | [−15.7, 29.6] | 0.0 | 0.9996 |
. | DID estimates by OLS . | DID estimates by GLM . | Differences of DID estimates between OLS versus GLS . | |||
---|---|---|---|---|---|---|
. | Impacts of RP . | 95% CI . | Impacts of RP . | 95% CI . | Chi2 . | P . |
Monthly volumes and costs | ||||||
Inpatient sector | ||||||
Total volumes (1000 DDDs) | −9.4 | [−15.0, −3.8] | −9.4 | [−15.0, −3.8] | 0.0 | 0.9845 |
Original drugs | −19.6 | [−24.6, −14.7] | −19.6 | [−24.6, −14.7] | 0.0 | 0.9985 |
Generic drugs | 10.3 | [8.1, 12.5] | 10.3 | [8.1, 12.5] | 0.0 | 0.9941 |
Total costs (1000 CNY) | −420.0 | [−586.6, −253.4] | −420.0 | [−586.6, −254.8] | 0.0 | 0.9986 |
Outpatient sector | ||||||
Total volumes (1000 DDDs) | 12.0 | [−19.9, 43.8] | 12.0 | [−19.9, 43.8] | 0.0 | 0.9854 |
Original drugs | −6.5 | [−35.8, 22.9] | −6.5 | [−35.8, 22.9] | 0.0 | 0.9564 |
Generic drugs | 18.4 | [8.8, 28.1] | 18.4 | [8.8, 28.0] | 0.0 | 0.9996 |
Total costs (1000 CNY) | −70.1 | [−186.2, 45.7] | −70.1 | [−186.2, 45.6] | 0.0 | 0.9976 |
Monthly hospital visits | ||||||
Inpatient admissions (1000) | 0.3 | [−0.6, 1.2] | 0.3 | [−0.6, 1.2] | 0.0 | 0.9931 |
Outpatient visits (1000) | 7.0 | [−16.0, 30.0] | 7.0 | [−15.7, 29.6] | 0.0 | 0.9996 |
Appendix 12. Literature review
Pharmaceutical policies in China
In the 1990s, the Chinese government set a price ceiling for individual drugs using the cost make-up method. Although this price ceiling had been adjusted more than 30 times between 1997 and 2013 (Wu et al., 2015), it was still largely ineffective in controlling drug prices, as 20% of drugs were of higher prices than in any other countries worldwide (Hu and Mossialos, 2016). Because of high prices and volume growth, drug costs were still increasing and became unaffordable for the public (Yip et al., 2012). In 2009, China adopted a new round of extensive healthcare reforms because its pricing and reimbursement policies had changed significantly. Specifically, the supply-side policies included tendering for off-patent drugs and negotiation for patent drugs and market authorization. Meanwhile, the demand-side policies included the Essential Medicines Program, secondary negotiation and a list for reimbursement (Hu and Mossialos, 2016). The Essential Medicines Program covered the whole life cycle of drugs, and two major components of this program, namely centralized tendering and ZMDP, were targeted directly at cost saving (Chen et al., 2014). In 2015, tendering had become the strongest and most widely used pricing policy, replacing the direct price ceiling. Tendering was conducted mainly at provincial levels with different assessment criteria, and the tendering results contained fixed manufacturers and prices during the implementation period, which was usually one year. However, studies on the impact of tendering are limited. Only one study has been conducted on essential medicines, which revealed a 25% decrease in drug prices (Hu, 2013). Moreover, as the tendering process emphasized price more than quality to win the tender, it led to major concerns about procured drugs’ quality. In 2018, the National Healthcare Security Administration launched the collective pharmaceutical procurement policy. By gathering 11 pilot cities’ yearly drug procurement volumes and ruling the manufacture bidding the lowest would win, this policy succeeded in reducing drug prices (Jiang et al., 2020), without empirical evidence on other drug procuring–related indicators like volume and cost. However, as the bid-winning enterprise would obtain the entire bidding object, the collective pharmaceutical procurement policy might lead to risks of monopoly, low accessibility caused by the market exit of manufacturers and a rise in the prices of other drug substances not covered by this policy (Jiang et al., 2020).
Impact of the ZMDP in China
The ZMDP was a major component of the Essential Medicines Program and was a drug policy widely evaluated in the literature. As it abolished the 15% markup from the hospital’s procurement price to the retail price, it theoretically could reverse the distorted economic incentives in hospitals, which relied extensively on drug revenues to operate (Shi et al., 2019).
Most studies evaluated the impacts of the ZMDP on costs, based on prescription or claim data. Generally, these studies revealed the associations of the ZMDP with reductions in drug costs and medical costs per visit or prescription (Li et al., 2013; Yang et al., 2013; Chen et al., 2014; Wei et al., 2017; He et al., 2018; Li et al., 2018; Zeng et al., 2019), and reductions in drug revenues and shares of total revenues (Yang et al., 2013; Chen et al., 2014; Yi et al., 2015; Tian et al., 2016). However, after the ZMDP, hospitals and/or physicians had strong incentives to increase revenues from other sources to offset losses in drug sales, leading the unintended policy impacts of increased use of diagnostic tests and medical consumables (Fu et al., 2018), health services (Li et al., 2013) and even traditional Chinese medicines (Wen Chen, 2014). They might also extend the length of hospital stay (Li et al., 2013) and encourage hospital visits (Li et al., 2013;Yi et al., 2015; Fu et al., 2018) for more profit. To conclude, though the ZMDP removed physicians’ direct incentives of overprescribing drugs, they would turn to other revenue sources for profit seeking, and even the incentive of overprescribing drugs could not be completely eliminated, as the government subsides were highly related to drug sales (Yi et al., 2015). These unintended effects crippled the ZMDP’s impacts a lot, leading to ineffectiveness in the long term (Ding and Wu, 2017; Zeng et al., 2019).
Studies evaluating the policy impact on drug prices and consumption volumes were limited. Of these, a study conducted in Hubei province estimated the increases in antibiotics’ volumes and costs by 17.30% and 14.79%, respectively, and no impact on prices (Tang et al., 2018). Another study conducted in four Chinese provinces revealed a 34.4% reduction in drug prices (Song et al., 2014), which was a similar proportion to the 30% reduction estimated by the government.
International impact of RP policy
The RP system was an international measure to control increasing drug costs and was mostly implemented in developed countries because of its high prerequisite for the domestic drug market (Kaplan et al., 2012). Most of the empirical studies were also conducted in developed countries such as European countries, New Zealand, Canada (British Columbia), America (Medicaid, and Australia, and showed that the RP policy could steer the demand to low-priced drugs and indirectly compel producers to reduce prices, particularly for high-price drugs (Zweifel and Crivelli, 1996, Danzon and Chao, 2000, Pavenik, 2002, Danzon and Ketcham, 2004, Brekke et al., 2007; Puig-Junoy, 2007; Brekke et al., 2011). The implementation of the RP system could also reduce the market share of high-priced original drugs and increase that of generic drugs (Marshall et al., 2002, Aronsson et al., 2001, Schneeweiss et al., 2002a; Schneeweiss et al., 2003; Grootendorst et al., 2005; Grootendorst and Stewart, 2006; Brekke et al., 2011), decrease (Marshall et al., 2002; Schneeweiss et al., 2002a; Schneeweiss et al., 2003) or have no impact (Pavenik, 2002; Grootendorst et al., 2005; Grootendorst and Stewart, 2006; Stargardt, 2010) on total consumption volumes, decrease the costs of targeted drugs and the costs paid by insurers (Narine et al., 1999, Grootendorst et al., 2002; Marshall et al., 2002; Moreno-Torres et al., 2011; Puig-Junoy, 2007; Kibicho and Pinkerton, 2012; Kaiser et al., 2014; Koskinen et al., 2015; Mardetko and Kos, 2017; Mardetko and Kos, 2018) and have no impact on patients’ health outcomes and healthcare utilization (Thomas et al., 1998; Grootendorst et al., 2001; Grootendorst et al., 2002; Hazlet and Blough, 2002; Schneeweiss et al., 2002b; Schneeweiss et al., 2003; Stargardt, 2010; Lessing et al., 2015).
In developing countries, only South Africa has implemented and evaluated its RP system, revealing discrepant findings in two sample drugs. The implementation of RP policy in South Africa was consistently associated with decreases in the prices of original drugs and decreases in total costs, but its impact on drug utilization varied, with a 19.6% reduction for candesartan and a 15.6% increase for rosuvastatin (de Jager and Suleman, 2019).
Reforms in Sanming, Fujian Province, China
Sanming City in Fujian Province, China, was a benchmark of medical system reforms in China because of its systemic reforms to release the deficit of the medical insurance fund, which was mainly caused by the aging of beneficiaries. Sanming conducted the systemic reforms in all public hospitals simultaneously in January 2013. Three crucial areas were reformed, namely, the hospital’s governance structure, payment system and compensation method. The payment system was modified to reverse the distorted economic incentives of healthcare providers by increasing the prices for labor-intensive services and the implementation of the ZMDP. One study evaluated Sanming’s systemic reforms as a whole and compared 22 public hospitals in Sanming with 187 hospitals in Fujian province. It revealed decreases in medical costs and drug costs per visit or per admission. In September 2014, Sanming implemented the RP policy, which grouped drugs by generic names and set the reference price at the lowest price of the generic drug. The RP in Sanming was adopted only in the inpatient sector of hospitals and covered only 15 drug substances as a pilot project. To our knowledge, the RP policy in Sanming was the first RP system in China and has never been empirically evaluated to date.
Author notes
Bin Jiang, Ruo Jing Zhou contributed equally to this work.