Recognizing Bedside Rationing: Clear Cases and Tough Calls

What Is Bedside Rationing?

Bedside rationing is the withholding by a physician of a medically beneficial service because of that service's cost to someone other than the patient. Three conditions must be met, in our view, before a physician's action qualifies as bedside rationing. The physician must 1) withhold, withdraw, or fail to recommend a service that, in the physician's best clinical judgment, is in the patient's best medical interests; 2) act primarily to promote the financial interests of someone other than the patient [including an organization, society at large, and the physician himself or herself]; and 3) have control over the use of the medically beneficial service.

An example helps show what types of actions qualify as bedside rationing.

“A patient arrives at his local emergency department with the classic signs and symptoms of acute myocardial infarction. The emergency department physician decides to administer thrombolysis with streptokinase rather than tissue plasminogen activator even though the latter is slightly better for this type of heart attack. Tissue plasminogen activator costs 10 times as much as streptokinase, and the physician thinks that the benefits of this therapy are not worth the additional cost.”

This is an example of bedside rationing because the physician withheld a service that would have been in the patient's best medical interests; tissue plasminogen activator would have been a better treatment for this patient than streptokinase [11]. In addition, the physician had control over the resource and withheld it because of a desire to save money for society as a whole. This is a relatively clear example of bedside rationing, but other cases are less clear. Thus, we elaborate on these three conditions through a series of additional examples. In some cases, we provide examples that fail to meet one or more of the conditions to emphasize the importance of each condition. In these cases, we propose changes that would make the cases qualify as examples of bedside rationing.

Condition 1: The Physician Withholds a Service That Is in the Patient's Best Medical Interests

For a physician's action to qualify as bedside rationing, it must first and foremost be an example of health care rationing. Although there is no single accepted definition of health care rationing [8], for our purposes, we consider health care rationing to be the withholding of a service that is in the patient's best medical interests. Defining the first condition in these terms helps physicians decide whether withholding a service is an example of health care rationing of any kind. (The other two conditions help the physician decide whether the rationing in question is bedside rationing or some other form of rationing.) The case of thrombolysis used to treat myocardial infarction, described above, is an example of health care rationing because the physician withheld a service, the administration of tissue plasminogen activator, that was in the patient's best medical interests. If streptokinase had been better than tissue plasminogen activator, then the case would not be an example of health care rationing.

Treatment of Mild Hypertension

“An otherwise healthy 42-year-old man presents to his physician with mild hypertension. He has been dieting and exercising for 6 months, but his blood pressure is still elevated. His physician institutes therapy with generic hydrochlorothiazide. The physician is chided by her colleagues for using this old, inexpensive medicine when newer antihypertensive agents are available. The physician feels uncomfortable because she knows that she prefers this older agent partly because it is less expensive than the newer ones.”

At first glance, this case appears to be an example of bedside rationing: A physician prescribed hydrochlorothiazide rather than a new antihypertensive agent because hydrochlorothiazide was less expensive. However, although this physician considered costs, she did not ration at the bedside, because the more expensive alternatives were no better than the less expensive one. Newer, more expensive antihypertensive agents have not been shown to be better than hydrochlorothiazide at controlling mild hypertension [12, 13], nor have they been shown to have better side effect profiles. Thus, the physician should not have felt uncomfortable about considering the costs of treatment. Instead, she should have recognized that she eliminated waste by offering an equally beneficial and less expensive medicine.

Many persons argue that physicians do not need to ration health care but merely need to eliminate waste in the medical system [2]. This assertion may or may not hold up under close scrutiny, but physicians should certainly eliminate waste whenever possible. Considering costs when services are of equal benefit is morally praiseworthy because it can provide other needed services in the context of limited resources without producing additional burdens. Considering costs is not necessarily bedside rationing.

One can easily imagine a change in this case that would make it meet the first condition of bedside rationing. Suppose that the patient had been diabetic. In this case, there would have been good reason to think that angiotensin-converting enzyme inhibitors, a more expensive class of antihypertensive medicine, would be better for this patient than hydrochlorothiazide because of their beneficial effects on kidney function [14] and their lower potential for aggravating glucose intolerance.

Although providing an equally beneficial but cheaper service is clearly praiseworthy, it is not always easy to know which medical options are in a patient's best medical interests. Many health care services have not been proven to be better or worse than others. When such services are involved, it is up to physicians to do their best to evaluate the available evidence. If evidence is lacking or if a service is of disputed benefit, it is often morally justifiable to use cost as a reason to choose one treatment over another. One needs to recognize, however, that even judgments about the strength and quality of evidence are value judgments and are influenced by culture and context, including cost constraints [15]. These judgments may be a particularly well-hidden form of rationing because they are disguised as “medical” or “objective” determinations. What counts as justification for a particular alternative may also depend on the nature of the choice that is confronted. Forced, high-stakes choices, such as promising but unproven treatments for the acquired immunodeficiency syndrome, may be made on the basis of weak evidence if strong evidence is lacking [16].

Physicians themselves may differ in their skepticism about new treatments, their judgments about evidence, and their assessments of the importance of the patient's medical interests that are at stake. Thus, judgments about benefit are likely to vary among physicians. A caution, however, is in order and is illustrated in the next case.

Prostate-Specific Antigen Testing

“A 60-year-old asymptomatic man visits his primary care physician and asks whether he should have the ”blood test for prostate cancer.“ The physician describes the risks and benefits of prostate-specific antigen testing and the controversy surrounding its use. The patient says, ”Doctor, I'd still like to get the test. My closest friend just got diagnosed with prostate cancer and has it all over his body. I'd rather go through anything than have that happen to me.“ Although the physician thinks that, in general, the risks of this testing outweigh its benefits, he knows several urologists and primary care physicians who routinely use it for screening.”

Where reasonable physicians would disagree, the patient's preference may influence the physician's judgment about whether the service would be medically beneficial in a given case [17, 18]. Physicians (and other health care providers) are in an ideal position to account for such strongly felt preferences, and they should consider these preferences relevant to judgments about risk versus benefit and cost versus benefit. The medical benefit of a service needs to be evaluated from the patient's perspective in individual physician–patient decision making [19, 20]. Refusing to authorize a test, even if the patient would pay for it, abuses the power discrepancy in the physician–patient relationship and fails to respect the patient's values and goals. To prevent these considerations from overwhelming all judgments of clinical effectiveness, however, the request should fit within the range of reasonable medical opinion. Where consensus exists about the potential benefits of a service, requests for particular services should be given less weight. Similarly, cases in which the potential harms of a withheld intervention clearly and uncontroversially outweigh the intervention's benefits (as with exercise stress tests for young, asymptomatic women) are not examples of bedside rationing because the decision to withhold the intervention is based on the balance between risks and benefits and not on cost considerations.

Considering costs is thus a necessary but not a sufficient condition for bedside rationing. For rationing to occur, a medical benefit that is valued by the patient must be withheld. Despite the difficulty of always knowing what the best alternative is for a given patient, the moral point should remain straightforward. Bedside rationing does not occur unless there is good reason to think-taking into account the physician's best clinical judgment and the patient's preferences and values-that the patient has not received a service that would have been medically beneficial. One way to address this question in the clinical setting is to ask oneself, “Would I be willing to provide or authorize this service if it were free or if the patient were paying for it himself or herself?” In this way, one can at least separate cases in which the harms outweigh the benefits from cases in which the benefits do not seem to be worth the costs. In addition, however, one must be open to patients' preferences and must recognize that one's own medical opinion is open to challenge where controversy or disagreement exists among physicians.

Condition 2: The Physician Acts Primarily To Promote the Financial Interests of Someone Other Than the Patient

According to the first condition, health care rationing occurs when a service that is in the patient's best medical interests is withheld. In such cases, consideration of the second and third conditions helps to determine whether the rationing is bedside rationing or another form of health care rationing.

Antibiotic Treatment of a Urinary Tract Infection

“A 36-year-old woman presents to her physician with urinary frequency and urgency. Urinalysis shows leukocytes and many bacteria. The patient has no known allergies but has never taken sulfa drugs. The physician prescribes a 3-day course of trimethoprim-sulfamethoxazole, even though ciprofloxacin, an expensive broad-spectrum antibiotic, would have a better chance of curing the patient and a smaller risk for side effects [21]. She chooses this treatment primarily because of concern about increasing antibiotic resistance.”

This case meets the first condition of bedside rationing: a service that was in the patient's best medical interests was withheld. It does not, however, meet the second condition. Ciprofloxacin was withheld not for financial reasons but to prevent the development of resistance to an important broad-spectrum antibiotic agent [22]. To maintain the effectiveness of broad-spectrum antibiotics in patients with severe illnesses, it is necessary to use less potent antibiotics for common conditions, such as the one afflicting this patient. Thus, as this case shows, the second condition of bedside rationing is important because many of the decisions that physicians make are intended to promote public health or other important objectives rather than the financial interests of someone other than the patient. These decisions do not qualify as examples of bedside rationing because they balance the nonmonetary benefits and harms accruing to others against the medical benefit to the patient. Instead, these decisions qualify as other forms of health care rationing.

In this case, the choice of trimethoprim-sulfamethoxazole is an example of noneconomic rationing. In noneconomic rationing, the patient does not receive the best medical treatment for a nonfinancial reason. In this case, the reason was the promotion of public health. Another example of noneconomic rationing is the decision to withhold scarce transplant organs from some patients because those patients are less likely than others to benefit from transplantation [23]. The next case is an example of an economic form of health care rationing that does not qualify as bedside rationing.

Out-of-Pocket Costs of Antihistamines

“A 20-year-old college student presents to her physician with seasonal allergic rhinitis. The physician discusses treatment options with the patient and recommends a trial of antihistamines. He mentions that inexpensive, over-the-counter antihistamines are often effective for allergy symptoms but can cause drowsiness. More expensive antihistamines that do not cause drowsiness are available. Because the student has no prescription coverage, she asks to try the less expensive antihistamine first.”

Although cost was a factor in this physician's decision about treatment, the decision was clearly made primarily to promote the patient's interests by reducing her out-of-pocket expenses. Physicians need to remember that patients' interests are not purely medical. In addition, this decision was made primarily by the patient: The physician explained the relative merits and drawbacks of the treatment alternatives, and the patient was able to weigh the cost-worthiness of the more expensive medication using her own valuation of costs and benefits. Thus, this is an example of rationing, but the decision was not made by the physician and was not made with interests other than the patient's in mind.

Imagine that the physician had prescribed the less expensive medicine without discussing the more expensive alternative. If that had occurred, the physician may have been motivated by the patient's financial interests, but it would be unclear whether those interests had been served, because some patients might be willing to spend more money to avoid possible sedation. The only way to know what a patient wants is to discuss the alternatives with the patient. Similarly, patients often make their own rationing decisions about services that are available only at a great geographic distance. Here, the “cost” of travel and inconvenience may tip the balance against an otherwise medically beneficial service. Discussions about cost when patients care about cost are an integral and important part of informed consent [24].

If this patient had not had out-of-pocket expenses or had had a copayment that did not vary with the cost of the medication and if the physician had offered only the less expensive medication without discussing the alternative treatment, this case would be a clear example of bedside rationing. The physician would have judged that the additional benefit of the more expensive drug was not worth the extra cost. This decision may be justifiable if, for instance, the patient had known that physicians make such judgments as part of a health care plan or had known that physicians are expected to follow established guidelines for the use of expensive medications. It may also be justifiable if the physician had told the patient, “There is a more expensive medication that causes less drowsiness, but it is much more expensive and, in our managed care plan, we reserve it for truck drivers and others whose need for it is more pressing.” The patient could choose to pay for the medication herself or could explain that she, too, has a special or pressing need. Physicians in these situations need to be sensitive to the power discrepancy inherent in the physician–patient relationship. Physicians, because of their knowledge and power, can influence patients' decisions and even the way that patients express their values. It is important to recognize that when the patient makes the value judgment, bedside rationing does not occur; in bedside rationing, the physician makes the value judgment that a service is not worth its cost.

Condition 3: The Physician Has Control over the Use of the Medically Beneficial Service

Hospital Control of Contrast Agents

“A physician orders intravenous pyelography for a patient with no known allergies. The patient is injected with a high-osmolar contrast agent and has moderate nausea. Later, the patient learns that low-osmolar contrast agents are less likely to cause this uncomfortable side effect [25]. The physician is sorry that the patient had the side effect, but hospital policy had precluded her from ordering low-osmolar contrast agents for a patient without a high risk for a serious adverse reaction.”

This case meets the first condition: A service that was in the patient's best interest was withheld. It meets the second condition because, in general, policies such as the one maintained by this hospital are in place to conserve societal resources [26]. However, it does not meet the third condition. This physician did not have control over use of the low-osmolar contrast agent. Instead, it was the hospital that limited the use of the better, but more expensive, contrast agent.

Rationing often occurs outside of individual physician–patient encounters. Formulary committees, utilization reviewers, and third-party payers can all limit physicians' actions [7]. Physicians may even play crucial roles as sources of information or expertise in these organizational rationing decisions. For example, many physicians work on hospital formulary committees. Although formulary committees must think about the moral implications of their decisions [27], physicians working on formulary committees are not bound by the same moral duties that apply when they work directly with patients. Similarly, physicians may work in government-designed rationing plans, such as Oregon's Medicaid experiment [28]. Physicians who limit the use of expensive medicines through formulary committees are not involved in bedside rationing but are making population-based, organizational-level rationing decisions that may later influence what is available for their individual patients. Similarly, physicians who follow policies established by governments or organizations are not rationing at the bedside.

Use of Scarce Magnetic Resonance Imaging Slots

“A neurologist works at a county hospital that does not have a magnetic resonance image (MRI) scanner. The hospital puts money aside each year so that six patients can receive an MRI at a nearby hospital. A physician evaluates a patient who has a ”soft indication“ for an MRI. The physician could order an MRI for the patient. However, he knows that if he requests an MRI for this patient, he denies an MRI to another patient, who may need it more. Thus, he tells the patient that an MRI is unnecessary.”

At first glance, this case does not appear to be an example of bedside rationing. It is not, one might argue, the price of the service that prevented this physician from ordering an MRI but rather the scarcity of time slots, the “absolute scarcity” of this service. In addition, this physician does not appear to have control over the service, because the hospital limits the number of MRIs that can be ordered in a year. However, these appearances are deceptive. In fact, it is the physician who decides when and whether to order the MRI; thus, he has control over the decision. The decision, indirectly, conserves services (scarce MRI “slots”) in the interests of other, presumably needier patients. Although it is not dollars per se that are being saved, restrictions on numbers of services-or intensive care beds, or staff members-are ultimately made for financial reasons. Money is, after all, merely the medium of exchange for all health care (and other) resources.

One could argue that these MRI slots, like intensive care beds or solid organs, are an absolutely scarce resource, whereas money is a relatively scarce resource. This distinction, however, is false. Resources are always limited. They may be stringently limited, as in this case, or more gently limited (as they would have been if this hospital had had its own MRI equipment and could have scanned hundreds of patients per year).

Thus, withholding and not recommending a potentially beneficial MRI counts as bedside rationing. It may be morally justified but should not be dismissed as a noneconomic decision. This example illustrates the importance of “economic honesty.” Only openness about these kinds of choices will allow discussion of their merits and morality to proceed.

Clear Cases and Tough Calls

As the above cases show, tough and often subtle calls need to be made when one tries to decide whether a clinical decision qualifies as an instance of bedside rationing. It is not always easy to know what is in a patient's best medical interests. Medical data about the risks and benefits of many health care services are frequently unclear, and assessment of evidence is itself a value-laden process. Nor is it easy to know what is in a patient's best financial interests when patients may not know their copayment responsibilities. Moreover, physicians frequently make decisions that have implications for public health.

There will always be tough cases, but it is helpful for physicians to be able to recognize when their actions definitely qualify or do not qualify as instances of bedside rationing. Figure 1 summarizes how physicians can recognize when the withholding of a service might qualify as an example of bedside rationing. Physicians should ask themselves the following three questions.

Is the service that is being withheld in the patient's best medical interests? If a withheld service is not in the patient's best medical interests, no rationing of any type has occurred. If the service is clearly in the patient's best medical interests, then the case involves some form of health care rationing. If the patient's best medical interests are unclear, the question of whether the case involved rationing is also unclear. In such cases, physicians should ask themselves the next two questions (Figure 1) to avoid missing any cases that qualify as instances of bedside rationing. It is better to err on the side of over-identification so that physicians can try to decide whether their actions in particular cases are justified.

Is the service being withheld primarily to save money for someone other than the patient? If physicians withhold a service to promote public health or to pursue other nonmonetary goals, they are not engaged in bedside rationing. Similarly, if a patient chooses a less expensive option because of its cost, the rationing is being done not by the physician but by the patient. Physicians may want to pay attention to this “self-imposed” rationing. For example, physicians may choose to take a more active role in influencing health care or organizational policies that cause patients to have high out-of-pocket costs.

Is the service in question under the physician's control? If a physician's use of a service is limited by, for example, administrative mechanisms, then the withholding of that service is not an example of bedside rationing but an example of rationing through administrative mechanisms. If the physician has complete control over use of the service, then the decision to withhold the service is an example of bedside rationing. In many cases, how much control the physician has is unclear. A physician may need to spend time to get approval for a service, and sometimes such an inordinate amount of time is necessary that it becomes nearly impossible for the physician to obtain the service. At that point, the physician needs to decide whether to spend that time.

In cases in which the answer to all three questions is “yes,” physicians have rationed at the bedside. In these cases, physicians need to seriously consider whether their rationing decisions are justified. The medical profession has yet to decide whether bedside rationing is ever morally justified, and, if so, under what circumstances. Physicians may indeed have an important role to play in limiting the use of marginally beneficial services because they are able to consider patients' individual characteristics and preferences (clinical guidelines, for example, cannot do this).

In cases in which the answer to one or more of the questions is “no,” physicians do not need to worry that they have participated in bedside rationing. In cases in which the answers to one or more of the questions are unclear-for example, if the best medical interests of the patient are unclear-then whether the physician has participated in bedside rationing is not obvious. Physicians should take cases in this gray area as seriously as if they definitely involved bedside rationing, and they should carefully consider whether their decisions are morally justified.

Only by recognizing what counts as bedside rationing can the next, more controversial step be taken: discussion among physicians, ethicists, patients, and society about the circumstances of, constraints on, and justifications for bedside rationing.

Dr. Goold: Department of Internal Medicine, University of Michigan Medical Center, 1500 East Medical Center Drive, 3116 Taubman Center, Ann Arbor, MI 48109-0376.