IDENTIFYING COST-EFFECTIVE

REGULATIONS TO COST-EFFECTIVENESS STUDIES:

AN ECONOMIC ANALYSIS OF A PROPOSED SET OF REGULATIONS

Zion Shohet

I.INTRODUCTION

Structural changes in the health-care industry and the increased emphasis on cost containment are fundamentally changing the manner in which drugs are purchased in the United States. Increasingly, managed care companies, who generally cover the costs of prescription drugs, influence the selection of drugs by creating formularies, lists of recommend drugs which physicians are expected to prescribe. In order to reduce costs, managed care companies use cost-effectiveness information to create their formularies. Drug companies, realizing the commercial importance of cost-effectiveness claims, are performing cost effectiveness studies to promote their drugs. This intense interest in cost effectiveness has sparked a new research field which is commonly referred to as pharmacoeconomics. Cost-effectiveness research is not directly regulated and there are no universal standards or commonly-accepted standards for performing such studies. In this paper, I perform a cost-effectiveness analysis of potential regulations of cost-effectiveness research, by examining their expected outcomes. The paper focuses on how to preserve some of the natural efficiencies from the unregulated system in a regime in which the FDA promulgates standards for conducting cost-effectiveness research.

H.. DESCRIPTION OF COST-EFFECTIVENESS STUDIES

Cost effectiveness studies are designed to measure the costs and benefits of prescribing a particular drug. Generally, they consider such factors as price, the efficacy of the drug, its side effects, how quickly it works, how easy it is to administer and the likelihood that patients would comply with the dosage requirements.[1] Thus, while cost effectiveness studies examine costs, they go beyond cost savings to assess the overall performance of a drug. Such studies can become very complex because a tested drug may lead to multiple outcomes, the probability of each of which and the costs and benefits associated with each of which must be assessed to determine the drug's overall cost-effectiveness.[2]

Currently, there are no standards or universally accepted practices for performing cost-effectiveness studies. Indeed, the field is rife with debate over how to conduct such studies and how to report their results.[3] Determining how to select the range of relevant clinical outcomes and designing proper research methodologies and statistical techniques has proven extremely challenging. Making assumptions about the value of such factors as: years of life saved, days of morbidity averted and quality of life changes is often highly arbitrary. And, tracking the entire course of a disease, which is essential for a meaningful cost-effectiveness study, can span a life-time for chronic illnesses. Equally as challenging is presenting the results of such studies because, unlike clinical trials which yield narrow conclusions about safety and efficacy, cost-effectiveness studies involve subjective opinions and interpretations.[4] Even the quantitative measures hardly ever come in absolute terms, rather they are stated in relation to a certain benchmark treatment. For example, the results of a study may indicate that an the examined drug requires a $50,000 investment per additional life-year gained over the benchmark treatment for the tested condition.[5]

A recent article demonstrates the complexity of cost-effectiveness studies.[6] It examines five different cost-effectiveness studies of the drug Misoprostol, used to prevent gastric damage. The studies examined whether the use of the drug is cost-effective, comparing it to non-use, rather than to another drug. The results of the five studies ranged from rating Misoprostol as cost-effective to a finding that it is excessively costly.[7] The article explained that the differing conclusions would generally be attributed to the use of different decision models, different probabilities, different target populations, different periods, different cost estimates or computational errors. The article concluded that, in this instance, different approaches to determining Misoprostol's effect on the development of ulcers, different assumptions about hospitalization rates for the condition and different assumptions about the costs of the drug and the costs of certain ambulatory care treatments were the most likely culprits for the inconsistencies.[8] The article criticized all five studies and concluded that Misoprostol cannot be recommended as cost-effective until a future study incorporates better effectiveness data on clinically relevant outcomes and better evaluation methods that integrate all relevant health effects into one common interpretable unit.[9] The article, therefore, goes beyond demonstrating how easily cost-effectiveness information can be manipulated; its suggests that without commonly agreed upon standards the information can be entirely meaningless.

III.RECENT PROLIFERATION OF COST-EFFECTIVENESS STUDIES

There has been a tremendous proliferation in drug cost-effectiveness research. The number of published cost-effectiveness studies about medical interventions has grown from virtually zero in 1966 to nearly 300 in 1990, with drug studies predominating.[10] In fact, in 1992, a new journal, Pharmacoeconomics, was founded, publishing 290 articles dealing with cost-effectiveness issues over the past two years.[11] A recent survey found that the average number of studies conducted by pharmaceutical companies has risen from 5 per company in 1991 to 24 in 1994.[12] Moreover, the percentage of pharmaceutical companies who do not have their own internal full-time pharmacoeconomic staff has declined from 85% in the 1980's to 31% in 1994.[13] In U.S. companies, such internal staffs often number between 6-10 people.[14]

It is no surprise that drug companies are behind much of the proliferation in this information. The primary reason for this is the increased leverage of managed-care companies, who are insisting on such information as the basis for their drug purchasing decisions. But there are other reasons. Drug companies are increasingly developing "me-too" drugs, drugs that duplicate competitors' products with just enough different chemical properties to warrant their own patents, at the expense of riskier breakthrough pharmaceuticals.[15] Indeed, the FDA rated more than 80% of all new drugs, over the past 12 years, as offering little or no therapeutic gain over existing ones.[16] In these crowded, inexpensive therapeutic areas, where clinical results are comparable, the only way to distinguish a drug is by demonstrating a cost-effectiveness advantage.[17] This is because, in many instances, the only difference between a new drug and existing one is a reduction in side effects or an improvement in the manner in which the drug is administered. Drug cost-effectiveness studies can determine whether such supposedly minor improvements give a drug a competitive edge over another, equally effective but cheaper drug.[18] For example, cost-effectiveness research may determine whether a new drug, whose only advantage over existing drugs is that its dosage is administered once rather than three times a day, is in fact more cost-effective, despite its higher costs.

Drug manufacturers are also using cost-effectiveness techniques to price their products and to distinguish between economic winners and losers as early as during Phase I testing.[19] At the same time, venture capital firms and investment companies, who provide early stage funding for bio-tech firms, are pushing for more information about the probability of a market niche for the products being developed, based on cost-effectiveness information.[20] Finally, the recent emphasis on cost containment has sparred lively activity in the academic field, some of which is funded through grants from the NIH and some of which is conducted through the sponsorship of the drug industry.[21]

The proliferation in cost-effectiveness research suggests that such information will at least be available for drug selection decisions; structural changes in the health-care industry make it central in such decisions. Traditionally, physicians determined which drugs to prescribe to their patients. They did so on the basis of their knowledge or perception of the efficacy of the available drugs. Physicians derived much of their knowledge and perceptions from information provided by drug manufacturers, either through direct contact with drug sales representatives or through other promotions and advertisements. Thus, this exchange of information influenced significantly physician prescribing behavior.[22] In fact, there has been a great deal of concern that drug companies, through a variety of promotional schemes, exercise undue influence on physician prescribing behavior. [23]

Drug distribution in the United States, however, is rapidly changing. Health Maintenance Organizations (EMOs) and other managed care entities are growing the drug benefits management business. By managing drug benefits for employers and other groups, the managed care organizations are increasingly making decisions about the selection of prescription drugs, determinations once made exclusively by physicians. It is estimated that 50 million Americans are enrolled in managed drug benefits programs, and drug makers expect this figure to double to 100 million by 1997.[24]

Managed care entities establish a formulary, a list of drugs from which participating physicians are expected to prescribe, by determining the appropriate medications for a range of diagnoses. In order to reduce their costs, managed care companies are increasingly relying on cost-effectiveness information to determine which drugs would enter their formularies. Although, in some instances, managed care companies conduct their own research, they often pressure drug companies to provide such information by making it a condition to getting drugs on their formularies.[25] Once the formulary is established, managed care entities' task is to encourage physicians to prescribe from the formulary; a task they take very seriously. Managed-care companies attempt to insulate physicians from external influence by restricting the access of drug sales representatives to physicians. [26]A recent study suggests just why they do so.[27] It found that physicians who belong to managed care programs are more likely to request additions to formularies if they have greater interaction with drug sales representatives,[28] suggesting that such interaction also causes them to prescribe outside the formulary. Moreover, the study found that the perception of physicians who interacted with drug companies about the performance of drugs matched closely the claims of drug companies, regardless of the drugs' true performance.[29] Thus, by limiting the access of drug sales representatives, managed care companies ensure that prescribing decisions are made on the basis of their original cost-effectiveness decisions. Just how successful are managed care companies? Kaiser Permanente, a large California based 1-1MG, manages to get its 9,000 physicians to fill 99% of prescriptions from its approved formulary. Thus, in this new world, cost-effectiveness considerations are at the heart of every prescription decision.

A large segment of the population, however, remains outside managed care. The physicians in this segment are not constrained by formularies, but rather rely heavily on information they receive from drug manufacturers. These physicians do not have information about drug costs or cost effectiveness unless it is provided by drug manufacturers. Indeed, a recent study finds that most physicians are not aware of drug costs.[30] This is likely to change! As drug companies beef-up cost effectiveness claims in their promotion campaigns, physicians will increasingly become more aware of such information. Moreover, as more physicians interact with managed-care companies, more information will permeate the indemnity world, because many physicians participate in both systems. The two sources of information are different, however. Drug companies' claims are likely biased, tailored to promote the use of their drugs. Information provided by managed care companies, on the other hand, is screened for such bias. This bifurcated system seems in desperate need for some standardization.

IV.PROPOSED REGULATION OF COST EFFECTIVENESS STUDIES

A.The Current Regulatory Void

The availability of cost-effectiveness information should be welcomed by industry participants and regulators alike. The information makes the selection of drugs a much more sensible and systematic process. However, there are several problems with this explosion of information that have yet to be tackled by the FDA. First, there are no standards or common guidelines for performing cost-effectiveness studies, questioning the reliability of such information, nor is there agreement on whether and when such studies should be undertaken. Second there are no guidelines as to how cost-effectiveness information should be promulgated, an issue of particular importance because of the ease in which such information is manipulated. Finally, there is a tremendous pressure by the managed care industry on drug companies to perform cost-effectiveness studies for unapproved uses of approved drugs. The FDA needs to decide how to address these potentially useful infractions.

So far the FDA has largely refrained from addressing the issue of cost-effectiveness. In a recent speech, David Kessler, the commissioner of the FDA, emphasized that the FDA does not want to become involved in making cost decisions as a basis for drug approval and that the most appropriate role for the FDA in controlling costs is to make sure that approval is only extended to drugs that work and for the FDA to progress systematically on the approval of generic drugs.[31] However, the commissioner conceded that the FDA is being drawn into the cost-effectiveness research field as a result of its regulation of prescription drug advertising.[32] Indeed, recently FDA representatives indicated that they intend to develop policies and white papers to more clearly define FDA's position on pharmacoeconomic issues.[33] In reality, the FDA has. already gotten involved in this area by issuing warning letters to several manufacturers who have made misleading cost-effectiveness claims in their advertising.[34]

B.Potential Approaches to Regulating Cost Effectiveness Studies

1.The Opposite Extremes: No Regulation vs. Cost Effectiveness as a Condition to Drug approval

The FDA may employ several strategies in response to the proliferation in cost-effectiveness information. It can choose to stay entirely out of the field, a posture consistent with FDA's traditional role, limited to evaluating the efficacy and safety of drugs. On the other extreme, recognizing the importance of cost-effectiveness information, the FDA may choose to condition the approval of new drugs on passing a certain threshold of cost-effectiveness. For example, the FDA may deny approval if a new drug does not at least match the cost-effectiveness of the most prevalent available treatment.

Both of these strategies are hardly feasible. The FDA cannot remain entirely out of the field because drug companies are increasingly including cost-effectiveness information in their promotion materials.[35] The problem is that every comparative claim is at heart a cost-effectiveness claim. Thus, the FDA must at least develop some standards for determining when such claims are misleading. Ignoring cost-effectiveness claims entirely would mean relinquishing all control over drug promotion, a stance the FDA is hardly likely to take. At the other extreme, conditioning drug approval on its cost-effectiveness performance is hardly workable either. Drug cost-effectiveness varies between populations, geographies and parameters of illness. Even drugs that are not cost-effective overall have some uses that are cost effective, and therefore should not be denied approval. There is also a practical difficulty in this approach. Cost-effectiveness studies, done prior to drug approval, are of lesser utility because they are limited to a highly controlled population, which often fails to mimic the real world operation of the drug.[36] Therefore, drug approval is, to a large extent, a prior condition to obtaining useful cost-effectiveness research. Finally, results of cost-effectiveness studies are highly malleable, and don't lend themselves to the on-off decisions required in the approval process. [37]

2.The Middle Ground Strategies

This leaves the FDA with three potential strategies. The FDA may establish guidelines for performing cost-effectiveness studies, without mandating them, using these standards to evaluate cost effectiveness claims when they are made by drug manufacturers. Alternatively, along with promulgating standards, the FDA may require drug companies to provide cost-effectiveness information within a certain period of approval, but without conditioning approval on the results of the studies.[38] Finally, the FDA may choose to establish an independent body, a cost effectiveness research institute, that would conduct cost-effectiveness research and make it available to the public.

a.Mandatory Testing by the Drug Companies or by an Independent Research Institute

The second and third alternatives, of either mandating cost effectiveness tests or forming a central research institute that would conduct such studies, have advantages and drawbacks. Both would ensure that cost effectiveness information will become available rapidly for all new drugs and possibly for existing ones. The third approach also adds a measure of integrity by having a disinterested body conducting the testing. Both, however, will substantially disrupt the natural market forces that have caused such information to become available in the first place, and will therefore disturb the inherent efficiencies in this process.

Cost-effectiveness studies and how to apply them is an evolving discipline. Industry is just now feeling its way around the field, increasing gradually the resources in the area, examining closely how to allocate these resources and experimenting continuously with new and innovative techniques. Mandating the immediate testing of every new drug by either drug companies or by a central entity will deprive the discipline from the continuous innovations that it is experiencing. In fact, it could be argued that the process, as it is naturally occurring, is already too rapid to absorb all the available innovation because of the unrelenting pressure for cost containment from the Clinton Administration. Moreover, rapidly introduced information may not sit well with physicians and patients alike. Because the information is novel and complex, physicians need to become knowledgeable and comfortable using it. Rapid promulgation of information may backfire by flooding physicians with information that they cannot process. Gradually phased-in mandatory testing may alleviate such concerns, but the question then becomes such a system compares to the gradual way in which the market introduces information. This point is discussed below.

And let's not forget costs. Universal testing is cost-effective if the benefits from testing, the savings derived from using the superior drugs, exceed the costs of testing. While universal testing may be superior to no testing at all, it likely falls short in comparison to targeted testing, which is in essence the market driven approach.[39] The relevant question is whether under market forces, every cost-effective cost-effectiveness study will be carried out, or whether regulatory intervention is necessary to ensure that testing occurs. This question will also be addressed below.

b.FDA Promulgating Standards for Cost-Effectiveness Studies without Requiring Testing

This leaves the standard-setting approach, which sets guidelines for studies without mandating them. This is the most realistic approach given the current FDA stance. This approach allows the market to determine what studies would be undertaken, without imposing the burden of testing every new drug or a subset thereof Drug companies will commission such studies either at the behest of managed-care companies who will continue to pressure for such information, or voluntarily for the purpose of promoting their drugs.

(1)Effect on Innovation

Setting standards may have also have an innovation-stifling effect. By promulgating standards, the FDA runs the risk of freezing the state of the art of cost-effectiveness technology at where it stands today. This is because drug companies will comply with all FDA regulations in order to ensure that they will be able to use their research and therefore stop experimenting with new approaches. Thus, the task of promulgating standards is an extremely delicate one. It needs to promote the integrity and consistency of information while, at the same time, providing for enough flexibility so as not to discourage experimentation with new methodologies and practices. This can be achieved by phasing in regulations gradually, starting with regulations that merely target clear abuses and, as the discipline matures, phasing in stricter guidelines that feature the best practices in the industry. The FDA will have to decide how to deal with old studies conducted under more lenient standards which do not meet later phased in standards. Allowing the continued use of such non-conforming studies will provide drug companies a strong incentive to test early. Proscribing the use of such studies will slow down research because drug companies will either wait or try to live up to the higher standards before they are promulgated, thereby also slowing-down research. This on-off switch may be a useful incentive scheme with which the FDA can control the pace of research.

Another way to avoid stifling innovation is through adjusting the standard used to determine non-compliance. It's useful to distinguish between two types of violations: ones that relate to research and statistical methodology, which are most likely to impact innovation, and ones that relate to the processing of information and the integrity of the assumptions,[40] which are less likely to affect innovation. By pursuing an aggressive standard of compliance against the later violations, and a more lenient one against the former, the FDA will avoid deterring innovation in the area in which it is most likely to bloom. Thus, for example, the FDA may sanction a drug company whenever there is a material deviation from the standard procedure for information collection, but only sanction deviations in the methodologies of cost-effectiveness studies when it finds a specific intent to mislead.

(2)The Increased Cost of Conducting Research

Strict guidelines will also increase the cost of testing, which may deter drug manufacturers from conducting studies. The problem of increased costs is central. If FDA standards are too burdensome, they will disrupt the current incentive scheme. Cost-effectiveness claims may become too expensive an advertising tool, prompting manufacturers to shift their advertising dollars elsewhere. A classic example of a guideline that may make studies more costly relates to the level of proof required. If the FDA requires that the results be stated with a 95% confIdence level, when current research is conducted to conform to a 70% confIdence level, the costs of new studies will increase substantially because they will require expanded research populations and more refined research techniques.

Arguably, managed care companies will still demand such information, ensuring that some cost effectiveness research will continue to come out. But, if the regulations are extremely costly, drug manufacturers will conduct non-complying research and share it only with managed care companies, so as to avoid detection by the FDA. Such limited distribution of research will become more viable as a greater portion of the market belongs to managed care, because public advertising and direct promotion to physicians will become less important. This will be unfortunate because it will deprive the information from the non-managed-care segment and may cause duplicate research to be undertaken for different managed care companies. Thus, it is crucial that FDA guidelines not be too burdensome. This means that FDA should attempt to standardize current practices rather than imposing stricter and costlier methodologies that will have an unpredictable impact on the system..

Guidelines may increase the cost of research in a more sinister way: they may limit the ability to make cost-effectiveness claims about inferior drugs, thereby, ex-ante, increasing the expected cost of each cost-effectiveness claim.[41] This means that by succeeding in its objective to limit misleading cost-effectiveness research, the FDA may deter drug companies from conducting such research. This is no reason to loosen standards, but it should alert FDA to the fact that even standards that merely enforce mainstream industry practices may increase the cost of research. It suggests that in order to maintain the current flow of information, the FDA may have to counter the damping effect of the new guidelines with incentives that will induce conducting such research.

If the standards are too complex, so that compliance is not easily ascertained by researchers, drug companies will turn to the FDA before conducting such research, so as to limit their expenditures on unusable research. This will create a sort of a prior approval regime, which may increase the cost of conducting research and therefore decrease the availability of information. The FDA must choose the penalties and the standard for non-compliance while being mindful of this possibility. A good-faith approach that would allow the possibility of remedying defective studies if the noncompliance is not intentional will reduce the possibility of creating a de-facto prior approval regime. Will such a system provide for enough deterrence? The FDA may have to ratchet-up the penalties if the drug company fails to meet the good faith standard, so as to provide sufficient deterrence against intentional violations of the standards.

(3)Areas in Which Studies May Not Be Conducted

The standard setting approach may have a number of additional problems that relate to the types of studies that will be conducted and the nature of the information that will be disclosed. First, if drug companies are not required to provide such information, they would only do so when it favors their own drugs. Thus, one would expect that cost effectiveness studies will always find that the sponsor company's drug is superior. At the same time, drug companies will suppress results that show their drug to be inferior. Second, drug companies will first test drugs that have a substantial impact on their bottom-line, delaying research in areas in which there is a greater disparity between drugs' cost-effectiveness but less of an impact on drug companies' bottom-line. Third, some drugs will never be tested if the cost of conducting the research exceeds the drug company's returns from testing. This is likely to occur in areas where drug patents have expired. Finally, when only one drug is available for a particular illness, its manufacturer will have no incentive to test the drug because of the natural monopoly it already possesses.

The first concern, that drug companies will only promulgate information that demonstrates their drugs' superiority is not as problematic as it first appears. Even if only the superior drug in any group of related drugs is tested, the study will point at the most cost-effective course of treatment.[42] If available information supports the use of a drug that is not superior, a result that may arise when a cost-effectiveness study does not compare the tested drug with all the relevant treatment alternatives,[43] the manufacturer of the superior drug will have powerful incentives to conduct its own research in order to demonstrate its drug's superiority. The problem therefore is not that the system will fail to identify superior drugs, but rather that there may be short-term inaccuracies in the system and that some information will take time to come out. Thus, as long as FDA guidelines are consistent and comprehensive enough to ensure that when two drugs are tested the results would indicate which is truly superior, the system will always move in the direction of improving the availability of cost-effectiveness information. Admittedly, short-term inaccuracies can be costly, but such costs don't seem significant enough to justify universal testing.

The concern that drug companies will first test the drugs that can potentially contribute most to their bottom line, neglecting at first areas in which there may be a greater disparity between the cumulative effectiveness of the available drugs, is legitimate. From a societal perspective, of course, it is more efficient to generate information first in areas of greater disparity in cumulative effectiveness. Of relevance here is the sequence or the timing of testing, and not whether testing would occur at all. This ties to an earlier discussion about gradually-phased mandatory testing. The question is whether this market inefficiency in the order in which testing will be conducted justifies FDA involvement in mandating testing. Since there are not guarantees that the FDA can prospectively identify which areas have the greatest disparity, the answer is probably no. At least, under market forces, we can be sure that managed care companies will apply pressure to test precisely in areas where large disparities are expected.

That some treatment areas will never be tested is also a legitimate concern; drug companies will only test when their expected return from testing exceeds testing costs. So, for example, in areas of expired patents no drug company will be willing to test. Even under pressure from managed care companies, drug companies will not invest in testing unless they expect to recoup their investments. Thus, some testing that would be efficient from a societal perspective will not occur. In this narrow area, universal testing is preferable to mere standard setting. But instead of mandating testing of all drugs, the FDA should target these easily detected areas through incentive for testing or through direct funding of research. Moreover, as managed care companies grow larger, they will routinely commission cost-effectiveness studies,[44] supplementing the information provided by the drug industry.

Finally, the concern that information will not be available for illnesses for which there is only one available drug treatment is not justified. In such areas, drug companies will still seek to prove that use of their drug is more cost-effective than other, non-drug-based treatments. However, if no other course of treatment exists, drug companies may be content to promote their drug on the basis of its efficacy, rather than its cost-effectiveness. Here, too, a combination of pressure from managed-care companies and governmental incentives and funding may alleviate the problem

Thus, it seems that a regime in which the FDA promulgates standards for the performance of cost-effectiveness studies, without mandating them, may be a workable one. The above analysis, however, demonstrates the sensitivity of the current market equilibrium, arguing for a gradual and cautious FDA intervention The above analysis also identifies the problem areas, which justify specific FDA incentives to promote research.

C.Regulating Cost Effectiveness Statements

Cost effectiveness information is easily manipulated. The results of such studies can be stated in ways that promote use of the tested drug, even when it is not the best drug in the market. Thus, it is important that FDA guidelines regulate the types of statements that drug companies can make about their drugs. This will prevent drug companies from misleadingly promoting their drugs in advertisements and drug labels, but it will not prevent detail-men, drug sales representatives, from manipulating the information when they interact directly with physicians.

To prevent this, the FDA may require drug companies who intend to make cost-effectiveness assertions about their products to compile a standardized fact sheet that will include a summery of the results of the cost-effectiveness study. FDA may require that the fact sheet either be incorporated into the drug label or that it be provided whenever any drug representative makes cost-effectiveness claims about a product. In either instance, individual physicians would have accurate information against which to examine the claims of detail-men.

D.Regulating Cost-Effectiveness Information about Off-Label Uses of Approved Drugs

Drug manufacturer have come under intense pressure from managed care companies to conduct and share information about the cost-effectiveness of various unapproved uses of approved drugs. Under Section 202. 1(e)(4) of the agency s advertising regulations, manufacturers cannot recommend or suggest any use that is not in the labeling of an approved NDA. The FDA has provided an exception to this rule in Section 312.7(a) of the JIND regulations, which states that the restriction of such statements in a promotional context "is not intended to restrict the full exchange of scientific information." While the FDA tolerates off-label uses as acceptable medical practice, it is clearly not ready to relinquish its role in preventing manufacturers from promoting off-label uses of their drugs. The exchange of information between a drug company and a managed care entity is promotional in nature; it is intended to further the use of the drug by the managed care company. Thus, such practices are clearly antithetical to FDA's regulation.

Off-label prescribing is an undisciplined practice, whose benefits and risks need to be assessed. It seems that cost-effectiveness studies may provide precisely the type of information that will prevent reckless prescribing of unapproved uses, and managed care companies' insistence on such information represents a measure of responsibility that should be lauded rather than restrained. In effect, managed-care companies' pressure has countered the traditional incentives that led drug companies not to seek approval for off-label uses. Traditionally, drug companies did not seek approval because, even without approval, they reaped the benefits of off-label prescribing, which left no incentives to expend costs on approval. But by pooling together the drug purchasing power of a large population and conditioning purchases on the availability of such information, managed care companies have injected enough of an incentive for such testing.[45] Drug companies and managed care companies are together creating an externality which benefits the rest of the public in the form of improved information about off-label uses. But, if the FDA maintains its stance against such practices, treating them as promotional in nature, drug companies may be reluctant to perform such studies and even when they do perform them they would limit disclosure to the managed-care company so as to avoid detection by the FDA.

Thus, a policy that strictly enforces the rule against promoting off-label uses, as it relates to the sharing of cost-effectiveness research, makes little sense. On the other hand, a policy that allows drug companies to make any claims about off-label uses, as long as they embody cost-effectiveness information, would neutralize the effect of the anti-promotion rules. The FDA may choose to condone the exchange of cost-effectiveness information only if it conforms to the guidelines for performing and disseminating cost effectiveness research, either the guidelines discussed above or stricter ones that will reduce the likelihood of errors. Such an approach would eliminate reckless promotion of unapproved uses, and at the same time prevent the suppression of useful information

But allowing the circulation of cost effectiveness claims is akin to a secondary use approval on the basis of cost-effectiveness tests.[46] This may not be as bad as it first sounds because these drugs have already passed all safety tests in the original approval process. Indeed, the FDA may want to consider allowing properly conducted cost-effectiveness studies to form the basis for formal approval of off-label uses. The risk, of course, is that FDA guidelines will make such cost-effectiveness research prohibitively expensive, deterring the cooperation of drug companies, who will secretively continue to conduct research under less stringent guidelines. Thus, the FDA must tread a fine line between ensuring that the tests are comprehensive enough to prevent reckless promotion of off-label uses, but not too burdensome so as to prevent the cooperation of drug companies.

The above analysis leaves out an important problem: drug companies will never reveal information about off-label practices that are not cost-effective because they wouldn't want to deter their continued use. This is not significant in the long term because at some point testing will be so prevalent so that an off-label use about which there is no information will be presumed to be not cost-effective. In the short term, the only way to ensure the information will come out is through compelled disclosure.[47] Such a requirement may not be wise however since it will further deter drug companies from openly testing off-label uses.

V.CONCLUSION

The explosion in cost-effectiveness research leaves the FDA no choice but to set some guidelines in the area. But the FDA must recognize that cost-effectiveness information is a valuable commodity, and that misguided regulations may block the flow of information and stifle innovation in this still evolving discipline. It should try and standardize current practices, targeting abuses and clear manipulations of information, rather than creating entirely new norms of conduct. The FDA should not forget, however, that an increasingly large portion of drug selection decisions are made on the basis of cost-effectiveness information. Therefore, its primary concern is to ensure that the such information is accurate and reliable.

---

[1] Greg Muirhead, Pharmacoeconomics: A Still-Fuzzy Buzzword, Drug Topics, May 9, 1994 at 74. 2 Robert J. Beck, How to Evaluate Drugs: Cost-Effectiveness Analysis, 264 JAMA 83 (1990).

[2] Id.

[3] See Muirhead, Supra note 1.

[4] Alan L. Hiliman et al., Avoiding Bias in the Conduct and Reporting of Cost Effectiveness Research Sponsored by Pharmaceutical Companies, 324 New Eng. J. Med. 1362, 1363 (1991).

[5] See Beck, supra note 2.

[6] Gerold Stucki et al., Is Misoprostol Cost Effective in the Prevention of Nonsteroidal Anti- Inflammatory Drug-Induced Gastropathy in Patients with Chronic Arthritis? A Review of Conflicting Economic Evaluations, 154 Archives of Internal Med. 2020 (1994).

[7] Id.

[8] Id

[9] Id.

[10] Charles Marwick, Is a Drug worth Its Cost, 269 JAMA 18 (1994).

[11] Id.

[12] Industry Chiefs Slammed Pharmacoeconomics, Marketletter, August 15, 1994.

[13] Id

[14] Id.

[15] Id.

[16] Andera Rock, Cut Your Spiraling Drug Costs 70%, Money, June 1, 1993 at 130. '⁶Id

[17] Paul Freinian, Let's Use Pharmacoeconomics to Develop Economic Winners, Med. Marketing & Media, July 1994 at 18.

[18] See Beck, supra note 2.

[19] Freinian, Supra note 3.

[20] The Pink Sheet, 55(44) 8-9, November 1, 1993.

[21] See Hiliman, supra note 4 at 1363.

[22] Drug Promotion: Stealth, Wealth and Safety, 341 Lancet 1507 (1993).

[23] David A. Kessler Ct al., Therapeutic Class Wars - Drug Promotions in a Competitive Market-Place, 331 JAMA 1350 (1994).

[24] David R. Olmos, Drug Makers Can't Beat 'Em So They Join Managed Care, Los Angeles Times, May4, 1994 atDl.

[25] The Pink Sheet, 55(44) at 8-9, Nov. 1, 1993.

[26] George Anders, Rxfor Sales: Managed Health Care Jeopardizes Outlook For Drug Detailers, The Wall Street J., September 10 1993 at Al.

[27] Mary M. Chren and Seth C. Landefeld, Physicians' Behavior and their Interaction with Drug Companies, 271 JAMA 684 (1994).

[28] Id

[29] Id.

[30] Lucinda G. Miller and Alan Bluni, Physician Awareness of Prescription Drug Costs: A Missing Element of Drug Advertising and Promotion, J. of Family Practice, Jan.,, 1993 at 33.

[31] The Pink Sheet, 55(44), November 1. 1993 at 8-9

[32] Id

[33] The pink Sheet, 56(39), Sept. 26, 1994 at 8.

[34] In a recent warning letter, the FDA's Division of Drug Marketing, Advertising and Communications challenged Lilly's claim that Axid, a prescription drug, is the "cost effective choice for healing reflux esophagitis," because the statement was based on a comparison of acquisition costs only, without factoring in clinical efficacy. See The Pink Sheet, 56(30), July 25, 1994 at 3-5. Similar letters have been sent to Knoll and Berlex. See The Pink Sheet, 55(44) Nov. 1, 1993 at 8-9.

[35] Seesupranote34.

[36] The Pink Sheet, 56(24) T&G-3, June 20, 1994.

[37] I do not recite the more obvious problems of further delaying the approval process and increasing the cost of approval, which may deter innovation, in the text.

[38] In Europe and Australia, economic research is often required for government approval and for the pricing of new drugs. See Hillman, supra note 4.

[39] This assumes that in some areas cost-effectiveness research is not cost-effective, either because the results are obvious and do not require testing or because the differences between treatments are so marginal that the benefits from testing would not off-set the costs.

[40] Such tactical manipulation that impact the results of studies but have little to do with innovation are: failure to discount items, excluding subjects that do not respond to treatment, and using charges rather than the full cost of treatment. See Hillnian, supra note 4.

[41] This is because under the unregulated world drug companies may always make a cost-effectiveness claim, even when the tested product is inferior, by manipulating the results of their studies. However, under FDA standards, which 14i11 reduce such manipulation. they will not be able to make assertions about inferior drugs. Thus, drug companies, on average, will have to conduct more than one study per each set of cost-effectiveness claims about a particular drug, assuming that drug companies cannot perfectly predict, ex-ante, which drugs will prove to be superior. The cost of conducting research may further be increased if the FDA mandates disclosure of all cost-effectiveness studies. Under such a regime, a drug company that conducts a test that does not find its drug superior will not only forego the costs of the study, but will also have to suffer the negative impact of having its research promote a competing product (the additional costs are the losses in sales incurred as a result of promoting the competing product for the period between the date of compelled disclosure and the date in which it is expected that the superior product will be tested by its owii manufacturer). Of course, there may be questions as to whether such compelled disclosure is practically enforceable, particularly because drug companies will abandon studies before completion if they expect unfavorable results. Interestingly, such compelled disclosure may be operating in practice when drug companies sponsor academic research and the recipient universities insist on publishing the findings regardless of the results.

[42] Drug companies may predict which drugs will be yield superior results and proceed to test them. If an infenor drug is mistakenly tested and the results so indicate, drug companies will try to suppress the research. While information about the superior drug may eventually be promulgated when it is tested by its manufacturer, having the results earlier by compelling the promulgation of all cost-effectiveness tests is more efficient. See discussion of whether such a guideline is feasible and how it may increase the cost of testing, supra note 41.

[43] Notice that the standards may take care of this problems by providing that the selection of drug against which the tested drug is to be compared is based on clinical relevance and not on the potential favorability of the results. For example, the FDA may insist that at least the comparisons required by the FDA for controlled clinical trial be included in cost-effectiveness studies. See Hillman, supra note 7.

[44] Presumably, there is a threshold beyond which a managed care company becomes large enough to justiI~r conducting its own cost-effectiveness research.

[45] In reality, while the drug companies are paying directly for the studies, the managed care companies are assuming some of the costs in the form of higher prices.

[46] Only cost-effective unapproved uses will be de-facto approved under this system because drug companies will suppress information about inferior unapproved uses.

[47] The FDA may set a prior approval regime, in which drug manufacturers sign-up for approval before they conduct their research, so that the FDA will be able to enforce the compelled disclosure.