|
MARKETWATCHExtensions Of Intellectual Property Rights And Delayed Adoption Of Generic Drugs: Effects On Medicaid Spending
Rising prescription drug costs present a critical policy issue for Medicaid. Generic substitution can reduce costs, but the effects are undercut by extensions of intellectual property (IP) protection, elevated generic prices, and low substitution rates. Using Medicaid prescription data for amoxicillin/clavulanate, metformin, and omeprazole, we calculated the savings that could have been realized if generic drugs had been available and fully substituted at their lowest cost when IP protection first expired (an average delay of twenty-six months). The delay in availability, elevated prices, and slow uptake of generic alternatives for these three drugs alone cost Medicaid $1.5 billion in 2000–2004.
THE RISING COST OF medications purchased by government programs is becoming one of the most important health policy issues of the decade.1 Within Medicaid, the federal- and state-funded health care insurance program for the poor, annual prescription drug spending increased more than 18 percent per year from 2000 to 2002, topping $31 billion in 2003.2 We sought to analyze how unusual extensions of intellectual property (IP) protection and the availability and use of generic drugs affect Medicaid’s prescription drug costs. Background on patents. The patent is the primary means used to protect the intellectual property of new pharmaceutical discoveries. A patent is an effective tool to promote innovation and research because the patent holder can earn a monopoly lasting for up to twenty years. However, because manufacturers usually apply for patents before completing the clinical trials required for Food and Drug Administration (FDA) approval, the functional duration of such protection is less than the full patent life. In the Hatch-Waxman Act of 1984, the U.S. government restored some of the patent term lost to the regulatory approval process.3 As a result, the time from market entry to expiration of the original IP protection grew from an average of nine years in 1984 to 11.5 years by the 1990s.4 To improve the availability of generic drugs after such protection expires, the Hatch-Waxman Act simplified the approval process for generic products. (Bioequivalent generic products have nearly always been shown to have identical therapeutic effects.)5 Generics increased from 19 percent in 1984 to more than 47 percent of U.S. prescription drug sales by 2002, spurred also by proliferation of drug formularies and mandatory generic-substitution laws.6 In 2004, brand-name drugs cost an average of $84 per prescription, versus an average of $31 for generics.7 Several reports have detailed millions of dollars in cost savings achievable through generic-substitution policies employed by government and private payers.8 Delay and underuse of inexpensive generics. Despite some success in promoting the availability of generic medications, three important factors can hinder the optimization of savings from generic drug use. First, generic drugs are often delayed from reaching the U.S. marketplace after expiration of the patent or other IP protection. To extend monopoly protection of the underlying active ingredients in their drug products, brand-name manufacturers sometimes engage in a process known as "evergreening" by patenting peripheral features of products, including aspects of their formulation, their metabolites, or methods of administration. A 2002 report from the Federal Trade Commission (FTC) detailed instances where brand-name manufacturers maintained market exclusivity by listing improper or invalid patents with the FDA.9 Also, the U.S. government has authorized further extensions in market exclusivity after the Hatch-Waxman Act. For example, in 1997 the FDA Modernization Act (FDAMA) provided six more months of market exclusivity if a drug company studies its brand-name product in pediatric populations.10 Second, generic prices can remain elevated because the Hatch-Waxman Act provides six months of exclusivity to the first generic product on the market. A market controlled by such a duopoly does not provide the same cost savings as an open market with multiple generic manufacturers.11 Finally, physicians can be slow to switch to generic versions of brand-name pharmaceuticals, and payers’ formularies often do not require such substitution.12 Purpose of this study. In numerous instances, the market exclusivity of pharmaceutical products was extended and generic drug prices remained elevated even after IP protections were supposed to expire. We identified three such instances and sought to quantify their economic impact on Medicaid. Our goal was to assess the cost of market-exclusivity extensions, artificially elevated generic prices, and slow adoption of generics when therapeutically equivalent generics could have been available sooner and more widely prescribed at lower prices.
Drug selection and patent expiration analysis. We selected three drugs for detailed analysis: amoxicillin/clavulanate potassium (Augmentin, GlaxoSmithKline), metformin (Glucophage, Bristol-Myers Squibb), and omeprazole (Prilosec, AstraZeneca). These drugs were each awarded long IP-related extensions of market exclusivity for different reasons and thus can serve as effective illustrations of the economic impact of generic delay. For each drug, a physician-attorney with training in patent law (Aaron Kesselheim) reviewed patent filings and court documents to define the first date of patent expiration, as well as the timing of and reasons for subsequent extensions. These dates were then compared with the dates at which generic products actually became available. (In determining the original expiration date of the patent on the underlying clinical entity, we included patent life extensions approved by Hatch-Waxman.) Data sources. The Centers for Medicare and Medicaid Services (CMS) provides aggregated quarterly drug spending data from each state Medicaid program categorized by National Drug Code (NDC) number.13 We used the NDC number to link the study drugs to the National Drug Data File to determine generic or branded status, formulation, and strength.14 For each drug, we identified the total number of prescriptions filled, units of medication dispensed, and amount paid by state Medicaid programs. For forty-nine states and the District of Columbia (Arizona data were not available), we analyzed data from the first quarter of 2000 through the fourth quarter of 2004, the last quarter of Medicaid data available at the time of the study.
Calculation of drug costs.
For each study drug, we calculated the number of branded and generic units reimbursed by Medicaid per quarter. The analysis was limited to products sold as tablets or capsules, the usual form in which these drugs are supplied. The CMS data reveal states’ reimbursement to pharmacies for each drug product but do not include the mandatory rebates that states receive from drug manufacturers participating in the Medicaid system.15 We accounted for these rebates using recent federal government analyses of Medicaid drug spending.16 Dividing the calculated Medicaid reimbursement for each drug by the units dispensed yielded the adjusted average cost per unit in each quarter (Exhibit 1
Calculation of potential cost savings. First, we calculated the generic price-related cost savings if the generic version was made available at the lowest average cost observed. Next, we considered the additional effect of IP extensions. We calculated what Medicaid spending would have been if generic use had started at the original (pre-extension) expiration date of the patent or other IP protection covering the active ingredient, assuming that the proportion of generic use remained the same. Finally, we calculated the total savings for each quarter if there was full low-cost generic substitution after the drug’s IP protection should have expired. We compared the reported Medicaid reimbursement to the projected Medicaid expenditures to determine the total savings.17
Determination of patent expiration date and delays in generic availability. Augmentin (amoxicillin/clavulanate). Based on patents originally filed during late 1970s, and including Hatch-Waxman extensions, the last valid patent covering most Augmentin products expired in January 2000.18 GlaxoSmithKline received a six-month extension for pediatric studies. It attempted to enforce later-issued patents that could have extended market exclusivity another fifteen years, and generic drug manufacturers challenged the patents in court.19 After these patents were ruled invalid, the first full quarter of generic availability occurred in the fourth quarter of 2002, a total delay of thirty-three months. Glucophage (metformin). Metformin has been available in Europe since 1979. Bristol-Myers Squibb obtained FDA approval to market the drug in the United States, and in March 1995 it won exclusive rights to sell it as Glucophage for five years until March 2000. (Although this market exclusivity was not a patent, it served as a similar kind of IP protection.) The company received an additional six months of exclusivity for conducting pediatric trials and then petitioned Congress for a three-year exclusivity extension, based on a conflict between the FDA rules and the Hatch-Waxman provisions regarding drug labeling. Legislators instead clarified the language and closed the potential loophole. In part as a result of this legal uncertainty, the first full quarter of generic market availability for metformin did not occur until the second quarter of 2002, a total delay of twenty-four months. Prilosec (omeprazole). The basic patent for Prilosec expired in April 2001, including time restored to the patent life as a result of the regulatory process under Hatch-Waxman.20 Its manufacturer, AstraZeneca, received a six-month extension for conducting pediatric studies. In addition, AstraZeneca claimed that generic manufacturers infringed patents lasting until 2007 on the pill’s coating. In subsequent litigation the courts ruled that at least one generic formulation did not infringe.21 Generic omeprazole became available in the first quarter of 2003, a total delay of twenty-one months. AstraZeneca then marketed a patent-protected derivative of Prilosec, Nexium (esomeprazole), the s-isomer of omeprazole, which it promoted in place of Prilosec. Finally, the FDA granted AstraZeneca the rights to market an over-the-counter (OTC) version of Prilosec in the third quarter of 2003, which it sold at a greatly reduced price.
Determinations of generic cost.
For each study drug, the cost to Medicaid fell once the six-month generic exclusivity period ended and multiple generic products became available. This decrease in costs was 22 percent for omeprazole, 24 percent for amoxicillin/clavulanate, and 55 percent for metformin (Exhibit 1
Rate of generic use.
Physicians were slow to adopt generic equivalents even when they did become available. In the case of omeprazole, the highest level of generic prescribing observed in the data set was only 67 percent (Exhibit 2
Aggregated cost savings. Exhibits 3  –5 show the calculated Medicaid spending (including estimated rebates) as well as a projection of the potential savings to Medicaid for each study drug, starting in the quarter when a generic version could have been on the market.
For amoxicillin/clavulanate, the lowest average generic cost was $3 per unit in the fourth quarter of 2004. If the generic form had been available at this cost in the fourth quarter of 2002, Medicaid could have saved $23 million. If generic use had started in the first quarter of 2000, Medicaid would have spent slightly more than it did on the brand-name version through the first quarter of 2001, because the adjusted brand-name cost in those quarters was less than $3 per unit. However, starting in the second quarter of 2001, Medicaid would have incurred a savings. Repeating these calculations through the fourth quarter of 2004 reveals a projected savings of $28.4 million. If the generic form had been fully substituted in the first quarter of 2000, the total potential savings to Medicaid would have been $28.5 million (Exhibit 3  ).
For metformin, the lowest average generic cost was $0.33 per unit in the fourth quarter of 2004. If the generic version had been available at this cost in its first quarter of availability (the second quarter of 2002), Medicaid could have saved $100.3 million over the subsequent three and a half years. If generic metformin had been available at the lowest average generic price starting in the second quarter of 2000, when the IP protection originally expired, Medicaid would have saved $283.3 million. If the generic version had immediately achieved full substitution, the total savings to Medicaid would have been $318.3 million through the fourth quarter of 2004 (Exhibit 4
For omeprazole, the lowest average cost was $0.55 per unit, the cost of the OTC version in the second quarter of 2004 sold at prescription drug strength by AstraZeneca. Since OTC omeprazole is available only at a 20 mg dose, and the brand-name and generic versions are available at doses of both 20 mg and 40 mg, we calculated the results per milligram, weighted by actual use patterns. If omeprazole had been available at this price when it was marketed in the first quarter of 2003, the projected savings would have been $222.6 million. If generic omeprazole had been available as an alternative in the second quarter of 2001, the savings would have reached $608.9 million. Potential Medicaid savings would have increased to $860.1 million if there had been complete substitution of the generic version for Prilosec. Finally, if generic omeprazole had been fully substituted for Prilosec and Nexium, an essentially identical product marketed by the same manufacturer, Medicaid would have saved $1.2 billion aggregated through 2004 (Exhibit 5
Exhibit 6
The projected costs of market-exclusivity extensions for prescription drugs have been estimated in the past.22 This analysis quantifies the increase in drug spending from extensions of IP protections on prescription drugs beyond their initial expiration dates, the additional exclusivity granted to the first generic manufacturers, and the incomplete adoption of generics when they become available. This analysis also provides more precise calculations of these costs to the Medicaid program. As Medicaid drug spending increases, many programs are resorting to limiting coverage and increasing patient copayments or deductibles, which have been shown to reduce the use of clinically necessary drugs, and have adopted fixed limits in the number of drugs that enrollees can receive per month.23 We have calculated that stricter enforcement of initial patent expiration, combined with more-active generic substitution, could have produced more than $1.5 billion in savings for Medicaid during the period studied for these three drugs. Savings using these strategies would have less harmful consequences for patients than other strategies that limit coverage or restrict therapeutic choices. Study limitations. Some limitations to our analysis make these findings conservative estimates. We considered the extensions to patent terms granted by Hatch-Waxman to be part of the original length of IP protection because those extensions were enacted more than twenty years ago as a compromise for allowing generic drugs to be marketed sooner. If we had considered the patent terms of drug products to be linked to the original submission dates of their patent applications, the calculated cost savings in this analysis would be much larger. In addition, the average price for the three study drugs fell over time, and the lowest average generic price occurred in the final quarter for which data were available. Although generic drug prices in general have been rising, the lowest average generic prices for these three study drugs could continue to decrease in future quarters if more generic manufacturers enter the market. Finally, Medicaid represents only about 10 percent of the nation’s $200 billion drug expenditure; our findings suggest further billions of dollars in savings for payers other than Medicaid, especially since those programs often have a lower rate of generic substitution. On the other hand, our calculation of the pharmaceutical Medicaid rebate level might not accurately represent the rebates provided for the three study drugs in our analysis. Rebates might be even higher in states such as Maine that have developed state Medicaid drug formularies or preferred drug lists. Since details of the actual rebates are not publicly available, using the government’s calculations helps account for this factor in the Medicaid reimbursement process. In addition, there have been mixed reports over patients’ responses to prescription drug programs that require generic substitution or other similar cost-saving measures. Sebastian Schneeweiss and colleagues reported no effect on outcomes, while another study suggested that patients might stop taking certain medications.24 Most recently, William Shrank and colleagues reported an improvement in therapy after initiation of a pharmacy benefit plan encouraging generic drug use.25 Areas for policy reform. These findings suggest several possible areas for policy reform. In the three cases examined, market exclusivity was extended for an average of twenty-six months. All were granted six-month extensions for conducting tests on pediatric populations. Undeniably, it is important to investigate prescription drugs in populations such as children who are traditionally underrepresented in new product clinical trials. However, the FDAMA’s pediatric-exclusivity provision allows extensions regardless of the outcome of the trial. As a result, a company could spend just a few million dollars to study a drug in children without necessarily obtaining clinically meaningful results and win six months of additional market exclusivity for a product that generates a billion dollars or more in annual sales.26 In 2003 Congress gave the FDA the ability to require pediatric studies for limited indications, although studies could be waived or deferred until after a drug’s approval.27 The FDA also announced that it plans to allow applicants the opportunity to first qualify for the pediatric exclusivity incentives for potentially required studies.28 In 2007, when the pediatric-exclusivity extension must be renewed, Congress could reformulate the incentives by reimbursing companies for their costs in organizing pediatric trials or by providing a bonus equal to twice the cost of the trials. The FTC has made recommendations aimed at preventing other market-exclusivity extensions, including avoiding incorrect or duplicative listing of patents with the FDA, which can block generic drugs from entering the market as the propriety of these listings is litigated.29 It might also be helpful to change the patenting process to allow consumer representatives to challenge some patents and thus allow courts to weigh the inventiveness of a product against its projected implications for public health.30 The likelihood of such legislative solutions is slim in the current political environment, but their importance can be emphasized with further research to show the effects of IP protections on public health care spending, as well as patient outcomes. These findings also show that after a generic version of a drug becomes available, it is not consistently substituted for the brand-name version and even over time does not achieve complete substitution. As of 2003, only thirty-five states had mandatory generic substitution rules in place for their Medicaid programs, and some physicians are slow or unwilling to switch to equivalent generic medications. Limited use of generic medications might result from prescribers’ lack of education about the equivalency of generic formulations, as well as the failure of some state Medicaid programs to mandate generic substitution for bioequivalent drugs.31 In the case of Augmentin, for the first five quarters of our analysis, the cost of generic substitution was higher than what Medicaid actually spent on the brand-name version. Even though generic drugs often cost far less than brand-name drugs in the private market, Medicaid’s unique method of reimbursing for pharmaceuticals can lead to instances where brand-name drugs might be more affordable than their generic counterparts in certain states. This can result from pharmaceutical manufacturers’ publication of artificially elevated average wholesale prices (AWPs), the values on which Medicaid programs determine prescription drug reimbursement rates.32 A number of policy changes have been proposed to the Medicaid payment structure that could prevent these situations, such as basing pharmaceutical prices on average manufacturer prices (AMPs—the average price paid to manufacturers by wholesalers for drugs distributed to retail and mail-order pharmacies) instead.33 There has been much debate over how patents affect drug costs and how to best balance IP protections that provide incentives for research with the need to promote competition and lower prices in the pharmaceutical market, to improve access to drugs and limit public health care budgets.34 If the initial length of IP protection on a given drug were more effectively enforced, pharmaceutical companies could continue to earn appropriate revenues on true medical innovations throughout the patent life. Generic manufacturers, unburdened of the risks and costs of defending patent infringement suits, would be better equipped to bring their drugs to market at a lower price after patent expiration, and providers would better know when generic drugs would be available to their patients. Greater reliance on initial patent terms would also increase the incentive to pursue research into new and truly innovative products rather than to invest in protection of revenues from older ones.35 IT IS VITAL TO RESPECT basic IP rights in drug development to help protect innovation. However, inappropriate extension of such rights, combined with reduced generic substitution, can negatively affect insurance programs, such as Medicaid, that cover drug costs. The implementation of the Medicare prescription drug benefit (Part D) poses such challenges on a much larger scale, which makes this a particularly timely issue for national discourse.
Aaron Kesselheim (akesselheim{at}partners.org) is an attorney and a clinical fellow in medicine, Division of Pharmacoepidemiology and Pharmacoeconomics, Brigham and Women’s Hospital, Harvard Medical School, in Boston, Massachusetts. Michael Fischer is an instructor in medicine in that division; Jerry Avorn is a professor of medicine and the division’s chief.
This article has been cited by other articles:
Comments:Read all Comments
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Top
Study Data And Methods
Study Results
