FDA indicates US Food and Drug Administration; GPIs, Generic Product Identifiers; NDAs, New Drug Applications; and NDCs, National Drug Codes.
eAppendix 1. Variables and Data Sources
eAppendix 2. Methods for Extracting and Defining Drug Variables
eAppendix 3. List of 300 GPIs Meeting Inclusion Criteria
eFigure. Timing of 25% and 50% Price Increases in a Calendar Year for Sole-Source, Off-Patent Drugs, 2009-2018
Customize your JAMA Network experience by selecting one or more topics from the list below.
Identify all potential conflicts of interest that might be relevant to your comment.
Conflicts of interest comprise financial interests, activities, and relationships within the past 3 years including but not limited to employment, affiliation, grants or funding, consultancies, honoraria or payment, speaker's bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued.
Err on the side of full disclosure.
If you have no conflicts of interest, check "No potential conflicts of interest" in the box below. The information will be posted with your response.
Not all submitted comments are published. Please see our commenting policy for details.
Alpern JD, Shahriar AA, Xi M, et al. Characteristics and Price Increases Among Sole-Source, Off-Patent Drugs in the United States, 2008 to 2018. JAMA Netw Open. 2020;3(8):e2013595. doi:10.1001/jamanetworkopen.2020.13595
在这项涉及 300 种药品的横断面研究中，价格普遍大幅上涨。低价药品、抗肿瘤和皮肤病类别药品以及在 1990 年之前取得美国食品药品管理局的批准等特质与日历年内价格大幅上涨有关。
Some sole-source, off-patent drugs in the United States have undergone substantial price hikes in recent years. Despite increased attention by lawmakers, there are limited data to guide policy.
To describe key attributes of sole-source, off-patent, off-exclusivity drugs; to characterize the prevalence of price increases; and to identify attributes associated with price increases.
Design, Setting, and Participants
In this cross-sectional study, 300 sole-source, off-patent, off-exclusivity drug products met inclusion criteria and were selected for analysis from January 1, 2008, to December 31, 2018. Attributes were identified from multiple sources, and yearly wholesale acquisition cost prices were determined from First Databank.
Main Outcomes and Measures
The association of drug attributes with the following 2 price change thresholds was measured after adjusting for inflation: 25% or more price increase in a calendar year (wholesale acquisition cost) and 50% or more price increase in a calendar year. The rate of annual price increase over time was also measured.
Of the 300 drug products and 2242 observations analyzed, the overall inflation-adjusted mean increase in drug prices was 8.8% (95% CI, 7.8%-9.8%) per year. Ninety-five drugs (31.7%) increased by 25% or more during any calendar year, and 66 drugs (22.0%) increased by 50% or more during any calendar year. An initial price of less than $2 per unit (adjusted odds ratio [aOR], 2.36; 95% CI, 1.69-3.29), antineoplastic and immunomodulatory class (aOR, 2.72; 95% CI, 1.31-5.65), dermatologic class (aOR, 2.95; 95% CI, 1.80-4.84), oral route (aOR, 2.01; 95% CI, 1.45-2.79), and US Food and Drug Administration (FDA) approval before 1990 (aOR, 1.52; 95% CI, 1.14-2.03) were attributes of drugs that were more likely to be associated with a 25% or more price increase in a calendar year after adjusting for by initial price. Similarly, an initial price of less than $2 per unit (aOR, 2.68; 95% CI, 1.76-4.09), antineoplastic and immunomodulatory class (aOR, 3.07; 95% CI, 1.54-6.12), oral route of administration (aOR, 1.70; 95% CI, 1.11-2.60), and FDA approval before 1990 (aOR, 2.02; 95% CI, 1.40-2.94) were attributes of drugs that were more likely to be associated with a 50% or more price increase in a calendar year after adjusting for by initial price. Price increases of 25% or more were most common in 2014, and price increases of 50% or more were most common in 2013.
Conclusions and Relevance
Price increases among sole-source, off-patent drugs are common, and policy interest in this practice is warranted. These findings should inform state drug pricing legislation.
The United States spends more per capita on prescription drugs than any other country.1 In 2018, $335 billion was spent on prescription drugs2 (more than $1000 for every adult and child), associated largely with spending on brand-name drugs.3 Once the patent and exclusivity period expires on a brand-name drug, generic manufacturers may enter the market, and price reductions are often achieved through competition between multiple generic manufacturers. In 2018, generic drug use by Medicare and Medicaid enrollees saved taxpayers an estimated $137 billion.4
However, substantial price increases affecting generic drugs with limited or no manufacturer competition have drawn attention to a subset of the off-patent drug marketplace that is functioning suboptimally.5 Off-patent drugs that are sole source (ie, those produced by only 1 manufacturer) seem to be particularly prone to price hikes compared with drugs with more than 1 manufacturer.6 Despite opportunities for generic manufacturer competition, barriers to generic market entry have created monopoly conditions,7 allowing manufacturers to charge high prices despite patents and other exclusivities having ended. Notable examples of this dysfunctional drug market include the 500% price increase of albendazole (Albenza),5 the greater than 5000% price increase of pyrimethamine (Daraprim),8 and the 600% price increase of the epinephrine auto-injector (EpiPen).9
Policy solutions are being pursued because unpredictable and steep price increases that affect sole-source, off-patent drugs limit patients’ access to drugs essential to their health. State lawmakers have become increasingly engaged, developing legislation to improve transparency, lower reimbursement rates, or prohibit excessive price increases among off-patent drugs that reach a predetermined threshold.10,11 The thresholds and characteristics of drugs targeted in these policies vary substantially.
Despite substantial public interest and political attention to the problem, there are few data to inform policy solutions. Furthermore, although certain drug price hikes have gained the attention of the public and lawmakers, the full extent of this practice and characteristics of the market remain poorly understood. In this study, we characterize sole-source, off-patent drugs in the United States over a 10-year period to explore associations between drug attributes and large price increases occurring within a calendar year.
This cross-sectional study analyzed sole-source, off-patent drugs and was conducted in accordance with the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline. The HealthPartners Institutional Review Board deemed this study exempt from review because no human participants were involved.
We sought to identify off-patent drugs with 1 manufacturer in the United States for which pricing information is available. We identified potentially eligible drugs with their New Drug Application (NDA) number and active ingredient list using the US Food and Drug Administration’s (FDA’s) list of off-patent, off-exclusivity drugs without an approved generic,12 released in June 2019. We included NDAs listed within Part I and Part II of this list, representing drugs with an approved NDA and no Abbreviated New Drug Application (ANDA) referencing the NDA on the US market. We excluded NDAs discontinued from the FDA’s Orange Book, active ingredients that contained multiple nutritional components, and NDAs therapeutically equivalent to an approved ANDA. Intravenous drugs in plastic containers were also excluded if an ANDA with the same active ingredient, administration route, and formulation as the NDA was listed in the Orange Book. To obtain pricing information, we identified National Drug Codes (NDCs) associated with each NDA using the FDA’s NDC Directory database and the RxMix Interface database.13 Each NDC was linked with its associated Generic Product Identifier (GPI), a classification system that identifies a drug based on its drug class grouping, name, strength, and dosage form, within the Medi-Span database (Wolters Kluwer).14 We defined GPI as the drug product for the analysis. We excluded NDCs with no pricing records, missing GPIs, and GPIs in which 1 NDC linked to multiple GPIs. Finally, we excluded all GPIs with less than 2 years of pricing data available after the expiration date of the listed patent or exclusivity period, as indicated by the FDA’s 2008 Orange Book.15 A flowchart for NDA, NDC, and GPI selection is shown in the Figure.
For each NDA, we identified drug attributes that relate to how a drug may be used or priced. The following variables were identified from multiple data sources detailed in eAppendix 1 in the Supplement: the initial price of a drug at the beginning of the study period, drug administration route, dosage form, class, NDA submission classification, FDA review priority, orphan drug designation, inclusion on the World Health Organization (WHO) Model List of Essential Medicines,16 history of at least 1 drug shortage, the strength of recommendation for use based on the IBM Micromedex DrugDex compendium,17 the number of FDA-approved indications, the presence of alternative drug products used for the same condition, and the year of FDA approval. A detailed description of the methods used for extracting and defining a drug variable is available in eAppendix 2 in the Supplement.
Current and historical wholesale acquisition cost (WAC) prices were extracted for each NDC within First Databank,18 a drug information database stored within HealthPartners’ Administrative Data Warehouse, from January 1, 2008, to December 31, 2018. We receive weekly data updates from First Databank’s National Drug Data File. This data table is stored in our HealthPartners’ Administrative Data Warehouse, an Oracle reporting environment that is available to our HealthPartners Institute Research Informatics team. Some of the specific attributes reported include NDCs, WACs, and effective date. We assumed the WAC for an NDC was unchanged until a new effective date was listed, indicating the date that a price change occurred. National Drug Code obsolete dates were obtained from First Databank, indicating the estimated date on which the NDC was reported to be discontinued. If no obsolete date was listed, we assumed that the NDC was active. To hold inflationary effects on pricing constant, all drug prices were converted to 2018 US dollars.19 Within a GPI grouping, if multiple NDCs were present during the same period, an average price was obtained per year for the overlapping NDCs. Year-by-year price changes were calculated at the GPI level from 2008 to 2018 using the last price listed within a given year.
The percentage price change per year from 2008 to 2018 and the associated 95% CI were estimated using a hierarchical linear model with annual year prices in log scale as the dependent variable and calendar year as the independent variable. An autoregressive covariance structure was used to model the multiple drug prices over time per GPI. To analyze the association of drug attributes with a 25% and 50% price increase in any year over the study period, we used a logistic generalized estimating equation model and a compound symmetry covariance structure to account for multiple observations per GPI over time. We used 2 approaches to calculate the odds of a 25% and 50% price increase in each calendar year over the study period for each level of drug attributes: one is the crude odds ratio (OR), and the other is an adjusted OR (aOR) based on the multivariable model controlling for initial price and time association. Missing attribute values were grouped together in a separate level and included in the analysis. All analyses were conducted using SAS, version 9.4, statistical software (SAS Institute Inc).
A total of 197 NDAs, 840 NDCs, and 300 GPIs met the inclusion criteria and were included in the analysis, with a total of 2242 observations (eAppendix 3 in the Supplement). Descriptive statistics are listed in Table 1. Drugs with an initial price of $50 or more per unit were most common (104 of 300 [34.7%]) followed by drugs with an initial price of less than $2 per unit (77 of 300 [25.7%]). New molecular entities were the most common NDA submission classification (114 of 300 [38.0%]) followed by new dosage forms (95 of 300 [31.7%]). Injection or intravenous was the most common drug administration route (116 of 300 [38.7%]) followed by oral (85 of 300 [28.3%]). Eighty-two of 300 drugs (27.3%) received a priority review designation by the FDA. Most drugs received a strength of recommendation rating of either class I (recommended) or class IIa (recommended most of the time) (183 of 300 [61.0%]) by the IBM Micromedex DrugDex compendium. One-quarter (74 of 300 [24.7%]) were listed on the WHO Model List of Essential Medicines, and one-third (100 of 300 [33.3%]) had an orphan drug designation.
The overall inflation-adjusted mean increase in drug prices was 8.8% (95% CI, 7.8%-9.8%) per year of the study period (Table 1). Price increases of 25% or more were most common in 2014, and price increases of 50% or more were most common in 2013 (eFigure in the Supplement). Of drug products with at least a 25% or 50% price increase in a calendar year, most increased by less than $20 on an absolute basis (170 of 206 [82.5%] and 84 of 103 [81.6%], respectively) (Table 2). The top 20 drug products with the greatest percentage price increase in a calendar year are listed in Table 3. Among these, the absolute change in price observed over the study period varied substantially. For example, cupric chloride only increased on an absolute basis by $1.90 over the study period. In contrast, pyrimethamine increased by $750.00, and edetate calcium disodium (calcium disodium versenate) increased by $1056.50 over the study period (and from $18.08 per unit in 2012 to $291.45 per unit in 2013). Similarly, Samarium SM-153 lexidronam pentasodium (Quadramet) increased from $2603.25 per unit in 2016 to $12 825.06 per unit in 2017.
Approximately one-third of drug products (95 of 300 [31.7%]) had prices that increased by 25% or more during any calendar year, and approximately one-quarter of drug products (66 of 300 [22.0%]) had prices that increased by 50% or more during any calendar year (Table 1). For drugs with a price increase of 25% or more, we observed 206 price increases during the study period, averaging 2.2 price increases of those experiencing at least 1; similarly, for drugs with a price increase of 50% or more, we observed 103 price increases during the study period, averaging 1.6 price increases.
An initial price of less than $2 per unit (aOR, 2.36; 95% CI, 1.69-3.29), antineoplastic and immunomodulatory class (aOR, 2.72; 95% CI, 1.31-5.65), dermatologic class (aOR, 2.95; 95% CI, 1.80-4.84), oral route (aOR, 2.01; 95% CI, 1.45-2.79), the presence of 2 FDA indications (aOR, 1.69; 95% CI, 1.25-2.28), and FDA approval before 1990 (aOR, 1.52; 95% CI, 1.14-2.03) were attributes more likely to be associated with a 25% or more price increase in a calendar year after adjusting for by initial price compared with the average of all attributes. These results are summarized in Table 4.
An initial price of less than $2 per unit (aOR, 2.68; 95% CI, 1.76-4.09), antineoplastic and immunomodulatory class (aOR, 3.07; 95% CI, 1.54-6.12), oral route (aOR, 1.70; 95% CI, 1.11-2.60), and FDA approval before 1990 (aOR 2.02; 95% CI, 1.40-2.94) were attributes more likely to be associated with a 50% or more price increase in a calendar year after adjusting for by initial price compared with all attributes. These results are summarized in Table 4.
This study is the first to date to comprehensively analyze the drug attributes and prevalence of price increases affecting sole-source, off-patent, and off-exclusivity drugs in the United States. We found that price increases of at least 25% and 50% within a given year were common, and prices increased on average by 8.8% per year. Drugs with an initial price of less than $2 per unit, antineoplastic and immunomodulatory class, oral administration route, and FDA approval before 1990 were more likely to sustain price increases of 25% and 50% within a given year. Our findings confirm that price hikes affecting this drug market are not one-off events8,20,21 but rather are common occurrences. Furthermore, average prices increased by 8.8% per year after inflation, which equates to a doubling of the real price in less than a decade. These data are consistent with prior studies22,23 showing that limited manufacturer competition in off-patent drug markets is associated with substantial price increases.
A key finding from the present study is that drugs with low initial prices (<$2 per unit) were more likely to increase by 25% and 50% or more in a given year. Manufacturers may be more likely to target lower-priced drugs for price hikes because they are less likely to capture headlines and the attention of lawmakers. In addition, there may be more opportunity to raise prices up to consumers’ willingness to pay for essential medications for drugs at lower absolute price levels. Similarly, older drugs were also more likely to increase by 25% and 50% in a calendar year. This finding gives weight to the argument that drug manufacturers may preferentially target older drugs, such as pyrimethamine (approved in 1953) and epinephrine auto-injector (approved in 1987), for steep price increases.
We identified important outlier drugs with high initial prices that had extraordinary percentage price increases (>100%) and absolute changes in price. For example, edetate calcium disodium, used to treat lead poisoning, increased by 1512% from $18.08 per unit in 2012 to $291.45 per unit in 2013. Samarium SM-153 lexidronam pentasodium (Quadramet), indicated for the treatment of pain in patients with osteoblastic metastatic bone lesions, increased by 392.7% from $2603.25 per unit in 2016 to $12 825.06 per unit in 2017. Both drugs treat rare diseases, which we identified as being a common characteristic among drugs with large percentage price increases in our sample. A prior study24 found that price hikes were more prevalent among less frequently prescribed generic drugs.
Drugs used to treat rare diseases typically attract few generic competitors because of limited sales volume,25 which we suspect is a key factor explaining the lack of generic competition among sole-source, off-patent drugs and the corresponding price increases. Other potential barriers to generic entry in this market include companies having difficulty demonstrating bioequivalence for complex drugs and tactics by current manufacturers to limit access to necessary drug samples, as occurred with pyrimethamine.26
We found that price increases of at least 25% were more likely to occur among dermatologic drugs, a class known for price increases affecting off-patent drugs.23 Drugs in the antineoplastic and immunomodulatory class were more likely to sustain price hikes of at least 25% and 50%. This finding is noteworthy considering that all of the price increases among cancer drugs affected old drugs that have been on the market for at least the last 3 decades.
This study has important policy implications. First, the prevalence of price increases in this market provides support for interest in policies that aim to prevent excessive price hikes among sole-source, off-patent drugs. Current policies include certain state drug pricing laws to review price increases reaching a predetermined threshold, the FDA’s Drug Competition Action Plan to improve the efficiency of the generic drug development and approval process,27 and the FDA working group to explore the role of drug importation among sole-source, off-patent drugs.28 Also noteworthy is the high proportion of medically necessary drugs that comprise this market, including new molecular entities (38.0% [114 of 300]), drugs with FDA priority review designation (27.3% [82 of 300]), drugs with class I and IIa recommendations (61.0% [183 of 300]), and essential medicines (24.7% [74 of 300]) (Table 1).
Second, our findings indicate that state drug pricing legislation that reviews or prohibits certain price hikes affecting off-patent drugs should use meaningful thresholds that reflect actual price changes occurring in this market. Because most drugs with price hikes had low initial prices and increased by less than $20 per unit in a calendar year, only a minority of drugs would qualify for review under current legislation that uses a threshold of a WAC price increase of $3000.00 in any 12-month period or course of treatment.29 In this case, lawmakers should consider lowering the bar for review because such policies are unlikely to capture a meaningful number of sole-source, off-patent drugs.
Third, although the large number of state drug pricing laws are well intended, the effectiveness of these policies is unknown, including whether manufacturers will adjust pricing targets based on the thresholds set. Other policies, including the FDA’s Drug Competition Action Plan, may not be able to overcome the lack of incentives for generic entry in this market, particularly for drugs treating rare diseases. In such cases, alternative sources for generic drugs that are outside of the for-profit pharmaceutical system offer a potentially attractive mechanism to stabilize markets for rare diseases and improve affordability, such as nonprofit companies30 and government-sponsored generic drug manufacturing.31 For example, the model outlined in the Affordable Drug Manufacturing Act of 2019 establishes an Office of Drug Manufacturing within the US Department of Health and Human Services, prioritizing the manufacturing of certain drugs, including those in uncompetitive markets on the WHO Model List of Essential Medicines.32
Our study has several limitations. First, our pricing data relied on drug manufacturers providing accurate pricing information. Therefore, some prices may underestimate the true market price of a drug. Second, we were unable to control for changes in manufacturing costs that affect sale prices. Third, per-unit costs only partially account for the consequences of pricing changes among patients: the quantity matters too, and differential quantities could also result in differential burden in per-unit price increases. Likewise, we were unable to account for the population burden of these price increases by accounting for the prevalence in use of drugs that had large price increases. We intend to address these 2 limitations in future research. Fourth, although the FDA released a new list of off-patent, off-exclusivity drugs without an approved generic before completion of the study, we used the June 2019 version, which was the most recent list available at the time of data collection. Fifth, we attempted to exclude drugs from our sample with a patent or exclusivity during the analysis period by analyzing historical Orange Book data. However, manufacturers may have extended a patent for a drug at a later time, which would not have been identified using our methods.
Price increases among sole-source, off-patent drugs were common between 2008 and 2018. Price increases preferentially affected lower-priced drugs, the dermatologic and oncology classes, and older drugs. These data substantiate the interest in policies aiming to combat price hikes affecting sole-source, off-patent, off-exclusivity drugs and may help to inform lawmakers who are developing state drug pricing legislation.
Accepted for Publication: May 24, 2020.
Published: August 17, 2020. doi:10.1001/jamanetworkopen.2020.13595
Open Access: This is an open access article distributed under the terms of the CC-BY-NC-ND License. © 2020 Alpern JD et al. JAMA Network Open.
Corresponding Author: Jonathan D. Alpern, MD, CTH, HealthPartners Institute, 8170 33rd Ave S, Mail Stop 23301A, Minneapolis, MN 55425 (email@example.com).
Author Contributions: Dr Alpern had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Concept and design: Alpern, Kodet, Stauffer, Pawloski, Dehmer.
Acquisition, analysis, or interpretation of data: Alpern, Shahriar, Xi, Thapa, Stauffer, Vazquez Benitez, Pawloski, Dehmer.
Drafting of the manuscript: Alpern, Xi, Thapa, Stauffer, Vazquez Benitez.
Critical revision of the manuscript for important intellectual content: Alpern, Shahriar, Kodet, Stauffer, Pawloski, Dehmer.
Statistical analysis: Xi, Vazquez Benitez.
Obtained funding: Alpern, Stauffer.
Administrative, technical, or material support: Kodet, Pawloski.
Conflict of Interest Disclosures: All authors reported being funded by Arnold Ventures for this work. No other disclosures were reported.
Funding/Support: This study was funded by Arnold Ventures.
Role of the Funder/Sponsor: The funding source had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.