Do Physicians’ Legal Duties to Patients Conflict with Public Health Values? The Case of Antibiotic Overprescription
- First Online:
- Cite this article as:
- Coleman, C.H. Bioethical Inquiry (2009) 6: 181. doi:10.1007/s11673-009-9155-4
- 219 Views
Among the many explanations for antibiotic overprescription, some doctors cite the risk of malpractice liability if they deny a patient's request for an antibiotic and the patient's condition worsens. In this paper, I examine the merits of this concern—i.e., whether physicians could, in fact, face malpractice liability for refusing to prescribe an antibiotic when, from a public health perspective, the use of the antibiotic would be considered inappropriate. I conclude that the potential for liability cannot be dismissed entirely, but the risk is remote—even in cases where there is a chance that the antibiotic might have benefited the patient.
The overprescription of antibiotics is an important contributing factor in the development of antibiotic resistance, a serious public health problem that has recently become even more urgent with the emergence of multi-drug resistant tuberculosis. According to some studies, at least half of all antibiotic prescriptions are “unnecessary, poorly chosen, or incorrectly dosed” (Solomon et al. 2001). In addition to prescribing antibiotics for illnesses that are not caused by bacteria, physicians frequently prescribe antibiotics for conditions that, although bacterial in nature, are likely to clear up quickly even if no treatment is provided. Even when an antibiotic prescription is medically warranted, doctors may fuel resistance by writing prescriptions for broad-spectrum antibiotics capable of fighting multiple bacteria, instead of waiting for the results of a culture test that would enable them to prescribe a narrow-spectrum antibiotic targeted to the patient’s particular infection.
Among the many explanations for antibiotic overprescription, some doctors cite the risk of malpractice liability if they deny a patient’s request for an antibiotic and the patient’s condition worsens (Selgelid 2007). In this paper, I examine the merits of this concern from the perspective of malpractice law in the United States. Specifically, I consider whether physicians could, in fact, face malpractice liability for refusing to prescribe an antibiotic when, from a public health perspective, the use of the antibiotic would be considered inappropriate. I conclude that the potential for liability cannot be dismissed entirely, but the risk is remote—even in cases where there is a chance that the antibiotic might have benefitted the patient. Nonetheless, if physicians believe they could be held liable, this perception is likely to affect their behavior, even if it is grossly disproportionate to the risk they actually face.
From a public health perspective, when physicians think about the legal implications of prescribing antibiotics, they should be motivated to act cautiously. To the extent that fears of liability are pushing physicians in the opposite direction, policy-makers should adopt counterbalancing legal mechanisms that would create incentives for physicians to prescribe antibiotics more judiciously. I close the paper by offering a few preliminary suggestions about how this might be done.
The overprescription of antibiotics can be defined as the prescription of antibiotics in excess of legitimate clinical need. The most obvious example of overprescription is the physiologically futile use of antibiotics, such as when a physician prescribes an antibiotic to treat a viral infection. In these situations, the antibiotic will simply not work. Overprescription can also occur when an antibiotic might provide some medical benefit, but the expected benefits are small and are outweighed by the potential promotion of resistance. An example of this second category would be an antibiotic prescription for an ear infection likely to clear up quickly without treatment.
Numerous causes of antibiotic overprescription have been identified in the literature (see Laxminarayan and Malani 2007). Patient-related factors include misunderstanding about how antibiotics work, the attitude that a visit to doctor is not productive unless it ends with a prescription (Howard 2007), an unwillingness to delay treatment until culture tests are available, and the fact that the primary risk of overprescription—the development of drug-resistant strains—is a societal problem that may never directly affect the individual receiving care. For physicians, overprescription often stems from a desire to placate worried patients (or, in the case of children, who make up a large percentage of ear infection patients, their parents). As a recent report on antibiotic overuse observed, “The average patient suffering from a cold or an ear infection wants immediate relief and sees a prescription for antibiotics as the ticket to recovery, and the physician may be only too happy to oblige if writing it benefits her practice” (Laxminarayan and Malani 2007, 3–4). The economics of antibiotic production and distribution do not help matters, as manufacturers’ profits are based on the number of sales, not the medical legitimacy of the prescriptions, and the out-of-pocket costs to purchasers of antibiotics do not reflect the social cost of overuse.
Antibiotic Prescriptions and Malpractice Liability: What are the Real Risks?
Assessing the risk of malpractice liability for refusing to write a physiologically futile antibiotic prescription is easy: there is no risk. First, the “standard of care” cannot plausibly be interpreted to require physicians to provide medically useless treatment simply because the patient asks for it. This is true whether the standard of care is defined in terms of what physicians customarily do (the traditional approach to malpractice liability) or as what a “reasonable physician” would do under the circumstances (a more modern approach used in some jurisdictions that allows juries to evaluate the expert witnesses’ testimony and make their own assessment of what a competent physician would have done under the circumstances) (Peters 2000). Second, if the antibiotic would truly have been futile, a patient who is denied the antibiotic will not be able to satisfy the causation element of a malpractice claim—i.e., the requirement that the patient establish that it is “more likely than not” that her injuries were caused by the conduct about which she complains. No matter how sick the patient becomes after being denied an antibiotic, if the antibiotic had no chance of working, the refusal to provide it could not possibly have caused any harm.
Of course, physicians who express concern about malpractice liability for refusing patients’ requests for antibiotic prescriptions are probably not thinking about cases of true physiological futility. Rather, they are concerned with situations in which the antibiotic might provide some benefit, albeit a small and uncertain one—such as with an ear infection that might clear up a bit earlier if an antibiotic is provided, or a case of diagnostic uncertainty in which the potential benefits of an antibiotic cannot be ruled out (Selgelid 2007, 222). In these situations, if the patient insists on an antibiotic prescription, does the physician who refuses to comply face a genuine liability risk?
In jurisdictions that follow the traditional approach to malpractice liability, this question would be analyzed by asking whether it is customary for physicians to prescribe antibiotics under the circumstances at issue. Given how many physicians are known to accede to patients’ requests for antibiotics even when the potential medical benefits are minimal, it might appear that a custom-based analysis would be favorable to plaintiffs. However, despite the frequency of antibiotic overprescription, the practice has been widely criticized by the organized medical profession (e.g. Gonzales et al. 2001), whose views have a significant impact on the assessment of medical custom. Moreover, even if a plaintiff could establish that many physicians do prescribe antibiotics when the potential benefits are minimal, physicians could still avoid liability in a traditional malpractice jurisdiction by showing that a conservative approach to antibiotics is supported by a “respectable minority” of the profession.1 This should be fairly easy to establish in light of professional statements endorsing conservative use.
The standard of care analysis becomes more complicated in jurisdictions that use the “reasonable physician” approach to malpractice liability, under which juries are given greater leeway to evaluate, rather than merely to identify, what physicians customarily do. At first glance, it might seem that an objective reasonableness standard would be even more favorable to physicians than one that simply defers to customary practice, given that, in the cases we are considering, the potential benefits of antibiotics are minimal and the public health risks are great. However, resistance does not result from a single antibiotic prescription; it takes many prescriptions over an extended period of time until an antibiotic no longer works. Because the incremental risks of a single antibiotic prescription are minimal, a jury might conclude that the reasonable physician would take the risk if there is even a small chance that the patient’s condition would improve.
Further complicating the reasonableness analysis is the fact that the primary risks of antibiotic overprescriptions extend to the public at large, not to the individual seeking care. Because physicians are commonly regarded as fiduciaries of their patients (Rodwin 1995), courts may take the position that, in assessing the reasonableness of the physician’s conduct, the only risks and benefits the jury should consider are those the individual patient would experience. If the public health risks of antibiotic resistance are taken out of the equation, a refusal to provide a potentially beneficial antibiotic prescription may well be viewed as “unreasonable” medical care.
Finally, in jurisdictions that rely on the reasonable physician approach to malpractice, the patient’s subjective desire for an antibiotic may play a greater role in the analysis than in jurisdictions that define the standard of care solely in terms of physicians’ customary practices. In general, reasonableness assessments depend on a weighing of the risks and benefits of actions, so the issue in a “reasonable physician” jurisdiction would be whether the potential benefits of an antibiotic prescription outweigh the associated risks. In the area of informed consent litigation, courts have emphasized that identifying and valuing the risks and benefits of medical treatment depends as much on individual values and preferences as on objective medical information (see Canterbury v. Spence). If the patient’s subjective desire for an antibiotic is taken into consideration in identifying the potential benefits of an antibiotic, physicians might be expected to take greater risks for such patients to satisfy the standard of reasonable medical care.
This deference to patient-driven assessments of the risks and benefits of treatment finds some support in the handful of cases in which physicians have sought to withhold or withdraw life-sustaining treatment over the objections of a patient or surrogate. In many of these cases, courts have found that the patients are entitled to treatments they or their surrogates deem subjectively beneficial, even if those treatments seem unjustifiable from an objective point of view (Pope 2007). Read broadly, these cases could be construed as standing for the general proposition that, as long as a treatment has the potential to provide some benefit to the patient (i.e., as long as it is not physiologically futile), determining whether the benefits are “worth it” is for the individual patient to decide.
Despite these observations, it is doubtful that a physician who refuses to prescribe a minimally beneficial antibiotic faces more than a theoretical risk of malpractice liability.2 First, even in jurisdictions that rely on the reasonable physician approach to malpractice, the views of professional organizations remain influential. With a clear professional consensus in favor of conservative use of antibiotics, plaintiffs are likely to have a hard time convincing a jury that a reasonable physician would choose to take a more aggressive approach.
Moreover, while individual autonomy is clearly an important legal value, the relative importance of autonomy vis-à-vis other values depends on the specific situation. Thus, the fact that patients have the right to refuse proposed interventions based on subjective or idiosyncratic value considerations suggests that individual autonomy is paramount when the patient’s bodily integrity is at issue. However, in the antibiotic context the patient is not seeking information to help decide whether to accept a bodily invasion, but is instead seeking to compel a physician to provide treatment the physician believes is inappropriate. The fact that courts have emphasized the subjective nature of risk-benefit assessment in the context of informed consent litigation does not mean they would place an equivalent emphasis on individual autonomy when the patient is seeking to compel treatment, rather than to refuse it.
Admittedly, some of the end-of-life cases do appear to support a broader right to affirmatively insist on treatments, but in those cases the potential efficacy of the treatment was not seriously at issue. Rather, the physicians in these cases were claiming that, even though the treatments were likely to work—in the sense that they would keep the patients alive—the patients were so debilitated that keeping them alive would be pointless. Whether life is worth living seems quite clearly to be a philosophical, not a medical, determination. Thus, rather than reflecting a broad willingness to defer to patient preferences in all situations, the end-of-life cases can be read as standing for the more limited proposition that doctors have no right to tell patients that they would be better off dead.3
More broadly, despite the fact that physicians are normally required to act exclusively in their patients’ interests, there is precedent for allowing physicians to take the interests of non-patients into account. For example, physicians are permitted, and in some cases required, to breach patient confidentiality when a patient poses a direct threat to others (see Tarasoff v. Regents of the University of California). While an antibiotic prescription does not pose an imminent threat to an identifiable individual—the standard courts have generally used for permitting or requiring physicians to breach confidentiality—the third-party disclosure cases demonstrate that courts will sometimes subordinate patients’ wishes to broader societal concerns.
Similarly, the entire field of medical research can be seen as an exception to the general principle that physicians must always promote the individual’s best interests, given that research subjects are frequently exposed to risks from which they will derive no direct benefits (Coleman 2005). To be sure, individuals who participate in medical research usually consent to enter into a relationship in which promoting their individual interests is not the primary objective. However, in the U.S. the law permits limited forms of nonconsensual research when the risks to individuals are minimal, provided the research cannot otherwise be conducted and an independent ethics committee approves it (Department of Health and Human Services 2005). As with the third-party disclosure cases, the permissibility of minimally risky, non-consensual medical research suggests that societal interests can sometimes trump the usual supremacy of individual choice.
Finally, even if a patient is able to convince a jury that the physician acted unreasonably by denying her request for an antibiotic, she must also establish that the denial of the prescription caused compensable damages. If the expected benefits of the antibiotic to the patient are minimal, any damages caused by refusing to prescribe the antibiotic are also likely to be minimal—unless the physician grossly miscalculated the patient’s need for the antibiotic, in which case malpractice liability may well be appropriate. Unless the damages caused by the physician’s conduct are substantial, it is unlikely that a malpractice lawsuit would even be brought.
Changing the Incentives: Legal Measures to Promote Conservative Antibiotic Use
The foregoing discussion suggests that physicians who refuse to provide antibiotics to patients who demand them may face a theoretical risk of malpractice liability, but in the absence of other wrongdoing (such as negligent misdiagnosis) the real liability risk is probably remote. Yet, it is the perceived, not the actual, risk that is likely to drive behavior. In the area of pain management, for example, physicians commonly cite fear of criminal liability as a barrier to aggressively prescribing narcotics (Ziegler and Lorich 2003), despite the fact that criminal actions against physicians who prescribe narcotics appropriately are fairly rare events (Meier 2008).
Theoretically, one way to address this problem would be to educate physicians about the real risk of liability, in the hopes that better-informed physicians will be more likely to resist inappropriate patient requests. Yet, changing physicians’ perceptions about the law is notoriously difficult, particularly when the legal issues are ambiguous and there is no case law directly on point. Moreover, because physicians’ concerns about liability are just a small piece of the overprescription problem, it is not clear how much of an impact educating physicians about the law would actually have.
A better response would be to counterbalance the negative incentives stemming from fear of malpractice liability by creating additional legal incentives inspired by public health norms. For example, legislatures could draw on tools designed to discourage the overprescription of other potentially dangerous medications, such as requirements that antibiotic prescriptions be written on special forms, with copies of the prescriptions sent to state officials (Brushwood 2003).4 They could require general practitioners to consult with specialists before prescribing broad-spectrum antibiotics, or prohibit the prescribing of antibiotics without an in-person consultation.
Measures like these would not necessarily eliminate physicians’ fears of malpractice liability, but they would counterbalance those fears with incentives to prescribe antibiotics more cautiously. Rather than being a barrier to the pursuit of public health objectives, the law could become a tool for aligning physicians’ incentives with important public health goals.
As Barry Furrow explains, “If a defendant has established that distinctive practices are supported by a minority within the professional group, the judge may direct a verdict for the defendant rather than leaving the issue to the jury.” Barry R. Furrow. 1997. Managed care organizations and patient injury: Rethinking liability. Georgia Law Review 31: 419-509, at 486 n.300.
Indeed, “there are no documented cases in the medical literature of a physician being sued for failing to prescribe an antibiotic and a board member of a malpractice insurer reported that he was unaware of such cases” (Howard, 76).
As one commentator has observed, “It is as presumptuous and ethically inappropriate for doctors to suppose that their professional expertise qualifies them to know what kind of life is worth prolonging as it would be for meteorologists to suppose their professional expertise qualifies them to know what kind of destination is worth a long drive in the rain” (Ackerman 1991, 27–28 cited in Pope, 41).
These monitoring programs have been criticized for over-deterring the legitimate use of controlled substances. However, under-prescribing does not appear to be a significant problem in the area of antibiotics (as it is with controlled substances), so the risk of over-deterrence is a much less significant concern.