Which Orphans Will Find a Home? The Rule of Rescue in Resource Allocation for Rare Diseases

Abstract

The rule of rescue describes the moral impulse to save identifiable lives in immediate danger at any expense. Think of the extremes taken to rescue a small child who has fallen down a well, a woman pinned beneath the rubble of an earthquake, or a submarine crew trapped on the ocean floor. No effort is deemed too great. Yet should this same moral instinct to rescue, regardless of cost, be applied in the emergency room, the hospital, or the community clinic?

In health care, the desire to save lives at any cost must be reconciled with the reality of resource scarcity. As one example, the estimated cost for prophylactic Factor VIII to treat one patient with hemophilia for one year is $300,000. Costs of this magnitude have been accepted by public and private insurers in the developed world, even though, in principle, these sums could provide greater overall health benefit if allocated to pay for the unmet health care needs of many other patients. Looking forward, however, broad application of the rule of rescue will be increasingly untenable. But the moral instinct will remain: the desire to help those weakest among us, especially when their small numbers allow us to see them as unique individuals. What, then, is the ethical framework that can guide coverage and reimbursement decisions for orphan drugs into the future?

Article

I call this … the rule of rescue. Our moral response to the imminence of death demands that we rescue the doomed. We throw a rope to the drowning, rush into burning buildings to snatch the entrapped, dispatch teams to search for the snowbound. This rescue morality spills over into medical care, where our ropes are artifcial hearts, our rush is the mobile critical care unit, our teams the transplant services. —Albert Jonsen[1]

The rule of rescue describes the moral impulse to save identifable lives in immediate danger at any expense. Think of the extremes taken to rescue a small child who has fallen down a well, a woman pinned beneath the rubble of an earthquake, or a submarine crew trapped on the ocean floor. No effort is deemed too great, and seemingly unlimited resources are expended to save these individual lives. To do less would be deemed unconscionable. Yet should this same moral instinct to rescue, regardless of cost, be applied in the emergency room, the hospital, or the community clinic? In health care, the desire to save lives at any cost must be reconciled with the reality of resource scarcity.

The treatment of patients with rare, or “orphan,” diseases raises persistent ethical dilemmas of this nature. In America, an orphan disease or condition is defined by law as one that affects fewer than 200,000 people.[2] Six to seven thousand different rare diseases have been identifed, and approximately five new rare diseases are described in the medical literature every week.[3] Orphan diseases are mostly chronic, and many are life-threatening or signifcantly life-altering. Familiar examples include Huntington disease, amyotrophic lateral sclerosis (commonly known as Lou Gehrig disease), and cystic fibrosis. Drugs developed to treat orphan diseases are known as orphan drugs and are usually extremely expensive. As one example, the estimated cost for prophylactic Factor VIII to treat one patient with hemophilia for one year is $300,000.[4]

Costs of this magnitude to provide orphan drugs for individual patients have been accepted by public and private insurers in the developed world, even though, in principle, these sums could provide greater overall health beneft if allocated to pay for the unmet health care needs of many other patients. In addition, it is clear that paying as much per patient to treat conditions affecting large numbers of patients would be dismissed out of hand as financially impossible. In practice, therefore, orphan drugs are treated as special cases, exempt from the general framework that insurers use to weigh evidence on benefts and costs in making allocation decisions for health care services.

Why do insurers agree to pay extremely high prices for orphan drugs? One reason is the logic that manufacturers are forced to seek high prices per treatment in order to recoup the cost of their investment in the development of these drugs, which will only be used for a relatively small number of patients. Another reason, frequently cited, is the desire to offer some kind of treatment to help identifable, vulnerable patients—the rule of rescue. Satisfying this desire, even at extremely high prices, has been possible in the past because orphan drugs have a limited impact on overall health care spending. Even astronomical prices, if paid for only a few patients, will not divert a large proportion of available health care resources from uses that could help other patients.

Looking forward, however, the conditions that have allowed insurers to support full expression of the traditional rule of rescue will no longer hold true. Increasing numbers of expensive orphan drugs are expected to come to market. In addition, advances in pharmacogenetics will soon be able to separate many common diseases, such as hypertension, arthritis, cancer, and diabetes, into numerous small and distinct subpopulations of patients with specifc genetic profles. This movement toward “personalized medicine” will produce increasing numbers of drugs that have been developed to treat small, identifable patient groups. Instead of a new blockbuster drug to treat millions with hypertension, new targeted therapies will treat only those few thousand with a particular genetic makeup. As the eligible patient populations decrease in size, the arguments for accepting higher prices per treatment will turn manageable overall costs into unsustainable ones.

Under this future scenario, broad application of the rule of rescue—willingness to pay extremely high prices to help save individuals in need—will be increasingly untenable. But the moral instinct will remain: the desire to help those weakest among us, especially when their small numbers allow us to see them as unique individuals. What, then, is the ethical framework that can guide coverage and reimbursement decisions for orphan drugs into the future?

The purpose of this article is threefold: 1) to describe the historical and political circumstances that have shaped orphan drug policy to date and to explain in greater detail why blanket coverage at “market” prices will not prove sustainable; 2) to outline and deconstruct the argument from the rule of rescue that is made in support of coverage of orphan drugs at extremely high prices; and 3) to draw on this analysis in order to suggest factors that should be incorporated into coverage and reimbursement decisions for orphan drugs and other narrowly targeted therapeutics so as to make these decisions more transparent and ethically sound.

Orphan Drugs, Costs, and Concerns

Rare diseases affect an estimated twenty-fve to thirty million Americans (8 percent to 12 percent of the population), but the relatively small number of patients with any particular disease has made rare disease patients a historically unattractive market for pharmaceutical manufacturers.[5] Small numbers of patients and the possibility of low return on investment are disincentives for medical researchers and pharmaceutical companies to make the signifcant investment needed to develop therapies specifcally for rare diseases.

In the late 1970s, several American rare disease organizations united to address the need for medical research and translation of that research into drugs and therapies for their members. In order to raise public awareness, explained Abbey Meyers, then-president of the National Organization for Rare Disorders, the rare disease coalition “embarked on a public relations campaign, bringing television cameras and newspaper reporters into the laboratories where recovered patients helped to fill capsules with lifesaving medicines that drug companies refused to manufacture.”[6] In response to patient advocacy and growing public pressure, the United States government passed the Orphan Drug Act of 1983 to counterbalance market forces and incentivize orphan drug development. In the decade before the Orphan Drug Act was passed, only ten new drugs for rare diseases were developed. Twenty-fve years after the act's passage, orphan products represent roughly one-third of all newly approved drugs and bio-logics. More than 1,100 new orphan treatments have entered the research pipeline.[7]

When they reach market, orphan drugs tend to be expensive. Gaucher's disease, for example, is a rare inherited metabolic disorder affecting 2,500 Americans. Cerezyme, an enzyme replacement therapy for Gaucher's disease, effectively enhances quality of life for most patients but is priced at $200,000 or more per year and must be taken for life.[8] Prices at this level can be explained in terms of three primary factors. First, development and production costs need to be offset. By current estimates, it takes an average of $1 billion to bring a new drug to market, and only a handful of patients may consume a particular orphan drug.[9] This renders development costs proportionately higher for orphan drugs. Second, the monopoly position of manufacturers is entrenched within legislation and permits large proft margins. Third, even after market exclusivity expires, orphan products generally experience less generic competition—which would drive costs down—than other new drugs and biologies.[10]

Orphan drugs historically have been covered by public and private health insurers, and most patients have access to even the most expensive orphan drugs. Now, however, numerous stakeholders are closely scrutinizing the high prices paid for orphan drugs.[11] For public and private insurers, the expense of providing orphan drugs to all eligible patients may become prohibitive as the absolute number of patients with rare diseases continues to rise, more orphan drugs become available, and these drugs occupy larger fractions of health care budgets. For many rare disease patients and their families, the high copayments often required for orphan drugs present great financial burdens, and individual policyholders can feel pressure to leave their small employers when the cost of their own or a child's orphan drug render other employees’ insurance premiums unaffordable. In addition, given the high price of orphan drugs, the American health system may increasingly face real trade-offs in which payment for orphan drugs requires limits to benefts for patients with more prevalent diseases. From numerous perspectives, therefore, the historically accepted model of coverage and reimbursement decisions for orphan drugs—a model strongly guided by the rule of rescue—no longer appears sustainable.

Rescue versus Cost-effectiveness

Orphan drug coverage decisions highlight the tension that can arise in health care between doing the most good possible with scarce health care resources and the desire to assist identifable individuals regardless of cost. This may also be described as tension between utilitarianism and the rule of rescue. Utilitarianism generally holds the view that the morally right action is the one that maximizes the overall good. The rule of rescue has not yet achieved a comparably consistent (or concise) meaning. For example, Erik Nord defined it as “the sense of immediate duty that people feel towards those who present themselves to a health service with a serious condition,” whereas Molly Osborne and Timothy W. Evans defined it as “the powerful human proclivity to rescue a single identifed endangered life, regardless of cost, at the expense of many nameless faces who will therefore be denied healthcare.”[12] Despite this variability, the consistent core of the rule of rescue is, as Charles Fried tells us, “the apparent anomaly that we are prepared to expend far greater resources saving the lives of known persons in present peril, than we are prepared to devote to measures that will avert future dangers to persons, perhaps unknown and possibly not yet even in existence.”[13]

The rule of rescue is in obvious confict with utilitarianism as a principle for allocating limited resources within an insurance pool or health system. If cost-effectiveness analysis, a policy tool refecting utilitarian goals, were applied in an undiluted fashion to coverage decisions for health care in order to maximize health gains, rare disease patients would receive no special preference, and many orphan drugs would not be covered. Because it does not account for feelings that priority should be accorded to those identifed as the neediest amongst us—among other reasons—cost-effectiveness analysis has been criticized as overly stark and unpalatable. Conversely, coverage and reimbursement decisions guided by the rule of rescue are criticized for failing to maximize health as an aggregate measure across an entire population. If resources are limited—and implicitly, if not explicitly, they always are—then strict adherence to the rule of rescue forces trade-offs that deny resources to some patients who, in the aggregate, could beneft more.

There are at least three constituent parts of the rule of rescue—identifiable individuals, endangered lives, and opportunity costs. A productive starting point for an analysis of the rule is to try to elucidate the extent to which and ways in which these are ethically relevant to orphan disease coverage decisions. The aim is to find a way to respect the strong moral impulse captured by the rule of rescue while taming it so as to achieve a sustainable coverage policy for orphan drugs and other therapies for small numbers of patients.

Identifiable Individuals

Identifability is central to all definitions of the rule of rescue. In the context of orphan diseases, identifiability is usually considered to be a direct consequence of the rarity of the condition: when there are very few individuals with a condition, it is easier to see them as individuals rather than as anonymous members of a group of patients. But for policy purposes, it is important to recognize that the ethical impact of identifiability is often enhanced when a rare condition produces visible signs of illness and when individuals are publicized through photo campaigns and telethons.

Publicity can be a powerful and important tool for advocacy groups, but it is not an appropriate ethical justification for coverage of particular orphan drugs over others. Likewise, being able to identify individuals because of the rareness of a condition is—by itself—not a strong ethical rationale for spending more resources for some individuals than for others. And notable physical signs of illness, if primarily cosmetic, are equally unsuitable to serve as a strong guide to resource allocation. Many outward signs of illness, however, are related to a signifcant impact on the quality or quantity of life and on that basis should be considered as highly relevant in allocation decisions. We take up these implications in a subsequent section discussing an ethical criterion of “endangered lives.”

In our personal morality, we have greater obligations to known individuals—for instance, to our children and friends—than to individuals unknown to us. If one fails to act, the failure is morally culpable. Tom Beauchamp and James Childress state that the obligation to rescue exists if five criteria are satisfed:

1. Y is at risk of signifcant loss or damage to life or health or some other maj or interest.
2. X's action is needed (singly or in concert with others) to prevent this loss or damage.
3. X's action (singly or in concert with others) has a high probability of preventing it.
4. X's action would not present signifcant risks, costs, or burdens to X.
5. The beneft that Y can be expected to gain outweighs any harms, costs, or burdens that X is likely to incur.[14]

Although this may clarify some aspects of the role of identifability in our personal morality, personal morality cannot be neatly transposed on the public sphere. Providing orphan drugs to all identifable rare disease patients would violate the condition that rescue does not present significant costs, risks, or burdens. Although the effect on any particular individual may be small, the aggregate cost will be greater morbidity and perhaps even mortality than would result if the resources spent on rescue were devoted to other, more cost-effective health care services. Thus, there is no apparent obligation to rescue identifable rare disease patients based on a duty of rescue within personal morality.

A contractualist theory—one that appeals to the idea of a social contract between all those in a moral domain—might provide a counterpoint to this conclusion. If citizens are prepared to put greater resources toward the rescue of identifable individuals, then decision-makers should respect this desire. There are several reasons to think the public would support this. First, studies have repeatedly found that the general public prefers to give preference to patients with life-threatening or severe illness, even if this approach leads to a less cost-effective allocation of resources (although this is not universally true).[15]

Second, the literature suggests people desire reassurance that they live in a compassionate society. The “symbolic value argument” asserts that by spending more on the rescue of an identifed few than would be dictated by a utilitarian allocation strategy, a society demonstrates the value it places on human life and promotes social utility[16] But if our sympathy is engaged by several identifable individuals suffering and dying, should we not be moved even more by the suffering and death of numerous unidentifed individuals? It is not unambiguously true that individuals under the Rawl-sian “veil of ignorance” would choose the rule of rescue as a moral principle.[17] Although the benefts of social reassurance should not be wholly discounted, it strains credulity to say that the more caring society is the one that sacrifces several anonymous lives in order to save an identifable one.

Finally, fairness requires that we not discriminate on morally irrelevant grounds. No person should receive social benefts on the basis of undeserved advantageous properties or be denied them on the basis of undeserved disadvantageous properties. Here, ‘“undeserved” means distributed by the “lotteries of social and biological life”: it refers to properties that individuals cannot be responsible for either having or lacking.[18] Such traits do not provide morally relevant grounds for discrimination between persons because not all people have a fair chance of acquiring or overcoming them. This makes allocation of resources on the basis of identifability ethically problematic. For rare disease patients, identifability often results from undeserved properties, both advantageous and disadvantageous. As a result, they should not receive any preference in health resource allocation because they are identifable. Though all forms of identifability are problematic, it is particularly important to guard against using purely cosmetic signs of illness or strong fears elicited by certain illnesses, such as cancer, as the basis for distributing resources.

In summary, although identifability may elicit a rescue impulse and sometimes entails moral obligation at the level of the individual, no persuasive rationale exists for using identifiability in resource allocation at the policy level, and indeed, strong ethical arguments can be made against it.

Endangered Lives

In addition to being identifable, those who elicit the rescue impulse in others are “endangered.” By some accounts, the endangered face imminent death; others have suggested that the rule of rescue may encompass people who merely have diminished quality of life or are in immediate peril of a severe nonfatal harm.[19] Any of these formulations suggests that rare disease patients may evoke an ethically powerful sense of solidarity with the weakest among us that should be viewed as distinct from identifability. Although orphan drugs are not inherently superior to best supportive care, some orphan drugs save lives that would otherwise be lost, and others restore or maintain capacities central to individuals’ functioning in society. Examples of such capacities include—but are not limited to—hearing, eyesight, and the ability to care for oneself independently or to pursue meaningful employment. When orphan drugs offer such signifcant health gains, paying a premium for orphan therapies is justifed to ensure fair benefts for rare disease patients.

Prioritarianism is an ethical argument for favoring the worst off because additions to welfare matter more the worse off the affected person is.[20] Some would define being worst-off as currently lacking a valuable good, such as health; with that understanding, prioritarianism might be called “sickest first.” Empathy for the worst-off is an important trigger for the moral impulse to rescue, and sickest first is ethically relevant to orphan drug coverage decisions. In practice, however, a sickest-frst principle might require allocation of resources even when only minor gains can be achieved and the cost is very high, which is obviously inefficient. This principle has also been criticized for presuming implicitly that resource scarcity is temporary and that individuals with progressive diseases can and will be saved in the future, which is obviously a flawed assumption. Unfortunately, budgetary limitations on one hand and an increasing number of orphan drugs on the other suggest that resource scarcity will persist. For these reasons, when sickest first is applied to orphan drug coverage decisions, it is necessary to go a step further and explicitly incorporate posttreatment prognosis. A bright line between what constitutes a fair claim on health benefts and what does not will be difficult to draw. Generally, however, lifesaving orphan therapies and therapies that restore or maintain capacities central to functioning in society should be covered. Orphan therapies that do not achieve these health outcomes clearly should not. For example, interventions that prolong life by only a few weeks on average likely do not merit coverage at extraordinary expense, whereas a drug like Cyste-amine, which allows cystinosis patients to survive for years without dialysis and forestalls the need for renal transplant, does.[21] In order to demarcate this line fairly and transparently, it will be necessary to evaluate the probability and magnitude of the expected beneft associated with any given orphan drug along predetermined dimensions and in relation to predetermined thresholds. In some cases, strict clinical criteria based on data showing that certain individuals beneft more than others may define coverage eligibility within a rare disease population.

Health resources should be allocated so as to provide the sickest members of an insurance pool—the neediest amongst us—with meaningful health gains when possible. The difficulty lies in finding the appropriate balance between doing a little good for many people and doing a lot of good for a few. Therefore, coverage decisions must not only incorporate consideration of the benefts gained but the opportunity costs incurred when covering expensive orphan drugs.

Opportunity Costs

When motivated by the rule of rescue, decision-makers seem to grow less mindful of what is foregone if the rescue is funded. To some extent, advocates of blanket coverage for orphan drugs have tried to justify this lack of concern. For instance, the National Organization for Rare Disorders’ official policy position on insurance reimbursement for orphan drugs states: “insurers must recognize that the small number of people using the product in any MCO [managed care organization] will usually represent a small overall expense to a health insurance company”[22] This, however, is a presumption about the facts, not an argument, and the factual context is rapidly changing as more orphan drugs enter the research pipeline and come to market. Therefore, opportunity costs must be explicitly included in any coverage decision.

In the United Kingdom, cost-effectiveness analysis is openly used to make coverage decisions. While it may cost the National Health Service “only” 2.5 million pounds to provide a hypothetical orphan drug, that same expenditure would pay for more than 520 hip replacements.[23] The United Kingdom, unlike the United States, has a fixed budget for health care, and opportunity costs are therefore “real.” Within the fragmented U.S. system, it is not so clear that money saved by denying coverage of orphan drugs will be used elsewhere to improve health. Recognizing this limitation, it is essential to find a way to quantify the opportunity costs associated with coverage of expensive orphan drugs within an insurance pool or health system, to communicate these clearly to stakeholders, and to ensure that what is foregone is acceptable. Limiting care of others to cover expensive orphan therapies will, in the aggregate, likely result in greater morbidity or mortality; acceptability is achieved when the benefts to rare disease patients are seen to outweigh or justify these costs to others. Such transparency will serve as a mechanism for ensuring that the burdens of resource scarcity are consistently distributed.

Justifiable Rescues

Many arguments and explanations have been made for deviating from a purely utilitarian allocation of health resources in order to ensure that orphan drugs are accessible to rare disease patients. For those who advocate a departure from maximizing health gains across the entire population, the rule of rescue has been cited as an infuential motivation. This has been true when individuals are asked if expensive orphan drugs should be covered, but it also pertains at a policy level. Yet as a policy, unrestrained rescue is unsustainable. Having looked at identifable individuals, endangered lives, and opportunity costs, it is clear that these constituent parts of the rule of rescue contain important moral clues not only to why the rule of rescue is intuitively appealing, but also as to how it can be ethically restrained.

Discussions of the rule of rescue often focus on the powerful ethical instinct to aid identifable patients. This is unfortunate, as there is no ethically sound argument for allocating resources on the basis of identif-ability Yet the rescue impulse is not elicited by identifability alone; it is also a response to individuals who are threatened with death or signifcant harm. It is ethically appropriate to ask what can be done for these individuals and to consider covering interventions that, although expensive, offer a meaningful opportunity to save lives or to maintain or restore capacities central to functioning in society. Opportunity costs must also be weighed in orphan drug coverage decisions to ensure that an undue burden is not placed on others for the sake of a few. By shifting the discussion to focus on these two elements of the rule of rescue, it is possible to justify giving priority consideration to some—though not all—orphan therapies.

Consider how these criteria might guide coverage and reimbursement decisions concerning Cerezyme for the treatment of Gaucher's disease. While Gaucher's disease is seldom fatal, does not produce notable cosmetic effects, and receives little publicity, it signifcantly impacts patients’ quality of life; symptoms include anemia, extreme fatigue, and bone pain. In severe cases, patients may experience cognitive deterioration, seizures, or loss of muscle coordination. Treatment with Cerezyme produces dramatic improvement in the well-being of most patients. Cerezyme is very expensive, with treatment often costing $200,000 or more annually, but since so few patients have the disease, the overall impact on health budgets for large insurers is relatively small.[24] In the framework set out above, paying a premium for Cerezyme would be reasonable: the identifability of its patients is not driven by cosmetic or publicity-related factors; the quality of patients’ lives is truly endangered by the severity of the illness; the treatment produces dramatic improvements; and the opportunity cost of treating a relatively small number of patients at high cost is manageable.

But the decisions for other orphan drugs might turn out differently. The scenario associated with cetuximab to treat nonsmall cell lung cancer is very different, for example. The identifiability of patients with nonsmall cell lung cancer is driven largely by the heightened public consciousness surrounding lung cancer. Lung cancer is often lethal, but the marginal benefts of cetuximab are quite modest. The average survival advantage from adding cetuximab to the standard treatment regimen is approximately five weeks.[25] Cetuximab treatment is also associated with higher frequencies of rash, diarrhea, and febrile neutrope-nia, a condition that increases the risk of infection. And, lastly, cetuximab is expensive at both the individual and population level. Although nonsmall cell lung cancer is technically a rare disease, 60,000 patients are diagnosed with the illness each year in the United States.[26] The treatment costs for each individual patient average approximately $80,000, which translates into an expenditure of $4.8 billion dollars per year. Cetuximab is covered by public and private insurers in the United States, but our ethical framework would draw a sharp contrast between this drug and Cerezyme. Considering the sources of identifability, the marginal impact on the length and quality of life, and the implicit opportunity costs of this level of expenditure, our framework would suggest that public and private insurers would be justifed in refusing to pay for cetuximab.

This conceptual framework offers an advance over current decision-making practices because it allows for more transparent, more consistent decision-making and protects against the ill-advised and unethical extremes of a rescue mentality, yet it also leaves room for the possibility that expensive orphan drugs should be covered. Plainly, more research is needed to implement the framework. This will include empirical and conceptual research on the nature of opportunity costs, as well as clinical research to measure the benefts of orphan drugs and comparative effectiveness research to determine if comparable benefts can be achieved by other, less expensive means. Further considerations beyond the rule of rescue may be relevant to orphan drug coverage decisions. For example, conversations about resource allocation may have the effect of starting a public dialogue about the “fair innings” argument, which claims that people are entitled to some “normal” number of life years and generally favors allocating resources to the young.[27] Additionally, it is certainly possible to pay less for orphan drugs than current market prices dictate, though payment policies must be carefully constructed so as not to stife innovation. Because social preferences and budgetary constraints will vary by time and location, this framework is just that—a framework—and not a static solution to orphan drug decision-making.

Given trends on the research and development horizon, it is clear that ever more common conditions will be divided into subpopulations that may beneft from tailored therapy. The pharmaceutical and biotech industries are eager to meet this demand, and industry clearly perceives the optimization of drug therapy in accordance with patients’ genotypes both as a huge medical leap forward and as a high-value proposition.[28] The resource allocation problems posed by orphan drugs are therefore a harbinger of things to come. Tomorrows medical care will feature a growing number of expensive therapies that offer beneft only to small populations. As with todays orphan drugs, identifability should not be a factor in decision-making. Rather, the potential health gains must be evaluated in context to determine whether they provide a meaningful beneft beyond what is already available, and the opportunity costs must be weighed to determine whether they are acceptable. Since decisions not to cover those therapies will always be difficult and uncomfortable, the factors infuencing those decisions must always be transparent.[29]

Disclaimer

The opinions expressed are those of the authors and do not refect the position or policy of the National Institutes of Health, the Public Health Service, or the Department of Health and Human Services.

Acknowledgments

The authors would like to thank Alan Wertheimer and Franklin Miller for their helpful feedback on an earlier version of this paper.