Determinants of market prices for drugs under Japan’s national health insurance

Abstract

Aims

The Japanese government reimburses patients for drugs at prices specified in the Drug Price Standard (DPS) published by the National Health Insurance (NHI) scheme. It revises reimbursements for most drugs on the basis of their market prices. This study thereby identifies factors related to drugs or disease that impact market prices for drugs using the DPS list.

Materials and methods

This study first examined the 2018 DPS list to identify all listed drugs, their prices, and their stipulated reimbursements. We then excluded from this study all the drugs for which prices are set per alternate rules. We calculated the percentage divergence between market prices and DPS prices and designated it our dependent variable. We performed descriptive and a univariate analysis on each variable and constructed multivariate regression models featuring independent variables for drug characteristics that might affect market prices.

Results

We identified 1,775 drugs with prices revised only by the market. We observed higher percentage divergences between DPS and market prices for drugs with generic alternatives (p < 0.001), drugs listed in the Japanese Pharmacopoeia (p < 0.001), and drugs for which at least two new drugs entered the same therapeutic category (p < 0.001). Injectable drugs exhibited a more significant and negative correlation with percentage divergences (p = 0.009) than ingestible drugs. Drugs that treat specific organs (p < 0.001), affect metabolism (p = 0.001), and those prescribed for non-therapeutic purposes (p < 0.001) display significantly higher percentage divergence than drugs affecting the nervous system and sensory organs. Divergences are less for narcotics (p < 0.001) and drugs that counter pathologic microorganisms and parasites (p = 0.004).

Conclusions

Factors that elevate competition among pharmaceutical companies likely lower market prices for drugs, and the direction of prices under NHI in Japan is affected by the category of diseases a drug treats.

Keywords:

• Drug price standard
• national health insurance
• reimbursement policy
• market price
• Japan

JEL Classification CODES:

• I18
• I1
• I11
• I

Introduction

Japan’s national health insurance (NHI) scheme reimburses patients for prescription medications that are published in the Drug Price Standard (DPS). The Japanese government sets reimbursements for each listed drug, and the reimbursement becomes the drugs’ official price¹. The government revises reimbursements biennially using different criteria. With a few exceptions, revisions are based on prices wholesalers charge medical facilities and pharmacies². As market prices generally are below listed DPS prices, biennially revised prices are often below previous listed prices. Furthermore, the reduction in the revised price will be determined depending on how much the market price is below the DPS-listed price.

The pharmaceutical industry wants to maintain prices or mitigate biennial reductions. The Federation of Pharmaceutical Manufacturers’ Associations of Japan (FPMAJ), the Pharmaceutical Research and Manufacturers of America (PhRMA), and the European Federation of Pharmaceutical Industries and Associations (EFPIA) have expressed concern about curtailing the number of drugs that qualify for a price maintenance premium (PMP)^3,4. Given that the government limits the number of PMP-qualified drugs to reduce its outlays, market prices play an increasing role in mitigating price reductions.

Even though market prices are the only drug prices set without Japanese government intervention, few studies explore how they are determined. Only one study suggests that the number of competitors expands the gap between market prices and reimbursements⁵. Since off-year revisions started in FY2021, pharmaceutical companies need to understand what characteristics influence market price to predict revisions and to develop price maintenance strategies. The government needs to understand where it enjoys leeway in reducing DPS prices and reimbursements. This study identifies features of drugs that create divergences between market and DPS prices.

Methods

Selection of drugs

We consulted 2018 DPS listings and reimbursements⁶. To identify drug prices revised only by market prices, we excluded those subject to price maintenance (PMP, basic drugs, and minimum statutory DPS prices), additional reductions (long-listed drugs, market expansion re-pricing), price increases (recalculations for unprofitable products, premiums for orphan and pediatric drugs), and weighted average pricing (generics)⁷. The terms of pricing rules are explained in Table 1.

Table 1. Term explanations of the rules of price revision for listed drugs.

Term	Type of rules	Explanation
Price maintenance premium	Maintenance	The price revision of new drugs is protected under the patent period when the drug meets a certain condition. The sample conditions are (1) orphan drugs, (2) products developed due to the government’s open requests, (3) products granted pricing premiums, or (4) drugs with new mechanisms of action.
Basic drugs	Maintenance	Drugs whose efficacy and safety have been established through long-term and widespread use in medical practice; drugs whose prices are to be maintained due to the necessity of ensuring a stable supply to the market.
Minimum statutory DPS prices	Maintenance	The minimum drug price selected for each dosage form to secure the minimum profit covering the supply cost. The price is maintained at the time of price revision.
Long-listed drugs eligible for exceptional revisions	Additional reduction	Additional price reduction for original products with slow entry of the generic products.
Market expansion re-pricing	Additional reduction	When the annual sales exceed the expected annual sales by a certain factor, the price will further be reduced during price revision by up to 50%.
Recalculations for unprofitable products	Price increase	Increase in the price of a drug that cannot be continuously manufactured and sold by the manufacturer/distributor because the drug is significantly low-priced.
Premiums for orphan and pediatric drugs	Price increase	When pediatric or orphan indications are approved or true clinical utility has been verified after its launch, premiums are added to the price.
Weighted average pricing (generics)	Weighted average	The price of generics is categorized into the following three weighted average prices: One drug price is selected for all generic drugs with a calculated price of <30% of the maximum price. One drug price is selected for all generic drugs with a calculated price of ≥30% and <50% of the maximum price. One drug price is selected for all generic drugs with a calculated price of ≥50% of the maximum drug price.

Variables

Figure 1 shows the association between the market price and DPS price. As a dependent variable we used the percentage difference between market and DPS prices, calculated as: $$\text{Percentage divergence rate }\left(\mathrm{\%}\right)=\mathrm{ }\frac{\left(\text{Prior NHI price * }\left(1\mathrm{ }+\text{ adjustment }\left(0.02\right)\right)\mathrm{ }–\text{ Revised NHI price}\right)\mathrm{*}100}{\text{Prior NHI price}}$$

Figure 1. Mechanism of DPS price revision based on the market price.

The current DPS list is dated 31 March 2018⁸. Where prior and revised DPS prices were identical, we hypothesized a 1% divergence from market price as the midpoint of the 2% adjustment. We extracted explanatory variables from the final selection using public data for availability of generics, whether drugs are ingested or inoculated, therapeutic categories, number of new drugs in the same therapeutic category from April 2016 to March 2018, and publication in the Japanese Pharmacopoeia^6,9–11. Table 2 represents details for all parameters.

Table 2. Characteristics of drugs in the analysis.

Characteristics	Values	Description	Reference
Existence of generics, n (%)	307 (17%)	The drug has generics with the same gradient on NHI price list	6
Form of the drugs		The drugs were categorized into three forms:	6
Internal medicine, n (%)	720 (41%)	Examples: tablets, capsules, powders, dry syrups
Injection medicine, n (%)	772 (43%)	Examples: injections including prefilled syringe
External medicine, n (%)	283 (16%)	Examples: eye drops, nasal drops, inhalants, patches
Class of therapeutic categories		The drugs were categorized into seven therapeutic categories:	9
(1) Nervous system and sensory organs, n (%)	298 (17%)	Examples: Hypnotics and sedatives, anxiolytics, antipyretics and analgesics, anti-inflammatory agents, psychotropics
(2) Individual organs, n (%)	539 (30%)	Examples: Antiarrhythmic agents, antihypertensives, hyperlipidemia agents, bronchodilators, peptic ulcer agents, hormones
(3) Metabolism, n (%)	416 (23%)	Examples: Nutrients, blood and body fluid agents, dialysis agents
(4) Cellular function, n (%)	194 (11%)	Examples: Alkylating agents, plant extract preparations, radioactive drugs, allergic agents
(5) Pathologic microorganisms and parasites, n (%)	179 (10%)	Examples: Antibiotics, antivirals
(6) Not mainly for therapeutic purpose, n (%)	113 (6%)	Examples: Diagnostic agents
(7) Narcotics, n (%)	36 (2%)	Examples: Alkaloidal narcotics
Number of new drugs in the same therapeutic category		Number of the same class drugs listed from April 2016 to March 2018	6,9,10
None, n (%)	952 (54%)
One, n (%)	387 (22%)
Two or more, n (%)	436 (25%)
Listed in the Japanese Pharmacopoeia, n (%)	166 (9%)	The drug is listed in the Japanese Pharmacopoeia 17th edition	11
Total number	1,775

Statistical methods

We performed descriptive analyses upon each variable and applied univariate analysis to percentage divergences for explanatory variables. We applied Mann–Whitney U tests or Kruskal–Wallis tests to categorical variables. We then constructed multivariate regression models to explore independent variables as influences on market prices, which assume linearity, independence, normality, and identically distributed errors. All explanatory variables enter the model. We estimated beta coefficients and 95% confidence intervals using STATA version 14.2.

Results

We identified 16,434 drugs and their prices from the 2018 DPS list. We excluded 14,659 duplicated drugs, drugs subject to PMP (n = 562), basic drugs (n = 660), drugs with minimum statutory DPS prices (n = 2,840), long-listed drugs eligible for exceptional revisions (n = 1,157), drugs eligible for market expansion pricing (n = 30), drugs eligible for recalculation as unprofitable (n = 184), orphan or pediatric drugs eligible for additional premium pricing (n = 49), and generic drugs, including those listed pre-1947 (n = 12,489). We identified the remaining 1,775 drugs as solely market-priced (Figure 2).

Figure 2. Drugs in the analysis revised only by market price.

Characteristics of selected drugs

Table 2 summarizes their characteristics. Among the 1,775 examined drugs, 83% had no equivalent generics, 41% were used in internal medicine, and 43% were injectable. We sorted them into seven therapeutic categories: agents affecting the nervous system and sensory organs (n = 298), organ-specific agents (n = 539), agents affecting metabolism (n = 416), agents affecting cell function (n = 194), agents treating pathologic microorganisms and parasites (n = 179), agents with non-therapeutic uses (n = 113), and narcotics (n = 36). Among sampled drugs, 54% had no new drug in their therapeutic category, and 10% appeared in the Japanese Pharmacopoeia.

Univariate analysis

The mean percentage divergence is 7.15%. Table 3 shows the mean percentage divergence of each drug categorized according to the Japan Standard Commodity Classification. In univariate analysis, available generics and mention in the Japanese Pharmacopoeia correlate positively and significantly with percentage divergence (p < 0.0001). Manner of administration (p = 0.0001), therapeutic category (p = 0.0001), and number of new drugs in the same therapeutic category (p = 0.0001) correlate significantly with percentage divergence (Table 4).

Table 3. Mean percentage divergence of each drug category.

Japan Standard Commodity Classification	Mean percentage divergence (%)
(1) Nervous system and sensory organs	6.56
Agents affecting the central nervous system	6.34
Agents affecting the peripheral nervous system	5.81
Agents affecting sensory organs	6.93
(2) Individual organs	7.69
Cardiovascular agents	7.61
Respiratory organ agents	7.07
Digestive organ agents	7.60
Hormones	7.51
Urogenital and anal organ agents	5.63
Epidermides	7.24
(3) Metabolism	7.21
Vitamins	7.89
Nutrients and tonics	7.13
Blood and body fluid agents	8.43
Dialysis agents	5.76
Other agents affecting metabolism	8.24
(4) Cellular function	6.52
Antineoplastics	8.07
Radioactive drugs	1.09
Allergic agents	4.36
(5) Pathologic microorganisms and parasites	6.18
Antibiotics	6.91
Chemotherapeutics	6.87
Biological preparations	4.58
Parasitics	6.12
(6) Not mainly for therapeutic purpose	9.12
Intracorporeal diagnostic agents	9.87
Other agents not mainly for therapeutic purpose	8.00
(7) Narcotics	5.12
Alkaloidal narcotics	5.05
Non-alkaloidal narcotics	4.91

Table 4. Results of the univariate analysis: association between percentage divergence and covariates.

Categorical variables	Mean percentage divergence	Standard deviation	p-value
Existence of generics
Yes	9.67	3.05	<0.0001
No	6.62	2.75
Form of the drugs
Internal medicine	7.56	2.38	0.0001
Injection medicine	6.87	3.84
External medicine	6.85	1.59
Class of therapeutic categories
(1) Nervous system and sensory organs	6.56	1.62	0.0001
(2) Individual organs	7.69	2.16
(3) Metabolism	7.21	3.38
(4) Cellular function	6.52	3.92
(5) Pathologic microorganisms and parasites	6.18	2.65
(6) Not mainly for therapeutic purpose	9.12	5.11
(7) Narcotics	5.12	1.49
Number of new drugs in the same therapeutic category
None	6.80	3.13	0.0001
One	7.38	2.21
Two or more	7.70	3.36
Listed in the Japanese Pharmacopoeia
Yes	8.82	2.74	0.0001
No	6.97	3.01
Total	7.15	3.04

Multivariate regression analysis

Table 5 displays regression results. Availability of generics, manner of administration, therapeutic category, number of new drugs in the same therapeutic category, and mention in the Japanese Pharmacopoeia correlate significantly with percentage divergence. Percentage divergences are higher for drugs with generics (p < 0.001) or drugs mentioned in the Japanese Pharmacopoeia (p < 0.001). Injectables correlate significantly and negatively with percentage divergence (p = 0.009) compared to ingested drugs. Percentage divergence is significant (p < 0.001) for drugs with two or more new drugs in their therapeutic category. Regarding therapeutic categories, the organ-specific drugs (p < 0.001), agents that affect metabolism (p = 0.001), and that serve non-therapeutic purposes (p < 0.001) exhibit significantly higher percentage divergence than agents affecting the nervous system and sensory organs. In total, agents for pathologic microorganisms and parasites (p = 0.004) and narcotics (p < 0.001) exhibit lower divergences.

Table 5. Multivariate regression analysis.

Categorical variables	Coef.	Standard error	95% CI	p˗value
Intercept	6.06	0.18	5.69, 6.43	<0.001
Existence of generics	2.80	0.18	2.46, 3.15	<0.001
Form of the drugs
Injection medicine	−0.41	0.16	−0.72, −0.10	0.009
External medicine	−0.19	0.20	−0.58, 0.20	0.343
Class of therapeutic categories
(2) Individual organs	0.72	0.19	0.34, 1.10	<0.001
(3) Metabolism	0.71	0.22	0.28, 1.15	0.001
(4) Cellular function	−0.31	0.26	−0.81, 0.20	0.236
(5) Pathologic microorganisms and parasites	−0.76	0.26	−1.28, −0.24	0.004
(6) Not mainly for therapeutic purpose	3.08	0.31	2.48, 3.68	<0.001
(7) Narcotics	−1.72	0.47	−2.65, −0.80	<0.001
Number of new drugs in the same therapeutic category
One	0.19	0.17	−0.14, 0.53	0.259
Two or more	0.98	0.16	0.65, 1.30	<0.001
Listed in the Japanese Pharmacopoeia	0.99	0.23	0.55, 1.44	<0.001

Discussion

Factors that depress market prices are the availability of generics, appearance of new drugs in the same therapeutic category, mention in the Japanese Pharmacopoeia, drugs affecting individual organs and metabolism, and drugs with non-therapeutic purposes. Drugs associated with upwardly revised prices are injectable, treat pathologic microorganisms and parasites, or are narcotics.

Our results indicate that to remain competitive pharmaceutical companies and wholesalers must reduce prices for drugs with generic substitutes and drugs that share therapeutic categories with two or more new drugs. Also, mention in the Japanese Pharmacopoeia indicates a drug is used widely. Our results coincide with findings by Takayama and Narukawa⁵ and suggest that injectables are more susceptible to price increases than ingested medicines.

Drugs that treat pathologic microorganisms and parasites might be more susceptible to price increases because fewer companies are developing them and escalating microbial resistance quickly renders them ineffective. It is argued that incentives are needed to encourage their development¹².

Large percentage divergences between market prices and reimbursements induce the government to reduce prices and reimbursements. Starting in 2021, moreover, the Japanese government will begin revising DPS price lists yearly. To maintain prices, pharmaceutical companies must sidestep the competitive market by developing drugs that resist large discounts in price. Meekings et al.¹³ suggest that targeting orphan drugs is a component of successful biopharma R&D. As an alternative, companies should develop drugs that meet PMP criteria, such as those that command premiums for innovation or usefulness.

Among our study’s limitations, we assumed a percentage divergence of 1% when prior and revised DPS prices are identical. This assumption has little impact on our conclusions because percentage divergences ranged from 0–2%. We excluded drugs subject to alternate pricing rules, as they might exhibit differing percentage divergences. Finally, our basis for comparison is 2018 DPS prices. Future studies should explore longer-run trends in market prices.

Basing biennial price changes of DPS-listed drugs on market prices reflects a drug’s value. The pharmaceutical industry could maintain reimbursements by developing drugs that treat unmet needs. Government could incentivize pharmaceutical investment through price maintenance policies. As government’s drug expenditures are rising, improved pricing policy and closer relationships among market participants are needed.

Conclusion

Our results indicate that factors increasing market competitiveness will likely reduce market prices of drugs, and that the market price of a drug is affected by medical conditions it treats. It requires a strategy for the pharmaceutical company to maintain the value of drugs and for government to set reimbursements for sustainable NIH.

Transparency

Declaration of funding

No funding was received to produce this article.

Declaration of financial/other relationships

No potential conflict of interest was reported by the author.

Peer reviewers on this manuscript have no relevant financial or other relationships to disclose.

Acknowledgements

None reported.

Data availability statement

All necessary data used for this study are included in the manuscript.